Good morning. Thank you for joining us today. I have a few quick comments before I turn it over to management. We're handing out the power banks that can charge both your phone and the cell phones, and you should feel free to use them through the event, as well as we have a charging station in the back of the room. You can also take the power banks home with you. We'd appreciate it if you could put your cell phones in silent mode. The company will be making forward-looking statements. We encourage you to review the safe harbor statement. Now I will turn it over to Zvi Lando, CEO and VP of Sales.
Morning. There's a story about the famous violinist, Isaac Stern, that was walking the streets of New York one morning, and he bumped into a tourist who didn't recognize him, and the tourist asked him, "Excuse me, sir, how do I get to Carnegie Hall?" Isaac Stern answered, "Practice." We decided to take a shortcut and have the event here, and we're very happy that you can join us today. My name is Zvi, and I'm the CEO of the company. I've met some of you over the 10 years that I was heading sales and service for SolarEdge globally.
I've been in this role for the last three months, since the passing of our founder, leader, and friend, Guy Sella, that I know many of you have listened to over the years, probably met him, and I'm sure that you remember him with respect and fondness as much as we do. The agenda for the day is such that in the first part of the day, we'll cover our solar business and our position in that market and our plans for that market moving forward. We'll discuss some of the new acquisitions and new markets that we have entered with a special spotlight on the technology of the company and how it fits both in the solar market and in the new markets we entered. We'll close the day with Ronen covering, of course, the financial aspects of the plans of the company.
We took special attention to give you the opportunity to meet a broader representation of the company's management team. You'll see them throughout the day presenting, and you'll have the opportunity to interact with them as well. I'll move quickly to the first item on the agenda, and I'm happy to introduce the Chairman of our Board, Nadav Zafrir, to kick off the day with a few remarks.
Good morning, everyone. Thanks for joining us, thanks to Zvi. I stand here today because Guy Sella, our founder, is gone, I'm both sad and proud to be here with you. I followed SolarEdge from its inception, through Guy's eyes, from its inception in 2006. In the last three months, I've looked at it closely through my own eyes. I have to tell you what you probably already know, it's a remarkable company. It's not surprising for me, since I know the founders of the company for many, many years. As you might know, SolarEdge has its roots in an organization that I personally know very well, the Israeli Defense Forces. Specifically, the special ops and technological units within the intelligence organization of the Israeli military, where I had the privilege to serve for 25 years until 2013.
Many of these years, I served under Guy Sella's leadership, he was my friend and mentor for the last 25 years. One of the best-kept secrets of the intelligence units within the IDF is that it has access to 100% of the population when they graduate from high school through mandatory service. Perhaps the best-kept secret of the intelligence units is that they have the ability to screen 100% of the high school graduates two years before they graduate and try to predict who are the ones that have the highest aptitude to learn extremely fast, work collaboratively to face some of the biggest challenges that Israel faces as a country, sometimes existential needs. It's in this environment that I first got to work with the founders of SolarEdge, three of which, Lior, Yoav, and Meir, are here today and will talk to you after Zvi.
In this environment, I had the pleasure of working with them and with Guy. When they continued on to start SolarEdge in 2006, I remained in the service for a few more years. I became the commander of 8200. As I said, I followed the building of SolarEdge through Guy's and the other founders' eyes over many years. When I retired in 2013, SolarEdge was already up and running. Actually, Guy and the founders guided me and mentored me as I started my own company called Team8. What Team8 does is looks for big challenges in the area of cyber machine learning and data and where they intersect. Try to find deep tech to find sustainable, meaningful solutions for some of these daunting problems.
I read the safe harbor about 10 times yesterday to make sure that I'm not saying anything I'm not supposed to. Rachel, our general counsel, walked me through it again this morning. I do believe that when historians look back at our times, they may remember or refer to Dickens in "A Tale of Two Cities." I really think these are the best of times and the worst of times. I think as humanity, we've never been confronted with such great challenges on the one hand, perhaps existential challenges. On the other, we've never had the scientific ability, the tech depth, which is constantly accelerating, which may solve some of these daunting problems that we're facing. I think what SolarEdge is doing is focusing on one of the most challenging of these problems that humanity is confronting, focusing specifically on solar in sustainability.
I think we have the tech depth, the experience, and the capabilities to solve some of these problems. I believe that we are now going to continue to lead the solar energy, but also have the ability to break into new business segments, as you'll be briefed today. Zvi and the rest of the leadership team, I want to express my appreciation for you and the whole management, to your continued efforts and outstanding results so far, and I'm sure you'll continue to lead in the near future and beyond. Thank you very much and have a great day.
In the next 40 minutes, I'll be covering our status and plans for the solar industry. That will be followed by a presentation about our roadmap for the solar industry and our operational plan. After that, we will go into the new businesses. In discussing our solar business, I'll cover some of the trends of the industry, why we believe this industry is expected to grow, and what are our plans to continue to grow at a higher rate than the growth rate of the industry, as we have done in recent years. I'll describe in detail our growth strategies, and I'll focus a little bit more time on one of our strategies that we coined ARPI, which represents average revenue per installation, and I'll explain that later on during the presentation.
I believe that at the end of my presentation, as well as that of my peers, you'll be familiar with our core capabilities in the solar industry and what got us to the leadership position that we are in today, and that will set the basis to discuss how we use those same capabilities together with the acquisitions that we brought in-house to grow in the new markets that we've entered. SolarEdge has a track record of execution, as evidenced both by our top line and bottom line growth over the last few years. If you look at it in another pattern, looking at the rate at which the industry grew in megawatts installed and the rate at which we grew in inverter shifts, you see that over the last few years, the industry grew by about half the rate that we grew.
Another more specific way to look at this is to take a few select geographies. In this case, we chose some countries that represent a geographical diversification and are all part of our top 10 sources of revenue and looked at each of these countries. What was the growth rate of the solar industry in this country over the last five years, and what was the growth rate of SolarEdge in that country during the same period of time? There is also a notation of the size of that market in 2019, just to give you a feel for what are the sizes of the markets that we are operating in and what are the sizes of the markets where we are outperforming the growth rate of the rest of the market.
The combination of vision, technological innovation, and our execution capability has led us to become the number one inverter company in the world by revenue, as well as the number one single-phase inverter provider by megawatt shift. Here you can see in summary a few numbers that represent the achievements and represent what we've accomplished as a company over the last 13 years since the founders began the company in 2006. Now looking ahead, let's start by looking at the market. There are a few macro trends that are working in our favor in increasing demand for electricity and increasing usage of renewable power in generating that demand. Population of the world is obviously growing.
More of that population is moving into cities, and on a consistent basis, the average energy consumption per capita is going up. This is pretty straightforward and increasing the demand for energy of all kinds. The second trend that works in our favor in this case is the increased usage of renewable energy in supplying this increased demand globally. As you can see, there's projections that by 2050, roughly 30% of the global electrical consumption will be supplied from renewable energy sources. Splitting that down another layer takes a look at the sources for renewable energy generation, and within that, if a few years ago solar was a relatively small portion, it's a portion that is expected to grow at a higher rate, such that by 2030, somewhere in the range of 17% of the renewable energy delivered will be coming from solar sources.
All of these trends combined are obviously leading to an increase in the size of the market that we've been serving for the last years, and the market that is our primary market that we will be serving in the future. Looking at the amount of solar energy installations globally, divided by segments, you can see that all segments are expected to grow, and in general, the market is expected to grow by roughly 30%-35% over the next few years. Within the growth of the market, the segments that are expected to grow at the highest rate are the segments in which we've been active and the segments in which we hold a leadership position of residential, commercial, and industrial installations.
Another way to look at the potential in this market is by looking at how many big markets exist globally, and if big, we chose a arbitrary number of a market of 1 GW of installations per year. You can see that every year, the number of countries that have a 1 GW market is increasing. They're very diversified globally. What's driving this is more countries reaching grid parity and more countries adopting policies that are favoring solar installations. We are active in practically all of these markets, excluding China, and we believe that this trend will continue as the market grows and expands into new countries and new geographies. What are our growth strategies to ride the wave of the growth of the industry and grow even at a higher rate as we've done in the past?
Both historically and today, our growth strategy is based on a four-pronged strategy. The first one is geographic expansion. The second one is segment development. The third one is market share growth. The third one is the same ARPI from my first slide, average revenue per installation increase. I'll spend a couple of minutes on our plans in each one of these areas and trying to tie them together and show how we plan to grow in the coming years. Geographic expansion is pretty straightforward and correlates to what I said before. As you saw probably in previous presentations, we have today installations in more than 130 countries around the world. We have presence of teams of our own in 28 countries around the world.
We typically operate in a country from remote until we recognize that the market is sizable enough, is right for our portfolio, and it's sustainable and expected to remain and grow into the future. Once we are active in a certain geography, we recognize where there is potential and obviously increase the resources and investment in that market. In 2020, there are three markets on top of our regular markets where we intend to focus more as we see a lot of potential for growth, and those are Australia, Brazil, and India. All of them markets of 1 GW in size and more. On top of geographical expansion, we aim constantly to introduce products that bring us to more segments of the industry.
If you look at our position today and try to measure how complete is our offering and how strong is our position in the market, I think it is recognized that in the residential space, we are very strong, if not the leader, and we have a portfolio that we believe is complete and can serve all applications within that segment of the industry. In the commercial industry, we are one of the leaders, and we have a very competitive and compelling portfolio, but we still have more elements to add until the portfolio is complete and we can move to the level of leadership that we aim for. In utility, we are a very small player today. However, we have already a few hundred MW of installations of early adopters using our commercial and industrial products for utility type installations already today.
Lior later on will describe how we plan to add products to put us in a better position in this segment and the industry as well, such that in the future, we will reach a point of having a complete portfolio and strong position in residential, commercial, and industrial, and eventually in the utility segment as well. That was the second prong of our growth strategy. The third is market share growth. This has been, to a large extent, the bread and butter of our growth over the years. As you can see, only five years ago, we were ranked the number 10 inverter company in the world, and for the last two years, we've been consecutively ranked number one.
An illustration of how we work in a specific geography would be our growth rate in the residential North American market and how we've come from a single-digit position to a very high market share today. The methodology for growing market share is pretty straightforward. You need to have the right product, you need to have the right channel to bring them to the market, and you need to have the determination to fight and win every battle and gain market share in that way. This has been a big part of our growth to date, and this has been a big part of our growth plans going forward.
Actually, the more you are present in the market, the more you are close to the customers, the more you are willing to listen to the customers, then you can leverage your innovation to develop new products that are better tailored for the needs of those customers and keep the machine rolling and keep building on one after the other and continue to gain market share. Now to the fourth, the famous ARPI, hopefully famous by now. Historically, a solar installation consisted of inverters, optimizers when we came in, and modules. We were selling the inverters and optimizers into that industry. If you look at the solar installation today, it can contain, on top of the inverter optimizers and modules, batteries, meters, home automation systems, EV chargers, communication devices.
Today, we supply already many of those products to our customers, and in that way, increasing the average revenue for installation or translating it into the ASP for installation. To put some numbers behind this, if we look at 2015. In 2015, we sold roughly 3,000 complex systems. For the sake of the discussion, let's define a complex system as a system that has in it more than just an inverter and optimizer and some form of storage capability or home automation. In 2019, we expect to sell roughly more than 50,000 similar installations or complex installations with additional components beyond the traditional inverter and optimizers. This, of course, translates to more revenue out of each of those installations. Here too, our plan is to continue to introduce more capability, more products, link them to our system, and drive growth through that means as well.
Lior will get into the details of the product that we're referring to. The end result is that if the pie that we were able to take a slice from in 2014 had just the inverters, by now the pie is bigger, and we plan to have access and the ability to generate revenue for more slices of the pie, including smart modules, inverters, batteries, and many other additional smaller devices like home automation and communication. The end result of all of these plans and all of these capabilities is that we expect to be able to grow our revenue consistently year-over-year and make it more diversified geographically and product-wise, and continue to be a leading player in this industry that is growing as well. I'll now hand it over to Lior, who will share with you the product roadmap.
Hi, everyone. My name is Lior. I want to talk about products. Obviously, the most exciting presentation of the day. This is the climax of your day today. What we aim to show is exactly taking the lead from what Zvi said, is that we have a product plan to both increase our offering and deepen our offering and average revenue per installation, both in the residential and C&I space, as well as introduce more products to deepen our early penetration into the utility scale market. We're coupling that with a range of software products, as we've seen that these software products are greatly needed in order to enable the continued expansion and proliferation of solar energy. Let's talk about these energy trends that we have in the world today. The first one is decarbonization. There's more and more renewable energy being sold.
There's more and more renewable energy being installed. Renewable energy comes with a few challenges that require attention as it is more intermittent and less easy to control and may predict. We have the trend of decentralization. Our power generation is becoming distributed. There's more and more distributed energy generation. We have the very strong trend of digitization. Everything becomes connected. Generation systems become connected as well as loads. Everything in our home is now connected. Everything in our building is now connected. That brings a lot of product opportunities.
We at SolarEdge, we are enabling these trends, we are supporting these trends with cost-effective solutions, with hardware and software products to allow the control, management, and prediction of distributed energy generation, and with a range and suite of energy management solutions which A, increase our ARPI, but also enable the growth of solar throughout the various markets. Our products today, we already serve all three segments in the solar industry. We have a full product suite in residential, covering everything from the roof with Smart Modules all the way down to the grid, and even software solutions for grid operators to leverage and manage residential solar systems. We have an expanded suite of products in commercial, and we are building our leadership in that market. We already serve the utility market with Optimizers, inverters for utility scale.
We are in, as Zvi said, still attract early adopters in this market, not yet in the mass scale of our use. With the acquisition of Kokam, we already have utility scale front-of-the-meter storage that is being deployed in various markets. Let's start with residential. The product that you see here, this is our brand new, literally being released into the market these days, the all-in-one HD-Wave inverter. This is a residential inverter that can do both PV generation as well as energy management and storage and backup to the home. Up until now, we've had PV inverters, which were based on our HD-Wave technology, and for storage and backup, we still use inverters from the older inverter platform. This inverter is now being released into the market. It brings customers the ability to upgrade later.
You can install a PV inverter without storage or backup capabilities, and you can add that on later. It creates what we call future-proof your solar system opportunity, as well as some opportunity for continued revenue from systems. It brings the ability to literally one inverter act as the energy manager to the home. Of course, it has our regular advantages that installers love and know, like ability for module-level optimization based on our optimizers, reduction of balance of system because of our unique DC architecture, module-level monitoring, and ability to do remote O&M, and of course, safety, which is becoming more and more important. This inverter actually works in harmony with a range of residential products that we are selling now.
Anything from backup interface products for home backup, of course, Power Optimizers, metering for managing your home's energy, a range of software products for the installer and for the system owner in order to design, operate, manage your PV system. We're already selling EV charging inverters, which is very fast-growing segment, is of course, EV cars. Smart standalone EV chargers, which work in unison inside our system in the monitoring, there is the ability to optimize your charge based on rates and energy consumption in the home. For the European market, where the three-phase grid is much more common in residential systems, we have a three-phase storage inverter, which we already started production of. The next product that we started selling already is a SolarEdge Smart Module. In Australia and Europe, we already offer today a SolarEdge Smart Module.
This is a module from SolarEdge, manufactured at a tier 1 bankable production site. Customers get the benefit of getting the whole solar solution from one vendor. It already comes pre-installed with the Power Optimizer, so installation and logistics are a little bit simpler, and it offers customers the ability from the roof to the grid to get one solution from one vendor. It, of course, increases the average revenue per installation for SolarEdge. The next product that is coming in 2020 is the SolarEdge Home Battery. SolarEdge residential batteries today, when people buy storage inverters from SolarEdge, they are coupled with third-party battery, mostly from LG Chem. We will continue to support that, but we will be also offering our own battery. This is a 10 kWh, 5 kW peak SolarEdge battery that works seamlessly within our system. Installation is very easy and straightforward.
Everything in the configuration and monitoring, of course, service, sales, logistics, everything comes from one platform in one software platform. We are very eager to release this product into the market. Again, it increases our overall share of the pie for sort of systems with storage. Another range of products that we have started selling and we are expanding that offering is energy management solution. These energy management solutions allow you to control or allow our customers to control energy devices around the home. Hot water heating, heat pump, pool pump, air conditioning, all can be controlled in one system. All is controlled by the inverter in order to optimize your energy bill, in order to optimize your energy usage.
If with a battery, you can actually take solar energy, which is not used during the daytime and shift that energy into hours of the day where you have load but no sunlight. With energy management devices, you can actually shift loads and shift these loads into hours of the day where you have solar energy. With that, we are optimizing the whole energy usage around the home and providing one seamless, energy management platform to our customers. Of course, ease of life, because all of these interact and integrate with home automation platforms like Amazon or Google, allowing you also to get ease of life at the tip of your finger.
[Presentation]
I told you this is the highlight of your day. What you actually see is a range of products from smart modules, batteries, smart inverters, optimizers of course, monitoring solutions, monitoring applications. All of this together allows us to offer, as Zvi said, a bigger piece of the pie, and also to allow differentiated value to our customers. Because differentiation and value is everything that we sell. The last piece of the puzzle is our grid services software platform. As solar is becoming more and more prolific, utilities are having an issue with solar. Intermittent generation, less knowledge what is actually going on their network because they cannot control this generation, they cannot predict this generation. We've brought a software product, which we call the SolarEdge Virtual Power Plant.
This is a cloud software product that sits on top of our monitoring platform and allows utilities to visualize, interact, and control with a fleet of solar systems. You can actually pull together and aggregate thousands, tens of thousands, hundreds of thousands of residential solar plus storage systems into an interface that utilities, network operators, energy retailers can actually control and visualize. It gives them visualization of what is actually going on their network. Some of these 10 customers that you see here, some of these 10 agreements are actually data where they tap into our platform just to know what's going on their network. It also allows them the ability to control the system. They can dispatch energy from a pool of system into the grid as if it's a virtual power plant. They don't have just to dispatch.
They can offer frequency stabilization services by pulling and pushing power into the grid. They can dispatch energy in a specific area where the network is congested, energy retailers even can sell energy from these systems into the grid when the wholesale price is high. This platform, we are a technology player. We provide this to the network, to the market participants in the different segments of network management. It is still a very small part of our revenue, but it is an enabler for more hardware sales also because the more these networks are being deployed, by the fact that our inverters have this capability, we see more and more inclination to use them because of this grid flexibility ability. This actually turns solar to be one with the grid and not against the grid. It solves the problem for the network operators.
It reduces the friction between traditional generation and distributed generation. It's a revenue source for us. It's a differentiator for us. Again, it increases our piece of the pie with the software value that we provide. As I said, we already have more than 10 agreements of building and operating such VPPs based on our VPP platform. Okay, moving on to C&I. This was all residential. Let's talk about C&I. The first thing about C&I is that our flagship inverter today in the C&I market is the Synergy inverter. What you see up there, you may be wrong to think that these are three separate inverters, but this is in actuality one inverter which is modular, made out of three units. The Synergy concept that we've brought into the market has grown very popular with our C&I customers.
It allows them to install inverters as if they are small string inverters. You hang these three units very easily. You connect them quickly with a prefabricated connector. Once it's installed, it operates and wires as if it's one larger C&I inverter. With this, you reduce labor, you reduce AC cabling, and you simplify your overall PV system. It's kind of like a large inverter, but in terms of installation, you get all the benefits of smaller string inverters. Today, we sell our 100 kW Synergy inverter in the market. We're in the midst of bringing into the market a 120 kW inverter. The bigger they are, we leverage economies of scale and are able to better compete in places where dollar to watt matters.
In 2020, we also plan to bring a 150 kW Synergy inverter into the market, further leveraging these economies of scale. In 2020, we plan to bring our C&I battery into the market. This is a 40 kW peak C&I battery. This battery allows for our C&I customers to do value stacking in commercial buildings. Value stacking means that you can use the battery locally to manage your electricity consumption, avoid demand charges, maximize self-consumption, and then you can also stack additional values onto the same battery by, in some times of the year, leveraging time of use. In other times, you can participate in grid services based on the same grid services platform that we have already for the residential market. Even in places where energy resiliency is important, and we see that it's becoming more and more important in more markets. Look at California.
This allows the PV system to participate as part of a microgrid, maybe with a generator or maybe with another energy source in order to bring more grid resiliency to your C&I campus. As we did in the residential space, we are also working on commercial EV chargers, both the DC fast charger and AC chargers. These will couple into the same monitoring platform that we have, into the same energy management concept that we have for the C&I space, and they are in development these days. Let's talk about utilities. We today already provide inverters and optimizers to the utility market. There's more than 200 MW of utility ground mount systems with SolarEdge inverters in them all over the world. Basically, same advantages that we have in other markets apply also for the utility market, only with different scales.
Added energy is, of course, very important for the ROI of a utility scale project. Reduction of cabling. There's tens of miles of cable in a utility project, and the ability to reduce cabling with our DC architecture with long strings is very important for such a project. Also, reducing labor and, of course, monitoring. The ability to visualize so many modules in a utility scale project and improve and streamline O&M is also very important. With Kokam, we also have more than 160 MWh of front of the meter utility scale storage already deployed. Let's talk about our products for the utility market. As I said, in terms of inverter, our Synergy platform is the platform that is also servicing the utility market today. The main difference is that for the utility market, we also have a 4-to-1 optimizer.
That is an optimizer that can control an MPPT four modules rather than the 2-to-1 that we use in C&I or the 1-to-1 that we use in residential. With that, we improve economies of scale, something that is very much needed in a utility scale project. We have a utility optimizer that is being launched in 2020. We have the Synergy inverters that are already being used and deployed in utility projects. In 2020, we are bringing to the market a larger scale string inverter. This is a 380 kW string inverter. It's quite big, and it aims to help us penetrate the larger utility scale system market resistance are bigger, and people still want to consolidate their AC wiring even more.
We find that the 380 kW, 250 kW in Europe, there are different grids, so it's 250 kW in Europe, 380 kW in the U.S., type of inverter will allow us to continue the penetration from the utility ground mount segment to the larger utility ground mount segment. This is coupled with bringing more and more utility-scale storage systems through our Kokam acquisition to service utility projects where you need that storage capability, either as a buffer to grid intermittency or just to provide grid services and other grid balancing services. We see that is becoming a bigger need in more and more utility connections around solar in various networks.
I hope that I was able to relay how our product strategy aligns very well with our overall strategy, increasing average revenue from installation, further enhancing our product leadership in the residential space, completing the portfolio for the commercial space, and enhancing or hopefully even accelerating our penetration into the utility segment we just started and is now growing. Thank you very much.
Uri.
Good morning, everybody. I'm happy to be here. My name is Uri. I've recently joined SolarEdge, three months ago. Before that, I was having 20 years of experience in the electronic manufacturing industry. At my last role at Flex, I was the senior vice president of global operations for the Americas and the European sites, which consists something like 40 different sites in manufacturing and logistics, and more than $10 billion of revenue. I would like to, in the next 20 minutes, to take you through some of our key successes to our operation and show you where we want to go in the future. What's the thing or capabilities and processes and systems that we will need to use in our future in order to leverage them to our new coming businesses. First of all, our manufacturing has always been a key to our success in operations.
Therefore, taking these capabilities and the system that we have developed in the last 10 years and leverage them on our new businesses is something that we would like to introduce here. Automation as part of building power electronics in the most reliable product with a mass production capability and the right cost will require us to develop our automation furthermore to the next step. Last but not least, Zvi has talked about it, and Lior has talked about it, our product roadmap consists a lot of batteries. If you can see, in every product that we are going to be used in the future, battery is going to be a key. Such introducing independent supply chain for battery would be something that is paramount to our future product and our future roadmap and our independency of the supply chain.
Through the years, actually, when we had started, and I was part of the team that was the first contract manufacturer for SolarEdge 10 years ago. Ever since then, we have chosen contract manufacturing to be the platform of manufacturing for our product. We have started with the Flex Israel operation back 2010 and then develop it to Flex Hungary operation and then to our main site in Fuzhou, China, which is currently the biggest site that we are building our products in. Such, in order to do that, and through the years, we have developed kind of a playbook. This playbook help us to achieve whatever you see today. Coupled with that, we develop it and make it today to be what we call the SolarEdge manufacturing system.
I'm specifically saying system because it's really important to understand that you cannot produce a mass production power electronics without a system in the level that we have done. Such, this system was a vehicle for us to be able to manufacture these mass production volumes. We have produced over the years more than 45 million different devices that are today installed in a rugged environment at mass scale. Not only producing this product was a key for us, also producing this product with the right cost and the right elements. This whole operational success and achievement would not be able to be achieved without our deep knowledge team and very dedicated team. I'm very proud to be today part of that team and part of this financial environment.
I would like to take you through a little bit of numbers and what we have talked about our operations and where we are going and what is it that we do. You can see here at the beginning of 2015, we took the numbers from the IPO days until today, and it is able to see here our growth year-over-year. We have growth more than 50% year-over-year, and 2019 is going to be the year that we're going to produce almost 16 million Power Optimizers and more than 700,000 inverters. This is a massive number, but has been able to achieve by using our systems and our manufacturing platform worldwide. Looking to our footprint and where we are.
What you can see here is a footprint of our distribution center, which is our last mile to the end customer, and also our manufacturing sites. Starting with the Flex Hungary site, going to the Jabil site in China. Recently we have discussed with the investors about our new Jabil Vietnam factory. This is the second footprint. Last but not least, is our Kokam Nonsan factory. This is part of our strategy of building cell to system battery cells. Ain't that amazing? At the beginning of my presentation, we discussed about the system and processes that will enable you to produce such kind of volumes and high complex power electronics. Definitely when we thought of how we're going to do it, we have created this system that we call the SolarEdge manufacturing system.
This system is a comprehensive approach to everything that we have acquired through the years. It contained our quality and reliability processes. I just want to one second explain you what is quality and reliability. Our products are almost the only electronic products that are worn for 25 years. If you don't apply a cutting-edge, unique, innovating quality and reliability processes, you will not be able to supply these products, and you will not be able to operate them in the right way. I will walk you through that bullet later on, but as such, we have really created this advanced quality and reliability system. As a holistic approach to design and in SolarEdge, when we design something new, we design it in our R&D as one team to manufacturing. It's not that our production is siloed and our system is siloed.
We take a comprehensive approach to the way that we design new products, and R&D is involved from the very first moments of the product till the end of the product. Then we discuss about automation, data platform, and automated testing. This is our comprehensive system to achieve the targets that we want and the volumes that we want to get from our operations. Now, I'd like to deep dive into two of the elements of our system today. One of the most important element of the system is how do we really apply this SolarEdge manufacturing system? How do we really make it happen? How do we know what is being produced in our global manufacturing footprint? Therefore, there's a lot of discussion today in the industry of Industry 4.0 and what is connected device and what do we need to do.
I think at SolarEdge, it's an evidence. What you can see in this diagram is something that is very illustrating our ability to predict what is our product at production and to understand what we have produced, in which quality we have produced, and what would be the predicted quality of this product going forward or 25 in the years. We can see in the base of the diagram, this is our machine. This is our production line. Every production line is produced from a front end and then back end manufacturing. If we take our automation, that's the back end of our operations, and the assembly is the front end of the operation. Every machine in that production line is really connected to a repository and communicating with this repository all day long, and in a real-time situation. This repository will apply some engine.
One of the engine is machine learning. The other one is machine-to-machine. At the end of the day, you will see on the top of that, our application to control this. This application would be something that you can use wherever you want. It could be used by our headquarters, it could be used by our manufacturing site, it could be used by the operator, it could be used by everyone. In essence, I'll give you some idea, because this chart is something that I've seen so many times, but I've never seen it in action. I've seen so many companies that are claiming to have that kind of capability and data platform, but I've seen little that are really in that stage.
I saw SolarEdge, and I remember it myself when the SolarEdge team came to my factory and said, "This is what we're going to do." We kind of look at them and say, "Okay, yeah, you're going to do it." Nobody did it. However, today, if you can look at it, I'll take, for example, the automated testers. Automated testers are, you can call them the pulse of the manufacturing. Those are the devices that every product is coming through them and is being tested through them. For example, if a product is not tested or one result of the Optimizer is not in the right threshold, you will not be able to produce it later. You will not be able to ship it, you will not be able to do anything, you will not be able to scan it, put it on a pallet or anything else.
That's the simple thing. However, these machines have also some predictable engine. They would be able to tell you whether the issue that you have is a contamination that is part of the history and the algorithms that we put inside, or is it just a wrong component or wrong number. What is machine-to-machine? Now we know how to elevate this data. We know how to take it to the repository, and we know what to do with it. The second thing is our machine-to-machine. Think about it, that at the end of the day, if I have an issue with one of the machines, with the vast history of what I have done, I'll be able to talk to the previous machine and change the parameters so the process will be improved.
Therefore, you get an end-to-end, fully process control system to the most, I would say, complex power electronics that one can introduce to the field. That gives us a very well prediction of what is our process, what are our product, and what's the quality of the product through the life cycle of the product. I would like you to think about it as an autonomous operation. If you take one cell of automation from one end to the other end, it's kind of autonomous. It doesn't need training, it doesn't need a ramp-up, it doesn't need people. We're eliminating completely the human factor here because we don't need to train a quality manager to look at our process. Our process look at itself and have the vast data of these multiple lines over multiple countries as one neural network that takes all this data together.
I'm very excited about this capability. In the near future, this is a key for us to grow, to grow our current business and our new businesses. I've mentioned that our products, and you all know it, is power electronics that sit on the top of the roof of every place in the world, from Alaska to Brazil. The way we approach this one, the reliability of the product, is with a very advanced system that is a holistic approach to product design, to manufacturing, testing, our install base, and then the repeatable information from the field. As we said, we have a data platform. We have, as Lior explained you before, we have devices that are actually one of the biggest IoT implementation worldwide. Our inverters are an IoT device that sit everywhere and generate power and generate communication.
Taking all of this one in a real-time information, we can understand what is our product reliability, what is our quality, where are the malfunctioning in this process, and how we can, again, pull it into our R&D and improve the process. We like to see it as an ongoing circle process that improves our product and improve our process through the years. As we step forward and introducing new businesses, increasing our customer demands, you all have heard that our regions are booking business more and more day-over-day, and we have seen in the charts that every year we are growing our revenue and our product installation by 50%, we will need to onboard few elements to our system. I will take you through some of the four elements that we are onboarding into that system. The first one is the geographical expansion.
That's not only because our products need to be in different geographies, that's an answer that we need, too. I think the world has changed. If you look at the old manufacturing side, and that's coming from where I were, everything was siloed, everything was in one place. Today, with all the macroeconomic changes and the ability to reduce costs by not spending so much on shipping, the world is going slowly and slowly towards localization. Therefore, localization means that you need to change your geography and go to different geography. Again, as we said, with automation, we can do it rather easily. If you look at the automotive industry, for the automotive industry, as we take it, the automotive industry is the most high-reliability industry. It takes them years to transfer a product from one place to the other.
In our case, if you have this data platform and you have this automation line, it takes you only two weeks or four weeks at a max to transfer a line and educate three to six people. Those employees would be only material handlers and very simple capability to ramp them up. The next one would be our Copy Smart and our new business integration. Let's discuss a little bit about geographical diversification and where we are going and what is our plan in numbers. I've shown you the geographical span of SolarEdge, and I will talk a little bit about what are we doing in each of these factories. The first one is Flex Zala, Hungary. We're increasing our capacity over there by 200%.
I'm happy to say that we are halfway there, and we have finished our first installation, and by Q2 2020, we will be fully operated in that facility, and the capacity will be as we planned for 2020. The Jabil site in China, so this is our mass production site. This is our biggest site. That will be grown in 20% for Q1, and that's going to be implemented in Q1 2020. Last but not least is our Vietnam factory. We have talked about Vietnam factory. The first stage of putting the capacity there and ramping the site will be finished by the end of this quarter, and by the end of Q2, we will be fully ramped in Vietnam. If you ask why we are doing it, is for a couple of reasons. We still have time, no? That's yours.
Yes.
Okay. If you ask why we are doing, there are a couple of reasons. First is our customers. Our customers are increasing their demands, and we need to put a plan that we'll be able to deliver them in the right time, in the right geography, with the right product. Secondly, as you know, our air shipment costs have been extensively high in the last few quarters and will continue to do that until we will create some buffers and inventories to be able to reduce this dramatically, this shipping cost. The last thing is tariff. We want to be, by the end of 2020, tariff-free product for the North American market. Just to give you a glimpse into our Vietnam operation, it's not just something that is in the chart. You can see here our site.
You can see the front end, the back end, the products that are on the line being manufactured as we speak. We said, and I will emphasize automation couple of time, and I want you to understand that the digitization and the automation is something that is core to SolarEdge, and it's also very advanced as opposed to other manufacturer. Therefore, today we are in about a penetration of 20%-30% of automation. As for 2020, we are going to be, by the end of this year, 100% dependent on automation. That's going to be in the three sites that we have discovered before.
The reason why we want to do it is because we want to ramp fast, we want to be agnostic to human errors, we want to be in the high reliability and predictability of manufacturing product that we can, and we want to also have a faster ROI. For this, I guess ROI is very important for this community, and I have done automation through my years many times, and I think ROI of something between 12 to 16 months of an automation product is something that is out of the chart. I have not seen something like that before. That's because we have taken the holistic approach that R&D is responsible for designing the product that will fit this automation.
If you would look 10 years ago at the office of Lior, you would see all these parts that they were thinking how to get this product that will be eventually on automation. You should understand, this is a cutting edge. It's not something that you can develop in one year. If you come tomorrow and say, "All right, I want to put in place automation," it will take you many years to have the right product to the right automation. Who said that manufacturing is boring?
If you look to our challenges and what we are trying to achieve, basically what we're trying to achieve is take a specific product, in this case, a power electronics product, produce it in the high reliability, and quality, make sure that it meets the market demands in terms of quantity, reliability, and cost, which is very important, then be able to put it in each geography that we would need in the future. That can only be achieved with a holistic approach to reliability, to engineering, and R&D. Taking these all elements, we have come to a conclusion that having a proximity site next to our R&D reliability will take us to a speed to ramp this product in the most quality way that we can. Therefore, we have decided to build our Sella 1 factory. It's a factory that will be in Nazareth.
It's currently in the process of ramping up, fitting up the factory. It will be planned to be operated in Q2 2020. The idea of the factory is to innovate fast as a proximity to our R&D, to be able to take our product, perfect them in the factory, in the processes that we know with our SolarEdge manufacturing system, and then move them as a mature product to each of our regions. We talked about SolarEdge systems, our SolarEdge infrastructure, our platforms. This platform has helped us so far to grow our business. As you all know, we have acquired some other businesses. We acquired the e-Mobility business, which is also based on power electronics. We have acquired the UPS business, which is also based on power electronics, and we are planning to leverage our system into that new spaces.
For example, if you take today what we are producing, the PCBAs or the boards that we are producing, we'll see it downstairs in the showroom. If we take these boards, they are today produced in our infrastructure. Right? That gives us economy of scale and ability to control our processes as we onboard our new businesses. Same for our UPS businesses. The next business that we onboard is we talked about our cell-to-system strategy. That's a kind of a vertical, which we are going to rely on the Kokam capability together with our platforms and develop the capacity that we will need to our new products and to our product roadmap.
Today, our Kokam capability is 200 MW at theoretical capacity, and we are going to expand it in the next two years to a 2 GW factory capacity, which will enable us independence in these critical elements of the battery solutions. 2020 numbers. We're taking all the elements, the systems, the processes, the holistic approach of SolarEdge, our geographical expansion, and I want to say that we are ready for 2020. You can see that in 2020, we are planning to double our manufacturing, and to double our capacity at the site, and to double up the number of products that we are going to produce. There are four reasons why we want to do it. Number one, as I said, our customer demand. They want our products, and this is the capacity we are put. We're going to be more geographical to that capacity.
The second one, as you know, probably Ronen will talk about it, we have spent a lot of money through the last quarter and the quarter before on air shipment. This capacity increase will allow us to decrease dramatically our air shipment cost. Our plan is to be as a minimum of this air shipment cost at the end of 2020, while our production will be up and running in all regions. The third thing is the tariff. Both, I think this is something that worldwide is being an issue, and I'm happy to say that by the end of 2020, we'll be completely tariff-free for all our North American customers. The fourth one, which is very important, that we have developed that platform, think about the economy of scale.
If you look at this plan and the number of products that we are planning to build, this is over 40 million products. These products and this capacity will help us to further continue our cost reduction, our ongoing cost reduction, and will enable us to compete with the new businesses that we have acquired, as these businesses will be in this platform and enjoy the economy of the scale of what we have today. To conclude, I hope that what you have seen here today is that we are ready to deliver. We are ready to take our processes. We have a system to do that. We know how to do it. We have been experiencing that for the last 10 years. We have a very high-reliability product. We are using automation as a key element to be able to deploy our production worldwide.
We are well-positioned for our growth, and we are well-positioned to onboard our new businesses. By that, I would like to thank you all, and I think it's Zvi now.
We're going to take a quick break, 15 minutes, and we'll reconvene here at 10:45. There are refreshments in the hallway.
In the previous section, we discussed the solar market, our position, and our plans how to continue to grow fast in the solar market. It's now time to look at the new markets that we have entered recently with the acquisitions that we made. At the premises of our strategy for growing into new markets are two elements. One is that we believe that the way people consume and produce energy in the future will be very different than the way it is today. The second thing is that in power electronics, and in particular, inverters and batteries are a huge part of that transition. That puts us in the position where we have the opportunity to place ourselves in the middle of a transition that is growing large new markets and new applications. There are a few trends in the power industry that are driving this change.
Grids are facing more extreme conditions, and they need solutions in order to be able to deal with these conditions. Storage is becoming a bigger factor in the market, and of course, e-Mobility is a very significant trend that we'll dive into a little bit in more detail later on. As we all mentioned, devices are now more connected, and devices can produce energy, and devices can consume energy, and that creates more strain on the grid, more complexity, and at the same time, it creates more opportunity and flexibility. Of course, in general, power demand is increasing, and the need to supply more power and at high quality is at the core of the challenges of the energy industry today.
[Presentation]
The philosophy behind our entrance into new markets were markets where power electronics, and in particular inverters, are a key enabler, where lithium-ion batteries are a key enabler, and markets that are expected to grow at high rate in the coming years. Based on that philosophy, we looked into a few markets, and I'll start with the e-mobility one. Electrification of transportation is a hot topic. Everybody is discussing it. There's global commitment to shift, across the world, and industry experts expect that by 2030, roughly 50% of the new vehicles that will be produced that year will be electric. That represents approximately 50 million electric vehicles a year. This is causing opportunity in the automotive industry. The shift from combustion engines to electric engines is creating opportunity for new entrants and changing the landscape of the automotive industry, and it creates opportunities for us as well.
All forms of vehicles will be going through the transformation. Trucks, commercial vehicles, passenger cars, two-wheelers, three-wheelers. There's one common or two common threads to all of those. They will need an electric powertrain, and they will need batteries. Looking here at one of these segments, which we chose to be our entrance into this market, is the segment of light commercial vehicles and light goods vehicles. Basically, LCVs and LGVs, light goods vehicles, are the vans that go around the cities delivering stuff, and there's even a higher motivation to get them electric and to reduce pollution within the city. If you look at this chart, the expectation is that in 2030, approximately 2 million such new vehicles will be produced. Approximately 2 million light commercial vehicles will be produced in 2030, and roughly about half of that in 2025.
When we looked at new markets to enter, we targeted this market and tried to find a way where we take what we have and what we've developed over the year, together with some additional capabilities we bring from the outside, and have a good entrance into this market. This is an illustration of the same thing. Basically, a solar system has inverters, batteries, communication, and control software, and the electrical part of an electrical vehicle has inverters, batteries, communication, and control software. Through the acquisition that we did in Italy of a company called S.M.R.E., on the basis of that, we formed the SolarEdge e-Mobility division. What we have today or can bring to the table today on the basis of this division over here is, first of all, a complete offering for the LGV and LCV market.
We have a powertrain with batteries, with an engine, with everything that is needed to electrify a light commercial vehicle. With the company that we acquired, we are already in pre-production phase with leading automotive manufacturers. That, together with our in-house capability of R&D, of scale of manufacturing, we believe that that puts us in a good initial position to be a player in this e-mobility market. Obviously, this is a process of years, and it's a long-term strategy, but it's a strategy that works to our strength and gives us entrance into a very large future market. If we try to quantify the opportunity of this market and this business, a light commercial vehicle uses an engine of 60 kW continuous and roughly 40 - 80 kWh of batteries.
If you take prices of power electronics today and make all the aggressive assumptions of ASP and cost reductions of power electronics and batteries, you'll reach a conclusion that on a 2 million vehicle basis in 2030, the size of the market for the solution that we are offering already today is a market of roughly $20 billion-$30 billion into 2030, and it is going to grow gradually from where it is today to that point. That's obviously a very big opportunity, and that is just in the large commercial vehicle space. This gives us the basis to develop products and offerings for the other segments of the e-Mobility industry on the basis of what we already have in hand and on the basis of the improvements that we'll do with our capability in power electronics.
This is the first, call it obvious entrance into a new market for a power electronics company, and that was the thinking behind the acquisition that we did of S.M.R.E. in Italy. A second market of interest is the UPS market. Here too, a UPS, at the end of the day, it's an inverter with a battery, very similar to a solar system with communication devices and control software to manage the system. The market here has similarities to the inverter market. It's roughly $7 billion-$8 billion in size, and it's segmented from single phase, typically up to 10 kW or 20 kW inverters or UPSs, and three-phase UPSs ranging from roughly 10 kW into 500 kW and megawatt scale UPSs. Very similar to the solar industry. One of the differences is that in recent years, there has not been a lot of innovation in this market.
Growth rates have been reasonable but not very fast, and there hasn't been a lot of investment in development in R&D. That is beginning to change with the evolution of more cloud applications, mega data centers, co-located data centers. The need for uninterrupted supply of power, in a cost and reliable way, is driving innovation in the UPS market as well. Here, too, we did our entrance to the market through an acquisition of a company in Israel that has been active in the UPS business for almost 40 years, a company called Gamatronic. They had a not very large global footprint, but they had two assets that were of value to us.
The first is decades of experience in this market. UPS is not only about the hardware, it's also about the application and knowing what type of scenarios might evolve and what type of scenarios you need to be capable of dealing with. The second thing that they had was a competitive offering in the three-phase segment between 10 kW-500 kW. Obviously, the synergy to our operation is very obvious. The products are practically the same. The manufacturing process is the same. The network of sales and service that we have globally, while different customers, is set in the right places and with the right technical capability to serve this market, similar to the way that we serve the solar market. Our plans here are pretty much to repeat the solar business, hopefully maybe faster.
We will be adding products to the portfolio in order to cover all segments of the market. In particular, a single-phase UPS that represents about a third of the market that we plan to introduce in 2021. We will use the methodologies that I discussed before in terms of sales, service, marketing, customer attention to grow market share one step at a time globally in all of the markets that we are present. I'll spend a few minutes on the other aspects of the Kokam acquisition after Uri gave you the operational view. Practically everything that we spoke about today, everything that we discussed, ends up with a lithium-ion battery. The demand for lithium-ion batteries has been increasing in recent years, and it's expected to increase because it's serving three of the high-growth markets. It's serving e-Mobility, it's serving consumer electronics, and it's serving energy storage.
The last two years, because of the growth in demand, capacity lagged demand, and there are always challenges of availability of lithium-ion from the base chemical to availability of complete cells and batteries. The market for lithium-ion batteries alone is expected to be about $70 billion in 2022. It's a huge market with the core of lithium-ion and connected to power electronics capability with battery management systems and so on. By acquiring Kokam, we acquired a global Tier 1 provider of battery solutions. They've been in the market for decades, and they've been selling batteries for all of the known applications, whether it's ESS, UPS, EV, and others. They already have 600 MWh of batteries installed worldwide, and they have know-how in the chemistry and the ability to customize chemistry of batteries for the applications that they are intended to.
This is, of course, very attractive to us as we are serving multiple applications that consume and use lithium-ion batteries. When we put a storage system with a solar system, we are not putting only a battery, we are putting a full system. Similar when you put a battery into an EV, or you put a battery with a UPS, it's not only a battery, it's going into a system. By acquiring Kokam, we are giving ourselves the ability to control the process from the cell, from the chemistry, to the system. That means to design the cells such that they're optimal for the end system that they need to perform within. This gives opportunities in terms of cost, it gives opportunities in terms of performance, it gives opportunities in terms of quality and reliability.
We will need, we will want to develop many battery products in the coming years for all of the applications that we serve. Having this independence to develop chemistries for systems that we are providing and to have uninterrupted supply is a huge advantage for us. It doesn't mean even that we will only be consuming our own batteries. We hope that the demand will be much bigger, but by having the complete product defined, we will be able to outsource and supplement our own capacity in a more easy way from other cell and battery providers. Of course, this is common to all of the businesses that we discussed during the day, solar, EV, and UPS. Quantifying the opportunity here is a bit more straightforward, as was mentioned. Our current capacity is limited at roughly 200 MWh a year.
We intend to build a factory that will be complete in 2022, with the goal to have capacity of 2 GWh . Capacity of 2 GWh , roughly, again, taking into account ASP erosion, current cell prices, current battery prices, because we intend to produce batteries, not only cells. You can call this an annual opportunity of roughly $500 million a year, not necessarily happening in the first year of production, but when we peak at the 2 GWh , that roughly represents the business opportunity behind this acquisition and behind entering into this market. If we look back at what we covered during the day, we discussed solar, we discussed the trends in the solar industry.
We discussed what we are doing, how we have outperformed the solar industry in the past, and what we are doing in order to continue and outperform the growth rate of the solar industry in the future. We discussed how we are taking the capabilities that we've developed in the areas of R&D, manufacturing, sales, and marketing, and finding markets where the technologies are similar, and matching our capabilities with strengths that we bought from the outside to build new business divisions that will be operating in these markets in the new era of grid that is going from being a centralized polluting grid source to an interactive distributed grid based on renewable energy.
Of course, this is a vision, and one of the sayings that Guy used to share with us frequently, he used to quote Shimon Peres, who was the Prime Minister of Israel, and Peres used to say that dreaming big and dreaming small is the same effort. May as well dream big, because then you have a chance of actually doing something big. When we sit around in the management team, we want to dream big. The actions that we're taking, the capability that we're developing, allows us to dream big of maybe being in 2030, a company of $8 billion or $10 billion. Now it's a dream, but dreaming is fun, and that gives us the opportunity to have some fun in our work. Hopefully, you can dream a bit with us and go through this journey together with us.
I think it's time for you guys to talk a bit and ask questions if you have.
Thank you, Zvi.
First of all, again, thank you very much for joining us today. Part of the benefit of having this kind of a day is not to allow you just to hear us, but actually ask questions that you have on your mind, and in a less narrow way than as we do in conferences where we usually look at financials, and we don't have the time to talk about the broader areas. One of the issues that we expect or expected to hear a little bit of question is related to our IP lawsuits, in China and in Germany with Huawei. Therefore, I would like Rachel, our legal counsel, to address this and allow you to ask some question about it, and then we can clear it out of the way and concentrate on the really interesting thing, which is more of technology and, later on, financials.
Hi. Hello. Good morning, everybody. I'm Rachel Prishkolnik, and I'm General Counsel to the company. Addressing the IP litigation matters and our IP in general, as you'll hear a little later on from Yoav Galin, our VP of R&D, technology and innovative development is at the core of what SolarEdge does. To that extent, we believe that our investments, which are very significant in R&D, should be protected. We have a portfolio of over 300 patents, and continuously are drafting new patent applications. To that end, when we suspect that a competitor is infringing on our patents, we will litigate and do what we can to protect our technology and to protect those investments. We've done that, as you know, and as we've published, in Germany and later, more recently, in China.
This last week, we had hearings, so I think it's already been covered a bit in the press, and I know Phil Shen and Brian Lee both put out short reviews about it in the last two days. We filed complaints against Huawei on three of our patents. There were two hearings in Germany last week, on the same day, the same panel of judges. In one of the instances, the judge was less inclined to accept our claims of infringement, and his decision was not in favor of infringement. We can appeal that judgment. In the second, he has appointed a court expert to examine the Huawei products, which we're very encouraged by. In the same week, it was a busy week for us in Europe, we had a hearing for one of our patents, and indeed, that patent was revoked.
What you have to remember in this process, and I'm sure some of you are familiar with it from the press and from other IP litigation matters, it's litigation, like Phil said in his report. It's not a sprint. It's very hard to get easy wins. Our technology is very complicated and difficult to explain in front of panels of judges. Ultimately, we believe that we will prevail. We can appeal. You know, most litigation proceedings in Germany, for instance, go through at least two, sometimes even three iterations before they're concluded. No matter who wins in the first iteration, there's an appeal. No matter who wins in the second iteration, there's usually an appeal. That's what happens. Patents that get revoked get appealed. This is a long process. It's three of our patents. We have many.
Three of our patents that we've challenged on. There's many. There's divisionals. As far as we're concerned, it's the very beginning and the tip of the iceberg on a process that we see it as necessary to protect our innovations and our technology. I hope that addresses any questions.
If there is any question, now could be a good time to ask Rachel if there's any question about this, and then we can move on.
Sure. Phil, yeah. Okay. Thanks. Just to be clear, the patent was revoked at the EPO level, not at a country level. That's the first thing. It doesn't have any impact on our business, on our day-to-day business, because people can freely operate, as long as they're not infringing other people's patents. This has absolutely no impact on our business day to day. Potentially, it could impact someone else's if we convince a court that they're infringing on us, but it doesn't have any impact on our business going forward.
Please, one thing, when you get questions, please wait for the mic so the listeners in our webcast can listen as well.
Yeah, thanks. I guess just a couple of questions with respect to the timeline. You mentioned, I guess, two issues outstanding. The patent that's been revoked, what's the sort of timeline expected for your appeal process? I know you said it's not a sprint, but any kind of sense you can give us on that? Secondly, on the court-ordered expert opinion, if I heard you correctly, is there an expected timeline around when that process has a resolution?
Sure. On the first instance, appealing a revoked patent, the whole process takes anywhere from about one and a half to two years, from what I understand from our Europe. I can tell you that a lot of the unexpected happens in these processes. Judges get replaced, panels change, so there's no guarantees. Again, as I answered to Phil's question prior, it doesn't affect our business, and it's one of a family of patents in an area of our innovative technology. Your second question was the timing on the expert. The judge actually asked for a swift appointment of an expert, and I think we agreed with the other party by the end of December to agree on the identification of who that expert would be. I really don't know what the timing is going forward after that, because again, it's just subject to the courts of Germany.
Other questions? Sure.
Rachel, what's your budget in the legal division for patent matters, and is that going up or down in the next two years?
Our budget obviously is not just litigation. Litigation is not something that we budget for specifically and say we're only going to spend X amount of dollars on our litigation. It's not budgeted. I have the privilege of a very free budget in that area. On our patent prosecution, we have in-house, a team of seven patent engineers, and a patent director who work directly under our chief technology officer with a dotted line to me, and they are working on continuous drafting of patents, which we do the initial drafting in-house, because we think we get better product that way. Our patent prosecution budget goes up annually because we're developing more. Now we're starting to assist the Kokam team in drafting their patents. We're working on e-Mobility patents, it's not going down.
Yep.
Sure. Hold on. The mic. We'll give the mic to Mike.
Yeah. As you get more patents on more integrated systems, we heard a nice presentation about all the integrated systems that are coming up in 2020 especially. Can you supersede the patents that are at issue here? Is eventually the company going to eventually have patents that will make the existing ones obsolescent at some point, and we don't even have to worry about it? Is that a fair assessment?
First of all, I don't think there's anything to worry about. Second of all, a company that has one great invention and then stops developing patents is in big trouble, obviously. It is definitely a moving target and process. It's just part of our development. Our R&D team, we have a patent innovation program. Our R&D team is incentivized to bring new ideas to our patent team. With every new development, of course, we issue patents and think that we are further enriching our technology portfolio, per se.
Last question to Rachel? Yeah.
Thanks. The patent that was revoked in Europe. While I understand that certainly it doesn't have an effect on your current business, it does have implications in terms of why that was revoked. Was that revoked for prior art? What would be the U.S. version of that patent?
That's an excellent question. There is a U.S. version of that patent. There's also a Chinese version of that patent. There's also divisionals related to that patent. We're very confident in the novelty and innovative step of the patent. It was revoked based on a piece of prior art. We disagree with the panel's conclusions about the lack of differentiation and inventive step, and we're very confident in our abilities to prove that. The process of proceedings is a long one, and sometimes things just do get confused in the process.
Just as a follow-up to my question, were there certain claims and one of the claims was what triggered, or did they say that the entire patent fell under the prior art ruling for the revocation?
Well, yeah. The meaning of a patent being revoked is that the main claim is, for whatever the reason, deemed not novel or not inventive, et cetera. That was the judgment of the panel after a 11-hour hearing.
Thank you.
My pleasure.
Okay. Now, we would like to open the floor for questions about the presentations that you saw or any other items that you would like to hear about. I'll just remind that there will be another session of questions after the financial one. You can basically split your questions between what you saw already and what you'll see in the next few presentations. Colin.
Thanks so much, Ronen. It's Colin Rusch from Oppenheimer. Can you talk a little bit about the maturity of the business for the light-duty inverters and how soon we might see you go into pre-production and production with some of those products?
Yeah. PD?
So-
On your mic.
I'm active. Yeah. As I mentioned, and I think as you've seen in our recent financials, and Ronen will discuss, we're not expecting or planning on significant revenue from this business in the near future. There is a maturity of the product, there's maturity of the capability and the engagement, but it won't be a large source of revenue in the immediate future.
Can you give a sense of timeframe?
It falls more into the timeframe of automotive qualification processes. When you are dealing with a manufacturer, let's say, of a motorcycle, the process is typically where they will test things out on a few samples, and then they will buy and get a few tens of samples and run them through a much more rigorous testing period that can range anywhere between, usually very long, maybe in the era of e-mobility, one to two years. On a very high level, that is in the phase that we are, either in the first few units delivered or in the first few tens of units delivered. The automotive qualification process is of a length of 12 to 24 months at least, usually.
Great.
Thanks. Hi, guys. Brian Lee, Goldman Sachs. Had a question, I guess two questions on the 2020 capacity plan. Maybe first off, is there a CapEx number you're willing to provide that's supporting the capacity plan? Secondly, if you look, it's a pretty robust capacity expansion here over the next 12 plus months that you're targeting, over doubling, it seems like. The revenue growth over the past several years, I know you guys have had some fast revenue growth years here, 50% plus or minus a couple of years running. Can you help kind of reconcile, are you building capacity multiple years in advance? Sort of seems like it's outpacing the revenue growth potential here, at least in the near term.
First of all, I'm not going to have a spoiler for my presentation, because what will I do after lunch? In general, I'll try to answer at least in concept. I think that the first thing that needs to be said is that our manufacturing is done today mostly through contract manufacturers. That means that any capacity increase is usually falling on the shoulders of these contract manufacturers. When we look at manufacturing capacity and manufacturing equipment, we divide it to two different buckets. The first one is the bucket of what we call the non-specific equipment. SMT machines, ovens, through-hole soldering machines. All of these are bought by the vendors themselves, and we basically do not invest in any of those. The parts that we're adding into this process is, first of all, the automatic assembly line.
As Uri mentioned before, we would like to be by the end of 2020, where all of our optimizers are manufactured in an automated way. Therefore, we're acquiring the machinery. As also as Uri mentioned, by the way, we have a very quick ROI on those ones. The second item that we buy are the testing machines, the testing equipment. Usually in the world of contract manufacturing, you get warranty of about a year for the product that you give. We're providing 25 years warranty for our optimizers. We give 12 years warranty for our inverters. If we're not testing this equipment in the factory before it is being shipped, the consequences could be very adverse, and therefore, we invest quite a lot in this kind of equipment. In recent years, it's been $33 million-$37 million of CapEx.
I will give you one note, and it will be around $40 million next year, when it comes to CapEx related to this. The second element of the question was how do we look at expansion of capacity from the business point of view? The first thing that we see here is the fact that in the last year, we saw growth that is way beyond what a company would usually plan. In 2018, we grew about 56% in the solar business. In this year, if you take the mid-range guidance of Q4 and street expectations, we'll grow at about 46% in solar. Usually, companies at a base of $1 billion are not planning for this kind of growth, and still it happens. We were lucky enough to be able to manufacture everything. We were lucky enough to have Jabil support, Flextronics support here, and our team.
It was at a very high cost of air shipments that was in Q3, approximately 254 basis points above another 200 basis points that we had in Q2. Therefore, this is a very large chunk of our gross margin that went away. The only way to take away these air shipments is by build enough inventory or enough capacity that will allow you, first of all, to deliver to our customers in Q4, Q1, and Q2 next year. To build enough inventory in excess to this amount that we can ship, then we can start rolling these ocean freights. Then, of course, to be able and to grow.
We believe that the capacity that we're building right now will allow us to be in a situation that by the end of next year, actually by the second half of next year, I wouldn't say eliminating, but reducing air shipments to the minimum. Of course, as I will say in my presentation, I will keep this for after lunch, is that we do expect to continue and grow, and therefore, we need more capacity. The way that we look at it, we should have enough buffer now to continue accelerated growth and to reduce air shipments, and this is how we come. This is, of course, not related to whether we need to sell all of those.
We hope that we will be able to sell as much, and we hope that Uri will have to deal with a lot of problems of how to increase capacity very quickly in the future. At least right now, we plan a little bit ahead in order to completely eliminate air shipment. Yes. By the way, question can go also to the other management members on the items that they said.
Great. Phil Shen with Roth Capital Partners. I was wondering if you could talk through what that geographic mix looks like, what the segment mix looks like, then if you can touch on margins, that would be great. Specifically, what I heard today was that you could still have tariff impacts until the end of the year, Q4 2020, whereas I think prior expectations may have been for no more tariffs after Q2 of 2020. Thanks.
We'll do the following. We'll defer your question to after my presentation because a lot of the answers will be there, and I do not want to say it before you can see it. I will just refer to the tariff. The tariff-free manufacturing is what we do by increasing, or I would call it non-subject to tariff product, is what we do in Vietnam and what we do in Hungary. We believe that with the growth that we will see, we will need more capacity than what we initially planned, and this is why we're increasing Vietnam. It's important to say that what Uri mentioned in his presentation is 2 stages of ramping up.
The first one is the planned capacity that we discussed all over the last few calls. This is to make sure that by the end of Q1, beginning of Q2, majority of our U.S. products will come out of tariff-free areas. With the second stage of Vietnam, we will be able to take it completely off. This is expecting, again, growth in our revenues in the U.S. Again, I'll answer all of the other elements of your question, Phil, in my presentation after lunch. Yeah. Shen?
Hi, thanks. Either for you, Ronen, or one of the other management team. Two questions, if I could. The first, the light vehicle, or the vehicle penetration. One of the trends that we've been seeing is the move to higher voltage architectures. V equals IR, so if you're going to higher voltage, you're going to have less amperage, which tends to have an impact from a battery perspective as well as the amount of cables. I'm curious, in your solution, how do you play into that environment or do you have a stronger position, I guess, with a lower voltage or amperage type solution? I have a second question.
The drivetrain that is currently in pre-production phases for the LGV and LCV range is a high voltage drivetrain. It's 400 V and a little bit above in voltage in order to support exactly that higher power, lower current application, especially in LCV and LGV, which require high current applications. It is a high voltage drivetrain.
You mentioned 400 V. We're seeing bus companies that are over 1,000 V, for example. Is there a limitation to your technology where, for example, you're highly competitive in that 400 V- 600 V, but kind of over that less so or?
No. There is no significant difference between, let's say, everything up to the 1,000 V. Once you go to high voltage in terms of the battery architecture, the inverter segment of the architecture, kind of like what we do in solar, the inverter segment of the architecture can work and operate at different voltages. The trend of high voltage, let's call it 1,000 V and above, is indeed in very high, heavy duty trucks and some buses, which is a market that we currently don't service. There are also a lot of regulatory issues regarding that higher voltage. What is currently, let's call it acceptable and most common in the LGV, LCV type of market is the 400 V- 600 V, but there is nothing in the technology that prevents from increasing the voltage to that 1,000 V range.
Great. This is my second question, just around the Kokam business. I'm less familiar with that business. In the battery markets, a lot of the differentiation is what nickel content is going to go into your cathode, for example. As I see you scale from 162 to the 2 GWh type nameplate capacity, I'm assuming that you are buying and mixing the chemistries yourself to create that, or are you buying cell production and is it more of an assembler? Could you just help me understand?
Kokam is a cell manufacturer with internal chemistry development capabilities. As you saw in Zvi's presentation, with that acquisition, we are able to integrate from raw material and basic chemistry development, and I think that in the technology presentation that is coming after lunch, I don't want to steal anyone's thunder, you're going to see those capabilities. Yes, we are able to develop from the chemistry level, from the cell definition level, and adapting cells all the way up to the system level.
I should assume or look at that as a pure scale from the 200 to 2 GW . That's the benefit there.
It's first of all pure scale. There are a lot of improvements in the processes and a lot of optimization that go into that 2 GWh .
Plans, predominantly scale with a lot of, I would call it, process optimization and process improvement that can be leveraged from high volume.
Thank you.
Mark Strouse from JP Morgan. Looking around, I see quite a few faces of relatively new investors to this space. Looking at all of the different companies that are coming out with storage solutions, when you go to sell the new residential system, can you just talk about what you lead with? How are you different than all these other systems that are popping up? Secondly, who are you selling it to? Is it the distributors? Is it smaller or medium-size installers, or is there potential to potentially win some business with some of the larger installers that already have systems? I'm thinking like a Sunrun or Brightbox, those kind of things.
Trying to remember both questions. Actually, it starts from the second question. The storage market is evolving in multiple geographies with different drivers. In Germany today, roughly one out of every two residential installation goes in with a battery, and that's purely on the basis of self-consumption. The tariff is very low. The homeowner wants to consume as much as the-- and the power prices are very high. The homeowner wants to consume as much as the energy as possible. U.S., California, are experiencing disaster and power outages, and that is driving the storage adoption in this area. One of the resulting trends is that storage is moving from specialty installers to a broader base of installers that are installing storage, excuse me, and SolarEdge. Initially in smaller volumes, and gradually it becomes a bigger part of their business.
That's a segment that is actually, for this type of segment, the advantage of having a full package from one source is very significant. It also ties into the other elements of our solution, because the small and midsize installers, if they can get the full system from one place, get the service, the warranty, and everything from that place, it makes their life much easier. We are positioned with storage solutions with the big guys and early adopters of storage, and we're seeing a trend of increasing storage sales by mid and small-size installers as well. The first part of the question, if I happened to answer it anyway.
Just how you differentiate yourself when you're selling. What separates your solution from a lot of the other solutions there are?
It comes back to the same point, basically. To a large extent, our target audience is the installer. I think it was referenced in the slide about market share, that one of the things that allowed us to grow was this focus on the installer level. For them, getting everything, getting trained how to sell even storage. That's a big thing. Installers want to sell storage, they don't know how to do it. We invest a lot of time. We brought in some cases, sales gurus and collected our installers and had the sales gurus talk to installers about improving their sales capability. That's the mechanism that we're taking this into the market.
Hi, Eric Liebs. Eric Liebs, Bank of America. Just actually as a follow-up question towards the storage offering, can you just frame for us expectations around attachment as well as market share based on your initial expectations as well as the ARPI increases in the context of storage?
The storage market, for the reasons that I mentioned, I don't feel comfortable in mapping out any specific market share number. It's a complex market with different people selling different components, and any type of specific number would be potentially misrepresenting reality. I can't answer with an accurate number, that question.
Could you talk about pricing in terms of dollar per megawatt hour and-
The storage product that we will be selling, as was presented by Lior, will be available in the market next year, and we'll publish the pricing as we come close and have our full cost structure understood and clear.
Okay, just one more question on that. How do you think about storage as a full storage system relative to your storage inverter that you're currently selling and pairing with, say, third-party battery systems such as LG Chem?
Sorry, I missed-
How do you think about selling as a full system versus just the storage inverter that you're pairing, for instance, with third parties such as LG Chem? Thank you.
Until now, we didn't have the offering of our own battery. What we've been selling was the storage inverter. It's two versions, the version for backup, which as I mentioned, is more common in the U.S. and in Australia, and the version which is named a bit differently, but the version for Europe of self-consumption. Those get attached to many types of batteries, but the most common one, as Lior mentioned, is LG Chem. Our idea is to add to that portfolio a battery from SolarEdge that is obviously going to be more seamlessly integrated with our inverters. Our inverters will still be compatible with other batteries the way they are today, and we expect sales of those to continue.
Hey, Mike Weinstein. Mike Weinstein from Credit Suisse. What's the advantage you get from actually having Kokam integrated within the company and actually making your own batteries that are, instead of going to other suppliers and maybe taking advantage of the free market out there globally, having people compete to give you the best battery instead?
If you looked at the chart that we showed earlier of demand for lithium-ion batteries, you saw, and I discussed three markets: e-Mobility, consumer electronics, and energy storage. Energy storage is the smaller one. It's pretty big for us, but for battery manufacturers, it's the smaller one compared to the two other applications. We've experienced already challenges in roadmap alignment because independent battery manufacturers might seek different opportunities, and then you have fluctuations in supply, or you have misalignment on technology and roadmap because of the fact that energy storage is a smaller part of their business. We believe it's critical, and to be able to develop the products, and as Lior mentioned, to develop the products from the chemistry level for the applications we're serving and to have security of supply in the market, which has been a big challenge in the last couple of years.
Yes. Wait, we'll give you a mic.
In terms of Kokam, you're going to develop your own chemistries. Are you ultimately, if it scales, going to go to a fab-lite or a manufacturing-lite strategy where you develop your chemistries, if you get a lot of demand, outsource it to a third party?
It's a good question that we're discussing a lot. We're taking the first step. It's a market that will shape itself in the coming years. We are confident that the step that we're taking now is needed in order to achieve the objectives that I mentioned in answering the previous question. Demand can grow and can be much higher than the capacity that we plan to manufacture. In which case, we'll have to figure out what method to go long term. Definitely, relying on outside suppliers similar to our SolarEdge manufacturing system is one of the options.
When I eyeball your capacity expansion of inverters versus Power Optimizers, is it fair to say they're both growing about 100%-ish, give or take?
In 2020 over 2019?
Yeah.
I believe so.
A lot of the applications you described today, seems like they would be more inverter-intensive rather than optimizer-intensive. I guess I would've expected, forget the scale of the inverter and optimizer growth year over year, I would've expected a much larger inverter growth versus an optimizer growth. Just can you explain if you move into utility scale, four modules, one optimizer, more commercial application storage with no optimizers. How does that translate?
Our industrial engineers forecasting our business deal with those questions all the time. When you look at between segments and geographies, we're covering such a broad range of sizes of installations. The ratio of optimizers to inverters will vary by the size of the installation. They'll vary whether we're using a two to one configuration like we do in commercial, but not everywhere in commercial, or one to one in residential. There are other elements that go into it. We look at the bottoms-up forecast in determining the exact mix because it's so multi-variable, that there's no one equation that you're going to put in, here's the number of inverters, what does that translate to in terms of the number of optimizers. Also, we make optimizers that are panel-specific in terms of their power rating, so that has an impact as well.
Maybe one last one.
Oh, yeah. Hi. Beyond the chemistry in a battery, Over here. Is there anything else you can do to improve the cost efficiency or effectiveness of the battery design over the roadmap to be a leader as you are in the inverter side?
I'll hand that over to Lior.
Are a multiple parameter beast. Being a power electronics engineer, I don't want to say that they're more complex than developing power in electronics, but it's a multi-parameter beast in terms of it's not just cycle rate, and it's not just amount of current or power that you can take out of the battery. You want to maintain temperature range. You want to maintain a high energy versus high power application number of cycles, throughput of energy. It's a multiple parameter beast. A lot of the batteries that are designed today are not specifically designed for the ESS market. In some elements, it helps in the cost structure, but in others, it interferes with the cost structure, at least of the overall LCOE of the system, which is much more important than up front price.
I think that we've been able to show that in PV inverters, and it's even more critical in batteries. Being able to optimize an ESS solution, especially coupled with solar that has specific behaviors that are not like an EV and not like a cell phone, can help drive the overall LCOE. Yes, there's lots to be done, and that's exactly another reason why we need to have that lower level expertise, because capacity can always be acquired, and you don't have to always install it. The ability to tap into the lower level design of a cell and then a pack and then a system is critical to win this market in the long term.
I'd like to add one more thing to Lior. I think that this is also related to the factory that we are expanding now in Kokam. I think that one of the major issues when it comes to the cost structure is not just the chemistry and what you try to get, it's actually what is surrounding it. It's an economy of scale kind of a game. Today, the main issue that we see, for example, in Kokam, is the fact that since they have a relatively small factory, their economies of scale, first of all, in the purchasing power when they buy the various components, is relatively small. When you look at the ratio of the amount of variable costs, direct labors and indirect labor workforce in the factory compared to the factory production.
To have a procurement person that buys for 2 GW or 200 MW is the same person, but of course, the cost is much more widespread. I think that other than what Lior mentioned before, the ability to optimize the chemistry and the processes themselves, I think that we have a major opportunity in the next Kokam factory to simply take away all of these disadvantages that we have today, as a result of the smaller scale production that we do, and therefore to be much more competitive. I think that this is the major thing that drives us now in order to get to the factory size that we see. If you remember in the call that we had two quarters ago, we discussed about 1.2 GW factory.
I don't know if it was passed unnoticed, but when both Zvi and Uri talked, we are going to increase this capacity to 2 GW . Part of it is assumptions around our ability to consume all of this capacity, but also a lot of it is coming from the ability to reduce the other costs in order to do the work there. I think that unlike other battery manufacturers, Kokam can do very nice work in a short-term period to reduce the cost on the manufacturing due to the higher scale of manufacturing itself. One more question?
I think that since we have a little bit of time, or we're a little bit ahead of time, let's send you to lunch with a little bit of technology taste, and we'll simply have Yoav and Meir, Yoav, our VP R&D, and Meir, our CIO, both founders, to give you a little bit of our technology and where we're heading with it. Yoav.
Hi, my name is Yoav Galin. I'm the VP R&D of SolarEdge, and I'll try to go over some of our technologies and capabilities that we use in order to make our different products. As you probably know, we believe the foundation of our success is actually our technology and our innovation in technology. We try to have something innovative in each and every one of our products. Our novel DC architecture provides the basis for high performance, for the rich functionality of our system, scalability, reliability, and low power, and we'll discuss this in a couple of slides. We develop everything in-house, we have control over all aspects of the system, and this is important if you want to have the ability to have the right product at the right price, at the right functionality. We have control over our own technology.
Most of our products are actually products which are digitally controlled, they all collect information. They are all controllable, thus we collect an enormous amount of data which we can use later on for added value to each one of our different customers. Regarding innovation, we began shipping our first product in 2010. It was our very first complete system, which was targeted for the residential market. It included the Power Optimizer, it included the inverter, single-phase inverters. It included a full monitoring platform. Since then, we have added more and more parts to our system, which enable us to scale to different markets and give value to different customers. In 2012, we had our first three-phase inverters. In 2015, we had our first storage system, which included batteries. In 2016, we had our first HD-Wave inverter, which was the first residential 99% weighted efficiency inverter.
Some other examples, in 2018, we started our grid services platform, which enables utility companies and other customers to actually manage large fleets of systems as power plants. We had our first EV-integrated charger in 2018. In 2020, we'll have our first residential battery storage, residential commercial storage, and so on. We try to innovate and add functionality and components to our system every time we can do so. Actually, each one of those developments, everything starts with the system. I'll try to explain why the storage system is so important and why it gives us the abilities we have. First of all, if you take the traditional PV system, which is used by most other companies, it has a lot of restrictions built into it.
When we design our own system, we actually thought of those limitations and tried to find a way where we can actually overcome those limitations and give something special, which is meaningful, valuable to our customers. During the time of system design, we actually split the system and the functionality of the system into those different components. We decide how those components interact with each other. We decide which component does what in the system. This split actually gives a lot of abilities to our system. Because once you have a good system, you can actually balance stress level on the different components. You can add functionality, you can actually increase reliability by controlling the stress. You can do all of these things while maintaining a low cost, just by deciding what function is done in which part of the system. This is a key point.
Another key point is once you have a good system, it's actually very easy to expand, to add functionality, to add components, to do different variations which help different customers do whatever they want with the system, get the functionality they require out of the system. Our system is based on a DC architecture. It enables to connect different DC components to a shared DC bus. Those components interact with each other without any regard to the actual physical attributes of the components or of the environment. Once you select to work with a component, a specific voltage, it doesn't matter if the temperature shifts, if other environmental conditions change, it still interacts the same way with all the other components. This makes it very easy to add different components into the system.
It also enables us to optimize the performance of each component and the way it interacts. We can add features like safety, which is something which is becoming very important for any systems this type. Our clients can easily utilize the different components without any or with limited restriction. We have in our system one point which connects to the grid, this is the inverter. Having one point which connects to the grid means it's relatively easy to adapt it and to change it according to grid codes or regulations. As we said, we have an expandable DC architecture. System elements can be either producers of energy or consumer of energy. If you take, for example, a PV module, it's a producer of energy. A battery can either consume or produce.
All of those different components can interact seamlessly, via our own proprietary protocol over the system, enable us to maximize the energy produced, make simpler and more reliable components, have a robust system, as I said before, have a single point which connects to the grid, gives us a lot of flexibility on how to adapt to different grid codes, and regulations. One example of the benefits of the DC architecture is, for example, when you connect a battery to a DC coupled system, if you have extra energy, you can push it into the battery, you're not limited by the actual inverter capacity, which is a lot of times limited by regulation and so on. You can enjoy the full potential of your system without any regard to the AC side.
If you have an AC coupled system, you will always have to go through the inverter and whatever its limitation is. Our systems are software controllable on many levels. It allows us to implement many new features easily and to add features as we go along to the existing fleet. We can do rapid compliance to different codes and regulations, whatever they are. In most cases, we were able to do the changes and have current system adhere to those standards. We have fleet level management, and we have the ability to remote assist our installers. In most cases where they have problems, we can actually detect what the problems are in the installation remotely and either solve them immediately, remotely, or tell them exactly what needs to be done. This is all due to the information we collect from the different components of the system.
A couple of words regarding the different capabilities we have in-house, which we use to develop our products. Power circuits, they're the heart of any power conversion or power device. We have a lot of expertise in physical layer topologies for power conversion and so on. Control loops are actually the way you operate the power circuit, how you do the conversion. The physical layer is the power circuit, and the control loop is actually how it is operated. We do all of our control loops development in-house. ASICs are a way to reduce the component count, which increases reliability, and they're also the way to miniaturize and reduce the cost of our systems. We use a lot of ASICs in order to get what we need. We have different ASICs for different things. On our optimizer, we're currently in our fourth generation of ASICs.
Real-time software, which is run on the field installable software, does the implementation of the control loops, all of the different integrations which are done inside the system. Mechanical and thermal design for any power device is important, especially if it's an outdoor unit which has to work in hard conditions. Being able to have the mechanical design which protects it from the elements and allows to dissipate the heat outward is extremely important in order to make sure you have a device which can run in the field for 25 years. Cloud software, we will discuss a bit later. Meir will show some of the examples that cloud-based software enables us to collect all the information to control all of our systems via our cloud. Part of our newly acquired abilities is chemistry and materials.
As has been said to you before, being able to actually control the chemistry of your cells and batteries enables you to get the right cells for the right task. Having EV designed cells for an ESS means you have to pay extra because you need more capacity, because you need more cycles, and EVs are designed for a specific amount of cycles. On the other hand, ESS or residential storage needs much more cycles. If you want to be efficient, you need to design the chemistry according to the specific task. Electric motor design, and not just the motor design, but all of the different components of integrating electric motors and electric drivetrains into vehicles is something that we gained in our new e-Mobility division, and this is also very important. A couple of words regarding the geographical spread of R&D.
We have R&D teams in several locations. We have R&D done in the U.S., in Italy, Bulgaria, Israel, and South Korea. You can see we have different methods and different types of engineers that work on the different aspects of our systems. We have over 330 granted patents in varying fields of engineering, all of them relevant to our own systems, and we have over 200 pending patents. I will give the floor now to my colleague, Meir, to describe the data-driven value.
Thank you very much. Hello, my name is Meir Adest. I'm one of the founders of the company and Chief Information Officer. If Yoav, in his part of the presentation, talked about the hardware and the technological capabilities that we have to develop the products, I'm going to focus more on the cloud software and the values that we could provide based on the data that we collect. It all starts with a single site. We have the Optimizers and other devices sending telemetries to the inverters, which are smart enough to run locally algorithms for managing consumption or optimizing energy usage. This could be done continuously, regardless of connection to the internet. Even if there's no connection, the inverter could run according to these policies.
As most of the inverters are connected to the internet, they continuously transmit real-time telemetry to the monitoring server and receives commands and policies from the central site. If I talked about communication, obviously this raises the question of security. We've developed security into the system from the get-go. There's built-in authentication and encryption in our communication system. I think it's important to note this security, we could really call it utility-grade security. It has been valued by a number of utility companies and utility-scale installations in numerous continents across the globe. It's been evaluated and accepted. Of course, besides the in-house rigorous testing that we do to guarantee security, we also have validation by third parties.
There's periodic validation by security experts and penetration tests, and we also have an ongoing bug bounty program, in which we let the public try and find problems with the system. If I was talking before about a single site, obviously there isn't just one single site. There are many sites. In fact, there are about 1.25 million sites which are monitored worldwide. Based from the tens of millions of Power Optimizers, there are about 3 billion telemetries collected every day. From the millions of inverters, there are over 150 million telemetries collected every day. All in all, over the last about 10 years, we have over 2.7 trillion different measurement points and data points, which were collected from every continent except for Antarctica, across different grid types, different climate zones, and many different types of equipment.
With this information, we could provide a number of benefits to the different stakeholders across the value chain. I want to focus on three of such stakeholders, and for each of them to show both what information they could benefit for and what actions they could do. The first is the system owner, and for them, we could provide visibility into system performance. They could see how their system is performing now, but they could also get an analysis of what the expected performance is, either based on what we see as similar systems performing or by comparing to historic performance of their own system.
Examples of such actions they could do around energy automation, whether it's something relatively simple like remote device operation or more complicated, such as policy-driven energy optimization, deciding when to consume what in order to optimize the energy costs. The second stakeholder is the installer, or in larger installations, EPCs, who could use the information to pinpoint installation faults, and this makes it much easier to troubleshoot and fix issues. Over time, by monitoring the performance of the system, they could detect impending problems and perform preventative maintenance before these problems become an issue. Some examples of operations are remote system activation, remote configuration, and remote upgrades of the site.
The last stakeholder is the grid operator, who could see, across thousands or tens of thousands or even hundreds of thousands points in their grid, the different parameters of the grid. This better visibility enables them to both plan better and enables them to give grid services. I wanted to dive into a couple of examples for these values. The first is identifying performance issues. What you see here on the right is the homeowner's dashboard. Basically, they could see all the modules that they have and how much power each module is producing. A more in-depth analysis can be seen in the middle here, where we could investigate the underperforming modules. What you see here is the power over a day, the power level of each of the modules. Most of the modules are performing as expected, but one of them is underperforming.
The last example on the right here is an example of a system with modules which suffer from PID, potential induced degradation. Basically, that's a phenomenon where when modules are exposed to high voltages over time, they start degrading and producing less power. Indeed, what you can see here is that different modules along the string, which have been subjected to different DC voltages, start degrading over time. A second example of a value is basically for the grid operators. What you see here is a map of California. Overlaid on top of it is a heat map based on telemetries from the SolarEdge systems, which in this case show the PV power production in each county of the grid. Of course, the same visibility could be also given to many other grid parameters.
Grid voltage, grid frequency, power factor, basically allow the grid operator to have better understanding of what their grid looks like at different points behind the feeders, behind the substations, where today not all grid operators have good visibility. Once they have this visibility, it enables them to then address the issues. Since we have this in-depth information of power generation at any given moment, we could couple that with our historic data that we've been collecting over years, on the one hand, and weather forecasting to accurately predict PV production forecasting. This is an example of a service that we provide, basically showing estimated PV production one, six, and 12 hours into the future, with pretty high accuracy, around 98%. As a final example, I'd like to tie these abilities together and kind of demonstrate a grid service event.
What you see here in this example is a neighborhood that's fed by a substation, which is near its maximum capacity. It's marked in red here. The grid operator could predict usage over the next few hours. In this example, it sees that there's expected to be high consumption behind the substation, which exceeds the capacity of the substation. This will basically, if unhandled, cause an outage in the grid. If the grid operator has a SolarEdge Virtual Power Plant, which I think Lior mentioned before, and they have access to the grid services dashboard, they could plan ahead of time a grid event in which they'll instruct the different distributed energy resources, which are centrally managed by SolarEdge, to provide power.
Actually, during that time, the excess power will be provided by the different batteries, this example, maybe even chargers, and prevent the outage without exceeding the capacity of the substation. Okay. Those were some examples of the values. Yoav and I talked about the core competencies of SolarEdge, the power electronics, inverters, converters, chargers, batteries, et cetera, in the context of the solar business. These same competencies are proving to be very useful in the new businesses that we've acquired, whether it's the UPS business, which technically speaking, is very similar to inverters or the EV powertrains, which is a combination of inverters and converters and batteries. These are the key takeaways that Yoav started with. I'm not going to repeat each of them.
I just would like to conclude by saying that our technology innovations are the key for our success, and the data that we collect is also a solid basis for further services. On that happy note, I think we're going to lunch.
Okay. We'll be down for lunch now. The lunch will be followed by a product showcase, an augmented reality product that we have downstairs. What we'll simply do is after lunch, we'll be divided into three groups. Each one will be taken by one of our founders in order to explain and to allow you to ask questions about the product and see a little bit of the augmented reality, which is serving as how we see future coming with our product. Without any further ado, this way to lunch. Okay. This is Lior's nightmare, to be in an event where the finance is the fun stuff. Right, Lior? Where are you?
In the presentation that I'm going to share with you, I'll try to talk a little bit about what we've done in the past and a little bit how you should look at our business moving into the future. I think that through the presentations and the lunch and hoping through the interactions, you were able to see where we are today. We're already a market leader in the inverter space, with still chance to capture, as you saw in Zvi's presentation. We have demonstrated a sequential fast growth that is coming with profitability that is unprecedented in the at least comparable companies around us. This profitability was actually accompanied by cash flow generation that cannot be compared to most companies in our space. This was based on our proven track record of execution. When we saw growth coming, we knew how to manufacture the products.
We knew how to ship those products, by the way, even at a cost of air shipment that impaired our gross margins over time. We felt that this is the right way to continue with our customers and to be able to support them. When you support them, even if it costs you a little bit more, it is something that pays off over the long term. I hope that our growth was demonstrating this one. The result of this growth, the profitability, is a very solid balance sheet. We have today over $400 million of cash and cash equivalents and investment on our balance sheet. We continue to generate cash flow, which will enable us to continue and grow.
The last layer is that the new businesses that we acquired opens to us opportunities in markets that we were not present before, all of them are representing potential for high growth, as Zvi mentioned in his second presentation. What I'll try to do here in this presentation is to basically capture the past and take us into the future. I think that the past was relatively nice for a company that started selling products only about 10 years ago. Over the last five years, we've demonstrated 35% CAGR in revenues, which will translate to about four times profitability, which is also translated to about cash generation from operations of $240 million this year out of about $21 million just about five years ago. This is something that we did through all of this innovation and execution that we managed to bring over the last time.
Let's start talking about the solar business, and then let's talk a little bit about the new businesses, where I will try to give you a little bit of a feeling of how we see the financial future moving forward. By definition, of course, I will not be able to tell you exactly what is going to happen and when, given the more new nature of these businesses and markets in which we didn't operate it, but I will try to share with you how we, as management, see the growth in these segments. Only four years ago, we were showing these slides during the IPO. Some of you sitting here saw this slide.
By the way, even at that time, we saw a lot of skepticism whether we can get there, because when we were showing a 2014 margins of 14%, sorry, 16% gross margin, we talked about 32%-37%. We were talking about operating expenses as a percentage of revenue of 15%-17%, and we were talking about net profitability of 15%-18%. Without writing it there, we also talked about double-digit growth that we are supposed or expected to generate over the next few years. If you remember, we used to say that long-term for us is about five years, not much more, because then it becomes very, very long-term. This is what we've done over time. In all of the years since IPO, since we're out of the gate, we demonstrated growth rate that is higher than the 15% threshold that we put.
Actually, in the last two years, the solar business grew rapidly at a pace of approximately 50% year-over-year. This is of a company that starts from a base of $600 million and $1 billion. To imagine how hard it is, you saw Uri's presentation. It's not just taking the Excel and saying, "Okay, now we're growing 20%." This is increasing the amount of machinery, amount of labor, amount of people, and amount of people in SolarEdge making sure that all of this growth is accompanied with the right quality and reliability of products. We were able to grow gross margins to about 35%, and actually, in some quarters, if you remember, we were already at 37% and 37.9%. Sometimes, two years ago, we even changed the tone and said that 37%, give or take 1%, is the long-term gross margin that we should see moving forward.
We took it back a little bit because of the tariffs. We were able to take all of the tariff effect and pass it along to our customers. This comes at an arithmetic effect on the gross margins because now we took the same profitability but simply divided it by a higher denominator, so growth margin was a little bit lower. Eventually, in profitability, you can see that operating expenses as revenues were slightly more than we expected over time. When you look at the net income on a GAAP and a non-GAAP basis, which is a little bit more pure from the economic point of view, we were able to be at around 16%-17% of net profit after taxes.
This is after taking into account the tax reform in the United States and the fact that we see today a lot of movement in the foreign exchange currencies, and being a relatively large company that operates in many markets, 28 markets, as Zvi said before, is something that takes its toll when it comes to currency exposures. There is still a lot of growth ahead of us, and we believe that we can basically now reaffirm a lot of what we said during the IPO for the next few years to come. I would like to address it by looking at the revenues, at the gross margins, operating expenses, and of course, the result is going to be the profitability. The main elements of our growth is going to come from three major areas. The first one is geographic expansion.
In 2015, when we went public, the United States represented 73% of our revenues. By the way, one customer called SolarCity represented 32% of our revenues at that year. Over time, we were able to expand beyond the United States, not by reducing the amount of sales in the United States, but actually by increasing dramatically the revenues in all other regions. As we saw the first three quarters of 2019, we're already at 46% revenues coming from the United States, 43% coming from Europe, and about 11% coming from the rest of the world. Of course, the rest of the world and Europe are growing relatively quicker than the United States. We also talked about it before.
We see in the future a situation where the U.S., Europe, and rest of the world, each one of them are contributing 1/3 of our revenues. Again, I'm talking only about solar, of course, and this is without reducing the amount of revenues in each and every geography. The second area that we see, and by the way, you can see that both in geographical growth and both in segmental growth, we're not just growing as the market, we're actually growing a little bit faster than the market. We simply take market share from some of our competitors.
When we look at where we are today from a segmental point of view. If in 2015, again, going public, we were at 90% residential, 10% commercial, the new suite of products that were demonstrated below brought us to a situation today where about 78% of our revenues are coming from residential, 21% come from commercial, and we saw some springs of utility still using commercial products, but utility installations. Long-term, utility, residential, commercial are forming approximately one-third of the revenue pie here or worldwide. We're talking about approximately $8.5 billion of inverter market that is divided between these three technologies. We are developing our utility products. We are continuing to increase the capabilities in the commercial space. We look ahead at a situation where one-third of our revenues will come from utility, as much as we will see it coming from residential and commercial.
The last thing is ARPI. ARPI comes through the sale of new products. It comes from the ability to serve higher or a bigger amount of the pie. By the way, we'll give the printed slides at the end of my presentation. I just didn't want you to go to all of the surprise at the beginning. Basically, we can continue and increase the ARPU over time, and when we look ahead, we believe that we can be in a situation where two-thirds of the business in the solar space will come from our inverters and optimizers, but 18% and 16% will come respectively from storage and smart modules that will have our products on them. The combination of geographical spread, increasing the segments, and increasing the ARPI will allow us to continue and grow our revenues. The second area is the gross margins.
The gross margin that you see today is about 34%-35% in the last few years, are affected today by three major areas. The first one is air shipments. If you take the cumulative effect of air shipments over the last quarter, and we actually discuss each one of them in a discrete manner in each and every conference call, we are today in Q4 going to be affected by approximately 500 basis points simply due to air shipments. Once we have more capacity online, we will be able to gradually get away from this addiction to air shipments, and we'll be able to take them away or at least reduce them to a minimum. This is about 500 basis points.
In addition to this, if you take the effect of the tariffs, the arithmetic effect in the fourth quarter of this year, it's about 185 basis points of effect on the gross margin simply because of this net arithmetic effect. By moving into Vietnam, moving into Flextronics, and having some capacity coming from Sella 1, we will be able to bypass this, and therefore to avoid this effect. We will, of course, transfer some of the prices back, but this will get us back to the gross margins that we used to see before. Therefore, when we look ahead, the 36%, give or take 1%, is something that we believe that can continue, and this is also taking into account more of the new products that are included in the ARPI.
Some of them, of course, are products with higher gross margins, some of them are products with lower gross margins, but the blended effect is approximately 36%, give or take 1%. Of course, operational leverage is also important again. We are investing all the time to grow our infrastructure. The first thing that we do is R&D. We are a technology company. If you do not invest in technology, as Guy used to say, "You're always as good as your next technology. You are never as good as the one that you had before," because everyone can basically develop it over time. Increasing our R&D spending is something that we would like to do. We continue, of course, to invest in geographical footprint. Our sales force is growing. Of course, you need to support all of those with G&A.
The way that you should look at the spread of these expenses moving forward is that we would like and hope, by the way, to increase R&D at 25% year-over-year. It's not very easy. When we have today close to, if I'm not mistaken, 800 engineers worldwide, to bring another 200 engineers every year is becoming a task. I hope that we do a good job in it, but it's not getting easier. We would like at least to grow by 25% year-over-year. Sales and marketing expenses should grow at a lower pace than the growth in revenue because we do have economies of scale.
If we have a Dutch customer to which we used to sell, let's say, $10 million in 2017 and $20 million in 2019, it's still the same salesperson that takes care of it, and therefore economies of scale happen. Lastly, without taking litigation expenses, G&A should grow by not more than 10%, leveraging on more and more processes and more and more computerized systems that we develop in-house or buy from the outside world. The result, of course, should be, as you can see in the green line, is to see operational leverage that is actually having our profitability going up. This is what we see for the years to come. For the years to come, meaning for the solar business for the next three, four years. We continue to feel comfortable with growth rate of 15%-25%.
On an anecdotal basis, if you look at the growth rate that is now projected to us and as was projected to us in all of the years since we went public, usually the Street was a little bit more conservative than ours. In any case, we're projecting a 15%-25% year-over-year growth in revenues coming from all of the elements that I described. We believe that we'll be at around 36%, give or take 1%. This will be after taking the air shipments, but directing some of these margins to increase other markets. If we want to bring new products at a lower margin, we can do it. If we want to penetrate a new region where the penetration cost should be a little bit more expensive compared to the more developed regions, we can do this.
We basically have now another kind of, I would call it, button that we can play with in order to have our growth continued in the same pace that we used to see. Operating expenses as percentage of revenues should continue and go down. Some of it is coming, on a GAAP basis, at a higher percentage that you see here. When stock price is higher, the cost of the expenses associated with RSU grants are a little bit higher. On a non-GAAP basis, we're already there, and we will continue to push this down. This will lead us to an operating income of 20%-23%. We don't talk about net income anymore because we do see the currency exchanges that are beyond our ability to control are happening. The world is becoming crazy when it comes to taxation.
On an operating income, we feel comfortable to be at 20%-23%. This was the solar business. Let's talk a little bit now about the new businesses. Of course, by definition, going into new markets with new companies that you just acquired involves a lot of uncertainty. This uncertainty comes from how quickly you're able to monetize these businesses and how quickly you can increase the business. I can say that none of the businesses that we acquired is an EBITDA-generating businesses. All of them were, in a sense, losing. All of them required investment. When we look at the revenues from these businesses, we expect them to be about 10% from 2022, 10% of total revenues from 2022, and we expect them to continue and grow at a double digit thereafter.
I would say that looking at the growth that we see in solar may be reasonable, but of course, the potential in each one of them is higher. As we will learn a little bit more, we will give you a little bit more color into it. We simply don't want to guide on something that, today at least, we don't know how to project very effectively. We feel very comfortable with this. The Kokam factory is even, or supposed to be easier to project if we're able, indeed, to meet the timeline in expanding the factory and if we're able to utilize or actually materialize all of our expectations. We believe that with the Kokam factory, the vast majority of this factory can be taken by SolarEdge for its own applications. This is the batteries that you saw downstairs, both the residential and the commercial.
There are ESS systems, of course, out there, that Kokam is using. We believe that in the future, again, UPS could be another area where we can take our batteries. If we're able to take all of the capacity there and assuming $250 per kilowatt installed ASP, this will be something exceeding $300 million in 2022 when the factory will be ramping up. Once we're in full ramp-up capacity, we can be at $500 million of revenues based on those numbers. From gross margin point of view, again, we're trying to give you a little bit of the way that we think of it. The UPS products, as Zvi mentioned before, are very similar to the products we sell today. These are batteries, these are inverters.
We believe that they should share the same gross margin as the solar businesses, and we expect it to be 30% by 2021. Why? It will take us a little bit to get there, simply due to the fact that we're now taking these products, and we take them from the old manufacturing methodology of Gamatronic, and we transfer them to the contract manufacturing and to the manufacturing capabilities that we have today. It's a process that takes a little bit of time, but we feel very comfortable with this 30%. Looking at Kokam, we're looking at 25% gross margins. 25 gross margin is coming due to the fact that once we have the factory, the economies of scale will allow us to be at a cost that are similar, again, to the other players in the industry.
We're not going to be maybe as cheap as someone that has 70 GWh or 40 GWh. When you build a factory, usually you have the building blocks, which are the machinery. Based on the bottlenecks of the, let's say, widest capability machinery, you build the building block of the factory. We're building or trying to build the most efficient factory that will allow us to have the best economies of scale based on the size that we have. Of course, we'll always have more expenses when it comes to the indirect expenses compared to the large companies. We feel that a 25% is achievable, and we believe that this can be the case in 2022. Lastly comes e-Mobility. e-Mobility is expected to be a lower margin business, at least at the beginning.
As Zvi mentioned, it's not something that will come very quickly because of the processes that it takes the car companies or the vehicle companies to adopt the product. We expect it to be 20% because these are products that are usually using a higher grade of products, and usually, you see that the large auto manufacturers usually have buying power that is higher than the usual customers that we see today. We see about 20% gross margins by 2023. All of this has to be translated eventually to profitability. In Kokam, we believe that we will be profitable by 2020. In Kokam, we believe that we can be profitable by 2022, both in the UPS and the e-Mobility businesses, compared to a situation where at least last year, all of them were not profitable.
A byproduct of our P&L results is our balance sheet. Today we have a strong balance sheet. We have a balance sheet that has sufficient amount of cash to take all of our expectations about expansions of factories or geographical growth. We have financial ratios of financial stability that are very high compared to most of our peers in the industry. Just look at the debt-to-equity ratio that you see there compared to some of our peers. Actually, all of this debt is debt that came from our acquisitions and not debt that we took as SolarEdge. This is something that, of course, we feel that this stability and balance sheet strength is allowing us to continue and grow.
One thing that we would like to do with this balance sheet is to take it and help us to grow the business beyond what we see today and to basically allow us to fulfill all of our plans. The first one is the growth. The growth that Uri showed you in his presentation is growth that requires CapEx installation. CapEx is coming from two major areas. The first one, automatic assembly lines. By 2020, we expect that by Q4 2020, all of our Power Optimizers will be manufactured using automated machinery. This is something that will cost us money as well as the testing equipment. We're looking today at about $32 million-$37 million in the last two years. We believe that next year it will be about $40 million, where we're going to increase capacity much more in the Vietnam and Zala areas.
The second area is the Kokam factory. Two quarters ago, during the earnings call, we talked about $50 million-$60 million. We talked about a 1.2 GWh factory. Over time, when we continue to analyze the factory, we believe that the 2 GW is the right measure for us to be in. It comes from various directions. First of all, we believe that we can take 2 GWh of product to our applications when we move forward. We also analyze the production and the cost structure. Of course, since economies of scale are playing out here, the bigger the factory is, the better you are on the cost per cell.
Therefore, we will build a factory with 2 GWh capacity, but an infrastructure to support even small growth out of this 2 GWh, that if we will see that demand is coming at a higher rate than we expected, we will be able to quickly respond and increase these capabilities without having to go and build another building or buy a different land. The Sella 1 factory in Nazareth is going to consume approximately $45 million-$50 million. Most of it will be paid by the beginning of next year and the end of this year, 2019.
This Sella 1 factory will be allowing us to continue and develop our products to be very close to our R&D and will enable us to continue and develop our automated machinery in order to make sure that then we can Copy Smart all of these methodologies into the contract manufacturers and the other sites that we have around the world. Of course, on top of all of this, we need to take into account $5 million to $6 million of annual spending on CapEx simply coming from the fact that new employees are coming. They need laptops, they need a place to sit, and they need another, sometimes, furniture. I think that all in all, you can see that we have a CapEx plan that is relatively extensive for the next two years. Our balance sheet allows us to be in this situation.
With this, I'll go back to where we started. I think that where we are today and what we were able to achieve puts us in a way in a situation where we can leverage our financial stability and financial strength to continue and grow our revenues at 15%-25%, to be at gross margin of 36%, give or take 1%, and to have operating income of 20%-23% on the solar business and continuing to increase all of the other elements in order to be able to show continued profitability and growing profitability in the future. With this, I'll open it to questions. Phil.
Great. Thanks, Ronen. Phil Shen with Roth Capital Partners. In terms of your 15%-25% year-over-year growth targets, can you walk us through your unit and pricing assumptions for the solar business? Do you expect pricing to decline at all? In terms of margins, historically, I think they're GAAP margins that you were talking about there.
Yes.
Do you foresee a time at some point in the near term where you maybe step away from non-GAAP, especially the stock-based compensation? You're very profitable, we need to be talking about that going forward.
Sure. Let's start, first of all, about the GAAP and non-GAAP thing. The main problem today is that the accounting is becoming a little bit more hard for us to explain how the core business looks like. With the new accounting principles like the leasing one, when you have to take long leasing agreement that you have with another company, and you need to present one side as an asset, one side as a liability, you need to accrue for expenses on the liability, but you cannot enjoy anything on the asset side. The accounting do not really present where the business is. By the way, when you do accounting for M&A, so many things like technology, like customer relationships, are being baked into the calculation of the P&L, and we simply do not know how to show you how we, as management, look at a business.
As long as we look at the business on a non-GAAP basis, we're trying to give you a flavor of how we look at it, and this is why we continue to provide GAAP, non-GAAP. I believe that unfortunately, as accounting becomes more economic in sense and less realistic in many senses, I think that we will have to stay with the GAAP and non-GAAP. One thing that we will continue to do is to give you all of the breakdown, and you see it, by the way, in all of the tables that we publish at the end of every quarter, so you'll be able to judge by yourself what's supposed to be there and what is not. First question you'll have to repeat because I forgot it.
What are the pricing and unit assumptions for 15%-25% year-over-year growth?
I think that it is becoming a little more problematic to project those, and I will tell you why. In the solar business, because we both have the solar business and we have the new products, which are representing the bigger ARPI that we see. On the solar business, we see today, and we do not expect to see major difference, we see today a relatively stable environment. You see an environment where most of the players in the industry were capacity constrained. You see that some of the players in this industry are still suffering from either component issues or tariff issues. Actually, you see that other than SolarEdge, nobody is making meaningful money out of their business, and therefore, the rationale of decreasing prices is not really present there.
Today, when we look at this growth rate, we assume that we will see a relatively small ASP erosion. When we talk about relatively small, this is lower than the 5%-10% that we used to see in the past. This is something that, at least now, we feel comfortable with. As for the unit growth, I'm not sure that I can give you a very good answer because, again, it continues to have every unit that we sell today, and Zvi showed it in one of the presentations. We sell much more units that are what we call complex units, and therefore, it's not just a game of how many units you sell. Actually, what is the content of the unit and what is the price of the unit when you sell it?
I'm not sure that I can give you enough granularity on this one.
Great. One more and I'll pass it on. In terms of your new businesses, can you talk through the OpEx assumptions for those new businesses? With the financial targets much more clear, what is the expected return on investment that you see by acquisition that you've made? Thank you.
From OpEx point of view, what we saw right now is that the numbers that we gave in solar are going to be pretty representative, also assuming the new businesses. The reason is that on at least for the next two, three years, the OpEx on these businesses starts from such a lower base that even if we increase it by 2 times more than solar, the effect on the overall solar expense is relatively lower. Therefore, I would assume that with the new businesses, you should see very similar effects. On R&D, we will see bigger growth in those new businesses, but the R&D expense is relatively large, and therefore, it will be swallowed there. The second issue is the sales and marketing.
We have today already infrastructure in many places that SolarEdge didn't use to have when it was at the same size as the new businesses. For example, if today we want to take a new person for our, let's say, UPS in, let's say, Italy, I don't need to form a subsidiary. I don't need to take an office. I have all of those here. Therefore, I would say that the new business assumptions should be the same, both for R&D and sales and marketing, and both for G&A, where all of these companies do not have the publicly traded company expenses related, and therefore, they can be at least 10%. I feel that it's about the same. The beginning was?
Sorry, about the Gamatronic on the-
Here, I'm not sure that I have a good enough answer. We try to tell only things that we know already today. Today, I can tell you that the only thing that we figured out is that first of all, Kokam factory should be around 2 GWh , and this means this $80 million-$90 million of CapEx. The assumption, if you take, for example, this $500 million and 25%, you'll see that basically the ROI on this one is going to be once we have the factory, about two years. I would assume that if you take the time until we get there, you should see about two years from the time that we have the factory to have a full return on the Kokam asset. When it comes to the Gamatronic, the amount was very small. We paid approximately $11.5 million.
I think that if we're able to grow, and once we reach profitability, you should see it relatively quickly. The only thing that it's hard for me to say is the e-Mobility, because as Zvi mentioned, some of the processes, especially of homologation certification by the automotive companies, are long enough that are not necessarily projected by us. I cannot tell you exactly how much is it. In general, this is kind of the business that once it happens, it's supposed to happen in relatively large volumes, and therefore, you should see a relatively good ROI on this one. Next question. By the way, questions can come also not on financials, if there are anything that you would like. Yes, please, Mark.
Hey, Ronen. A lot of moving parts in 2020 with the capacity expansion and the new products and seasonality from safe harboring. Can you just kind of give high-level thoughts on how we should think about seasonality next year of revenue?
Safe harbor, first of all, let's differentiate between the U.S. and non-U.S., because as you saw, the non-U.S. business is already more than 50% of the business. Usually you follow the regular seasonality in this market. That means that usually Q1 is relatively small, Q2 is strong, Q3 is stronger, Q4 is usually flat to down in most cases. In the U.S., the safe harbor is a little bit changing it, because the safe harbor basically prescribes that in order to enjoy the ITC at the higher rate in 2020, at the same rate as it was in 2019, two things need to happen. Either you need to buy equipment this year in 2019, or you can buy equipment where you put an order by the end of 2019. You need to pay for this order and you need to make a firm commitment.
You can get this equipment towards the first quarter and up until 15th of April. I think that what you will see, if the U.S. market was usually more skewed towards the second half of the year, I think that you'll see a little bit more flattened seasonality in Q1 and Q2 because of the safe harbor. To put everything into the right perspective, the ones who can really enjoy safe harbor are only the very large players, only those who actually own the asset eventually. By definition, not all of the U.S. market is going to take safe harbor orders in Q1 or even in Q4 this year.
Okay. Just lastly, just to be clear, the greater than 10% of revenue from new business in 2022 and double digits thereafter-
Yes.
That's solely from the three acquisitions you've done?
Yes.
Okay. I guess, how should we think about the likelihood for more M&A? Are you just kind of focused on integration at the moment?
Zvi, I don't know if you want to comment on M&A, but in general, we were very acquisitive even before we bought these companies because we, from the days of the IPO, thought of what are the kinds of growth that we can see, and one of them was, of course, acquiring other companies. I can tell you that we bought three companies in areas that were interesting for us at a relatively short time, and that means that we now, as management, take a lot of effort to integrate those, to learn those, and to be able to bring the values of SolarEdge to these companies.
I can tell you that, while we're always looking at new companies and we continue to do so, once an opportunity will come, if this is an opportunity that makes sense on the price, on the ROI, in a field that is of an interest for us, and we believe as management that we can handle the integration of it's something that we're totally open to do. I don't know if there is a very large stone that we see right now in front of us, but some of these things are simply happening over time. I wouldn't rule it out. At the same time, at least when we're looking at more acquisitions, we'll also look at what we've done until now with the acquisitions that we already made and whether we can actually afford one from management perspective. Yes.
Hey, Mike Weinstein from Credit Suisse. I guess the $242 million of cash generation is going to fund the CapEx program going forward. You don't see any additional need for outside financing? Just want to make sure.
I don't think that we need. Right now, again, we look at about $200 and something cash requirements next year, but of course, we believe that we will continue to generate. It's not only what we have right now on the balance sheet. We believe that we are fairly settled with what we have right now from cash perspective.
Just a quick question on the cash generation. What's the difference between the $207 million of operating income and the $242 million of cash generation for this?
Between the income and the cash generation? It's OpEx and expenses related to the PPA, purchase price allocation from our acquisitions. Also, by the way, depreciation, but to a smaller extent. Yes.
On the Kokam run rate revenues of $300 million and $500 million, I think one of the prior presentations anchored to a $250 per kilowatt hour number. Is that all external revenue or is that intercompany as well?
No, no. First of all, Zvi mentioned $250 per kilowatt installed, he talked about, again, 2 GW . This is how we get to the 500. We believe that everything that we say here is going to be reported to the outside world because whatever we do as intercompany is going to be eliminated in the consolidation, and since we are practically holding almost 100% of Kokam, we hope to hold all of these 100%, most of it will stay on us. Everything that you see here is external. It's not intercompany.
If you deploy a residential system with a Kokam battery, a SolarEdge inverter, those are just going to be accounted for totally separately, not as one unit?
No, no. Again, we will account all of those. Again, we'll have to decide how to report it. In general, this will be all revenues of SolarEdge, the company. In the way that companies are working, Kokam will be selling to SolarEdge, and SolarEdge will be selling to the outside customer. All of the intercompany transactions will be eliminated in the consolidated financials.
One last quick one. Do margins get worse before they get better at the non-solar businesses as you invest and ramp?
I don't think so. Right now, all of them started in a relatively bad point. For example, if you take the Kokam factory, the Kokam factory was very inefficient when we acquired. First of all, the amount of utilization was lower. Second, when we came, we did some efficiencies there. When you look at the UPS business, most of the manufacturing was not done using a high level, I would call it contract manufacturers, and the e-Mobility was done relatively internally. Once we are taking all of these businesses to us, two things are happening. First of all, they enjoy our economies of scale. When we're buying a component that Kokam is buying or that UPS is buying, we buy those in millions while they buy in thousands. We get much better benefits, therefore, we can use our purchasing power.
The second thing is that once we move into contract manufacturing, we know how to better utilize and get better prices for them because they see very large volume on the solar side. I think that the trend should be going upward. Yes, Brian.
Ronen, if we look at some of these pie charts, revenue growth by segment, obviously utility is a very small piece today. It looks like it's going to be a relatively big piece going forward in the growth algorithm. I know the product itself is sort of pending, you're not committing to any specific date in terms of official launch, how are you comfortable with the sort of 36% ± gross margin range for the solar business, given that you've got this whole new sort of end market, which is historically known for very low prices and lower margins?
It's like, you take a few ingredients, you put them into one cake, and something happens that is a little bit different from them. I can tell you that in general, the 36% gross margin that we guide here is not only built on this segment but also geographies. Some geographies have better margins than the others. Some new components on the ARPI, higher ARPI, have better margins, and we basically mix all of them together. When we look at our approach towards both residential, commercial, and then utility, you see two trends that are happening in the price, but they also happen in the cost. The first thing is the economies of scale. The bigger the inverter is, the cost per watt for the inverter itself is going down dramatically. That's number one.
The second issue is how many optimizers we are attaching to the inverter, or what is the ratio between modules to optimizers on this one. In residential today, we do 1-to-1. In commercial, we do 2-to-1. We already presented a power optimizer that knows how to support four modules. That means that when we will move towards utility, first of all, the margins may be slightly lower, by the way, it shouldn't be. Again, as we said, we're not committing yet to the price. The margin is still far away. In general, by concept, we should be able to drop the cost of both the optimizer and the inverter and play a little bit with the ratios so the overall margins should be similar.
Of course, again, once we'll have the product out and we'll see the prices, we'll guide about exactly how it is.
Okay. Just on the product category itself, I was kind of surprised to see the Smart Module. It's going from 0% to 16%. I know your peer has done a lot in that space. Can you kind of level set us as to where you're involved with module makers? What kind of visibility you have to that product category becoming a bigger part of the growth curve for you guys?
Maybe, Zvi, if you would like to answer a little bit about this.
Just to make sure that I understand correctly the question. We are not relying on the module makers to be the ones that are selling the product. We are actually the ones selling the product under our brand and through our channels. We have much bigger control of how this market will eventually evolve and the capabilities to grow it. Does that answer?
Yeah. I guess then, can you level set us as to sort of what the capacity is for that? Are you doing that through a partnership or-
Yes. As mentioned-
Moving into the module manufacturing business?
We are buying the modules from tier 1 bankable manufacturer. It is a product that is designed to our specification, and we have a lot of the control of the process, but we are not the manufacturer on our own today.
Plus, we don't have any of these hard commitments for product, meaning that, for example, in module prices full a year and a half from now, dramatically, we don't have any long-term obligations right now that can incur this one.
Thank you.
Yes. More questions? Anything? Yes, Eric.
Hi, Eric Liebs, Bank of America. Just to talk a little bit more about the revenue growth by product category. Could you just discuss expectations for storage and energy management growth? I know you have your long-term target of 18%, given that you're going to begin selling this in a more meaningful way in the first half of 2020, can you just talk about near-term expectations on the revenue composition there?
First of all, the product will come actually at the later part of 2020, not at the first part because the product is still under development and in testing, and therefore you will not see the very beginning. Again, it's a little bit hard to answer this one simply due to the fact that while everyone talks about storage, we haven't dipped our feet into selling batteries before. I can tell you that while Zvi, Peter here, the team, when they talk to customers, all of them express their desire to have as many products as soon as possible. We would, first of all, like to see how this is forming up.
In general, I can tell you that when you look at the overall cost of a system with a battery and without battery, Lior, the cost of inverter and let's say battery, what will be the ratio between battery and inverter in price, let's say, on a five kW system?
A battery is much more expensive in price. I think the ratio would be X5 or anything between X4-X6, something like that.
Even if you take relatively small percentages of attachment rate that you take, it's coming into a relatively large revenue number. Again, the reason, and you saw it in the past, we don't try to be overselling here on something that we don't sell yet. We believe that the potential is high. We need to basically dip our feet into this storage water and see how much we can sell, but the opportunity by itself is relatively large, given the size of the battery compared to the size of the inverter when it comes to the price structure. Yes.
Hi, Maheep here from Credit Suisse. Ronen, Hello.
Yep. Hello, Maheep.
Just on the Kokam plant, could you just talk about the debt for the Kokam plant? How much do you expect to add? Because I think in the past you have said that given in South Korea, you need to raise debt over there for Kokam specifically. Following to that, how do you think about incremental CapEx for Kokam beyond the $80 million-$90 million on a run rate basis to match up to industry standards?
I'm not sure that I know the industry standard in order to compare Kokam. I can tell you that at least when it comes to the overall investment that we do there, what we try to do, and this is why it took us a while, and I know that some of you were trying to chase us. I know that some of you were frustrated with our inability to give this data before, but when we looked at the manufacturing technology that we're going to use in machinery, we try to take a newer generation than compared to what most of the players are using today. Therefore, we believe that on the CapEx side, we could be a little bit less expensive than the CapEx that most of the players are doing today.
One of the things that you see, especially when you run big volumes, and this is part of our SAM, is to have harmonization of the manufacturing. You don't want to have many factories running with many different kind of equipment, and therefore, once you make a decision, you are pretty much tied to this decision for a relatively long time. Us coming as a relatively newer player into this market allows us to choose the manufacturing equipment and methodologies that could be a little bit more cutting edge, therefore reducing both the cost of the operation and also the cost of the CapEx there. I would assume that, beyond this $80 million-$90 million that you will see on the 2 GW , you shouldn't expect a very large expansion in CapEx once we would like to move forward.
Actually, once we would like to move from, let's say, 2 GW to 3 GW, you're going to have a relatively small increment because some of the equipment that we buy already today has the capacity of 3 GW. It's just, we took the equipment that the bottleneck today is about 2 GWh , and we left it there. You shouldn't see a lot of increment. As for, I think the beginning was about manufacturing in Korea, right?
Yes, just the debt in Korea.
Debt in Korea, we need to decide. We have the money to invest. You saw it. It comes now to taxation, and again, the tax world becomes to be very aggressive and very hard. In general, whatever you put into Korea today, you need to invest in a form of equity and a loan because there are thin capitalization rules there. Usually the rule of the thumb is that you put 2 units of debt on 1 unit of capital. If I'm to invest $90 million, if I put $90 million in, at least $30 million will be kept as a equity that if one day I would like to repatriate, and I would expect that we would like to repatriate those monies to the U.S. company simply because Kokam will be profitable enough, I will have to pay 25% dividend tax on those ones.
Our approach will be very easy. We will try to take as much as possible debt in Korea, assuming that we're able to get this from Korean banks, and to have the minor possible investment into Korea coming from monies that are coming from here in the United States. I hope that, the fact that we have so much cash here in the U.S. company, will allow us actually to get better terms on debt in Korea. I can tell you that from at least the initial discussions with Korean banks, they're interested, but I need to see if this is really materializing into a debt that we can take.
Got that. One small housekeeping, just on a tax rate assumption going into the forecast for the U.S. market, just given it's a big market. A high-level question on the cash, the $250 million, around $250 million of cash. How do you think about capital return or use of that capital? Because your cash more than covers the CapEx needs.
I'll start with the tax rate. Tax rates, we historically said they're 15%. I believe that it should be anything between 15%-16%.
The tax credits.
The tax credits here. I'm not sure that I understood the question, sorry.
No, sorry. Just your assumptions around the tax credits, do you have any extension?
Oh
Any probabilities around the extensions or anything like that?
Hardly know Israeli politics and lawmaking. Commenting on something in the United States is going to be far beyond our abilities. I can tell you that as a company, this is our view. We would like to see as little government intervention in the market as possible. I can tell you that from our point of view, we're happy with whatever the result is. What we used to see, by the way, and we saw it in 2015, that when the ITC was extended, you actually saw a lumpy year coming the year after, because now, nobody was fearing from the next step down of the ITC, so they felt comfortable not to increase the installations. I do not know what's going to happen there, especially on the lawmaking. For us, whatever it is, we'll be ready and we'll be able to support it.
That will be related to this. About cash position, as you saw and hear in the presentation, and I hope that we were able to show this throughout the presentation today, it's all about growth. We believe that although we grew very rapidly over the last few years, we can continue and grow in a relatively fast pace. Growth requires cash, and there is nothing better for a company than having all the cash it needs to grow without having to go and be at the mercy of the capital markets or debt markets out there. I would tell you that at least conceptually, first of all, working capital, then will come conceptually, CapEx investments, then will come M&A, and capital return will be at the last point of those.
I believe that at least between the two or the three of the beginning, we have enough for the next few years. I won't commit on anything, but this is the way that we see the priority. Yes, I think that we have time for another one. Yes. We have more question? Yes, Jonathan.
Can you help us understand the decision behind doubling your capacity in units in 2020 and your guidance for 15%-25% solar revenue growth?
Of course. It's a combination of several things. First of all, we say in Hebrew that whoever was burned from a soup is now on a yogurt. It's like growing on yogurt. We've burned a little bit due to the fact that we were very, I would say, quite responsible on the way that we planned our expansion, and therefore, when we saw the growth coming in, it came at a very high price of air shipments. Now we basically face three issues that we need to solve. The first one is that I'm paying right now close to 550 basis points of air shipments, and I do not see the demand of my products going down. What do I need to do? First of all, we need to manufacture enough in order to ship all of the products to the customers.
We need to ship to have excess capacity to start generating inventory that we can put on a boat that will take a few weeks until it gets to the United States or to Europe and build the inventories there. We must have excess capacity in order to eliminate the air shipments because you need to start rolling this flying wheel, I would call it, of cargo marine shipments compared to air shipments. That's the first logic. The second logic that we have there is the fact that, again, some of the growth needs to come not only to satisfy what we need now, but actually to have a kind of a safety plan.
For example, if tomorrow morning there is, I do not know, something happening in China, we would like to have a little bit of redundancy in the other areas in order to make sure that we can continue and supply our customers. There are so many customers that are today almost exclusively dependent on us. It's our responsibility to have the geographical spread, that means that if you have a large capacity in China, you need to build large enough capacity in the other locations as well. The third issue is the fact that, again, we were surprised by growth before, and although we guided in 2015 for 15%-25%, we actually grew much more, and therefore, we're a little bit more cautious now and want to make sure that if something happens, we're able to meet it.
There is one thing important to remember, though, here is that a lot of this investment is not done by us. It's actually done by our contract manufacturers, and they feel comfortable enough with making this growth with us as well.
Thanks. I think the third reason was the best one.
Okay. All of them are true, so it's pick one. Yes. This is it. Zvi, maybe closing remarks before we let people drink something?
Yeah.
Thank you.
This will be one minute because we really stretched, at least my talking abilities in one day. What we set out to do today was to share with you our, I would call it our plan, our view, and our vision. Our plan in the short term, our view for the next few years, and our vision for the long term of the company. I hope that we got that message across in a clear way. On top of that, we wanted to use this opportunity to expose you to the people.
The people that have helped us materialize the vision that we had in the past. It's a very unique situation, I think you would know of startups, that 10 years later, everybody is still on board and everybody is determined to carry us to materialize the view and the vision that we have for the future. I hope that you got a good impression and confidence from meeting the whole team or most of the management team of SolarEdge. We are very excited. We're very proud of what we together, of course, with all of the other employees and under the leadership of Guy, have been able to do with the company in the last 10 years. We're very excited about the potential and where we can together take this company going forward.
I hope we were able to share that excitement as well, maybe in a bit of a mellow style. We are, and I think that was evident by the first slide that I showed, we are a company that prides ourselves in more in doing and less in talking. We probably have an obligation to talk to you more frequently. This was our first such event since five years of the IPO. We will aim and plan with all of the safe harbor protection to do this at least once every two years and get together in this forum and share in a similar way our plans and vision at this periodic level. If there are no further questions, we'll try and get back to doing, and you guys can go do some drinking. Okay, thanks.