Good afternoon, and thank you for joining us. My name is Ki Bin Kim, Head of Investor Relations at Lineage. Today's presentation will focus on Lineage's warehouse productivity initiatives, in particular, LinOS, our internally developed next-generation warehouse execution platform. Our goal is for you to leave with a deeper appreciation of our unique operating and technology platform, which we believe will continue to set us apart from our peers. Now, I'd like to introduce our executive leadership team. All, except for Rob, who just joined us as our new CFO, have over a decade of experience with Lineage. Starting in the middle, we have Greg Lehmkuhl, our President and CEO. We'll kick off today's agenda by discussing the state of our industry and how Lineage is best positioned to win over the long term, and then to my right, we have Jeff Rivera, Chief Operating Officer.
We'll share our warehouse productivity achievements, beginning with our approach to people and processes. And Sudarshan Thattai, our Chief Information Officer and Chief Transformation Officer, will cover Lineage's technology edge and digital enablement journey. And to the far right, we have Elliot Wolf, Vice President and Chief Data Scientist. We'll dive deeper into LinOS. And finally, our new CFO, Rob LeMaster, who will join us on the stage in a second, will summarize upside opportunities and provide financial impact details of LinOS. With that, let me turn the stage over to Greg.
Thanks, Ki Bin, and thanks, everybody, for joining us today. We're excited to talk about Lineage a little bit and dive deeper into our warehouse productivity journey and LinOS, and so I'll start just with a quick company overview for any of you that aren't familiar with us, so after 17 years of hard work, we've built Lineage into the largest company in our industry. Our global network is comprised of 3.1 billion cubic feet of warehousing capacity and generated about $1.3 billion in EBITDA last year. With a global footprint, we have 488 facilities comprising 86 million sq ft in 19 countries around the world. Our strategy focuses on operating modern and strategically located assets in key real estate markets and ports. We have an orientation toward the distribution side of the cold chain, and we have the largest network of 82 automated warehouses around the world.
We are a values-driven organization with over 26,000 team members around the world who are committed to fulfilling our purpose. Importantly, and especially for today's discussion, our labor expense is 60% of our total operating expense. While we've made substantial progress increasing productivity in recent years, thanks to this guy, we still view incremental opportunities on the labor side as one of our biggest controllable levers moving forward. Let's move to the next slide and talk briefly about our purpose and values. While transforming the food supply chain to eliminate waste and helping to feed the world might sound a little lofty, we really do play a crucial and central role in the global food chain. For example, we estimate that we transport or store about 30% of the temperature-controlled food and therefore calories consumed in all the United States.
We're also a people and values-driven company where our core values drive our behavior and our decisions on a daily basis. So moving to slide six, let's talk about supply and demand for a minute, because I get more questions on this topic from investors than any other one by a long shot. And we discussed this on the last earnings call. But we understand that we're in a niche industry, and it's a specialized part of the real estate landscape with very limited publicly available data. And so as such, as the industry leader we've been working with CBRE to provide insights to investors on new supply growth in our industry. At this point, we focused on the U.S. market, where we have the most accessible data and where we're seeing the most acute supply-demand imbalances in some markets. So let me just walk through the slide quickly.
The upper left-hand chart labeled A shows that from 2021 to 2025, public refrigerated warehouse supply grew by approximately 14.5% on a square footage basis, and importantly, if you look at CBRE's outlook for new capacity next year in 2026, that's down substantially to 1.5%. From a competitive standpoint, all of the capacity that's added in this period was built at the highest cost-to-build in our industry's history, and since that period, the cost of capital has increased significantly. The upper right-hand chart labeled B is based on the widely used Nielsen data, that's the point-of-sale data for retail, and the Circana data, which is like the gold standard for food service, showing fresh and frozen food volumes. Importantly, the data shows demand for food categories stored in our warehouses grew cumulatively by 5% during the 2021 through 2025 time period.
And this is in spite of consumer price inflation on the food side, which has been super high, interest rates, other headwinds. And despite all that, end consumer demand in our business continues to grow. And if you look at kind of the tail of the tape of the last few years, this growth was clearly masked over the last two to three years because our customers overbuilt inventories during COVID and then subsequently destocked that inventory right as we were leading up to our IPO. And so while our customers' inventory holdings have seen truly unprecedented volatility in the last couple of years because of COVID, the underlying demand for the goods that flow through our warehouses has been growing at a steady pace.
So at the bottom left-hand side of the slide, we just bring these two concepts together to calculate the estimated excess capacity, about 9.5%, in the U.S. market over the last four years. But despite this nearly 10% imbalance, our 2025, and that's what we got into last quarter, 2025 total estimated average physical occupancy is 75%, and we think we'll achieve that. We'll see how December comes out, but we're on track to achieve that, is down just three points from 2021. And so nobody likes fighting 15% new capacity or the resulting 300 basis point decline in occupancy. But given the backdrop that we've been fighting through, we're proud of this result in a very, very tough environment. And certainly, as evidenced by what the team will share today, we're not standing still in the face of these headwinds.
We're laser-focused on improving our go-to-market approach, streamlining our operations, lowering our admin costs, and carefully managing our balance sheet. So a few more thoughts on supply and demand. Regarding supply, it's important to point out that the new supply is concentrated in select U.S. markets. While many markets in the U.S. are still at equilibrium and our European and Asia-Pacific business are not experiencing the same type of new supply pressure, and diversification is just one of the scale advantages that we enjoy versus our smaller competitors that may have more concentration in these markets that have had the new supply delivered. We also believe that there's a number of factors that will tighten this dynamic over time.
On the supply side, we believe new deliveries, as evidenced by the CBRE data, have peaked, and we're hearing that very limited interest in new developments, which we think is rational given the supply and demand imbalance in parts of the U.S. Regarding the existing supply absorption from the supply that's been delivered recently, major players like ourselves are idling buildings and exploring alternative uses like data centers and other energy-intensive alternatives like Waymo or electric vehicle charging, which we're uniquely positioned to supply those types of services because of the amount of power we have being delivered to our buildings. We're also seeing in this environment that older, less efficient buildings are becoming obsolete and are being repurposed for higher and better uses.
Finally, given the complexity of operating cold storage warehouses and the relatively high cost of entry, we believe a number of our new competitors will exit the market in the short to medium term, and we believe we're best positioned to absorb some of this capacity, the most attractive pieces of this capacity over time. On the demand side, there's no doubt in our minds that demand has been muted for all the reasons I stated before, and there are a number of potential catalysts that could accelerate demand, including tariff resolution, interest rate declines, food inflation moderating, and customers trading into frozen categories given the attractive value proposition of the frozen food category.
And so while we're certainly facing a challenging market in the near term, we're controlling the controllables, we're best positioned to win in the long term in our industry, and we believe there's a number of catalysts that will bring the market closer to equilibrium in the medium term. So moving to slide seven. So acknowledging what I just said about the market's macroeconomic dynamics, one of the things that gives me confidence is if you just look at this map, you'll see Lineage has a presence in all the key markets around the world where our customers need us most. This is very, very hard to replace, very strategically located, very modern real estate and buildings, and that gives us the largest and the best position, as well as the most diversified real estate portfolio in our space.
So in addition to that best-in-class real estate portfolio, we also provide our 13,000 customers with the broadest suite of supply chain solutions with our Global Integrated Solutions segment. And so if you look at our customer spend, 75% of their supply chain spends on the transportation side and 25% on the warehousing side. And we're the only one that can provide that farm-to-fork solution with the broad set of Global Integrated Solutions that we provide to help them impact their total landed costs and also get their product reliably to their customers. And lastly, on this slide, I could talk all day long about our lean operational excellence, but I'd be stealing Jeff's thunder, and he'll be double-clicking on that in just a minute here.
But I do want to say I think the way we approach that element of our business is differentiated and that our process for continuous improvement is excellent. So before I turn it over to Jeff, I'll just spend a minute outlining the value creation levers that we're working on as a management team to ensure that we deliver shareholder value for our investors. So a number of us will speak on each of these, but I just want to spend a minute outlining the four key drivers as a group. We see a lot of potential in the way we go to market and customer excellence, capital allocation, and network effects in addition to the warehouse productivity topics that we'll talk about today.
And as you've heard us say many times before, what underpins all those value levers is our laser focus on digital enablement that supercharges the impact of these levers over time and puts us in a differentiated position. So let's dig into it. And as I said, we'll double-click on the warehouse productivity today, and I'll hand it over to Jeff here, but we'll come back at the end to answer your questions and have a discussion after you learn more from the team here. So Jeff.
Great. Thanks, Greg. So we're going to start off really digging into the people and process part of our journey. And this is the foundation to enable us to realize the benefits of technology that's coming. So the focus we've had the last few years has delivered significant improvements in safety and productivity and quality and customer service. And we'll walk you through some of the key steps that we've taken here. All right. On the people side, it all starts with engaging our team on our purpose, as Greg started with earlier. Something that our team can emotionally attach to and be intrinsically motivated by. And this creates pride in their mission to help feed the world. We have multiple reinforcement mechanisms in place to be able to catch our team members living the values and to recognize them for that.
In 2024, we took another significant step by creating a stock ownership program for a significant percentage of our global workforce. So this has led to an act-like-and-owner mindset, and we've established quarterly ownership meetings where we provide an update to all of our team members on how Lineage is performing and what their individual role is in helping to drive continuous improvement and success. Next, we focused on talent development with an intentional approach on hiring to our values and competencies. And the focus on culture and talent, as you can see on the right there, has resulted in over a 50% reduction in turnover over the last four years. Moving over to the process side, we've built on the foundation of culture with our focus on leveraging a lean operating system to drive continuous improvement.
Lineage has been driving a lean operating system now for almost 10 years. Very similar to the Toyota Production System or to the Danaher Business System, we've structured a methodology to teach our team members and ensure we're driving a consistent approach around the globe. So a consistent approach, not only to drive continuous improvement, but a consistent look and feel for our customers. Given that we've grown through acquisitions, having this approach is critical to drive consistency and to engage our team members around the world. To date, we've had over 80 sites that are on the formal lean journey that are going through different certification levels of lean, and we have over 9,000 team members that are engaged in helping us solve problems driving our lean methodologies. And our lean journey starts with standardized work.
So we've created the Lineage operations playbook with over 400 standardized operating procedures that is used by our facilities around the world to drive continuous improvement and then also to audit to identify where there's gaps to be able to drive overall performance improvement opportunities. We have also created continuous improvement roadmaps in over 450 of our buildings. So a continuous improvement roadmap is a detailed plan that has smart targets, actions, owners, and dates for the team to be able to realize improvements throughout the year to be able to deliver financial performance. In the middle column there underneath visual management, we've deployed a real-time metrics platform called Metrics One. Refreshes every 15 seconds, and I can look on my phone or laptop and have visibility to 100 metrics for each facility around the globe.
We have a view of looking at the site level, region level, country level, business unit level, or a global view to see how we're performing. We have a saying that red is good or problems are good, and that we want to identify problems quickly so that we can attack those problems and improve overall performance for our customers. Lastly, on the problem-solving piece, we have over 9,000 team members that have been trained on how to solve problems on the floor. We've conducted almost 1,000 Kaizen events that have realized over $20 million in savings by engaging our team members to help us identify and eliminate waste. What does this mean to productivity? Our focus on people and process have resulted in significant impact over the last eight quarters.
As you can see on the left here, the dotted line shows industry wage inflation versus the dark blue line showing our actual labor costs staying flat over the last eight quarters. On the right, a similar view of total labor costs per throughput per pallet also shows our ability to offset wage inflation again through our focus on people and process. And with this foundation, it's setting us up for the next horizon of benefits as we commence on our tech journey. And so Sudarshan is going to start walking us through where we're going next.
Yeah. Thank you, Jeff. Much appreciated. I really want to kind of take this opportunity to kind of dig a little bit into both our technology journey and our digitization enablement journey, as well as LinOS. And we'll talk about that between me and Elliot a little bit in detail. Our technology enablement, our digital enablement really started over the last decade, right? We've invested significantly in our systems and digitization. It allows us to unify data, modernize our systems in pursuit of kind of lowering our transaction costs across both our warehouses and also our back offices, so it's our kind of backbone for execution. Those efforts have created a digital foundation that we now operate on, and they support really five pillars in our digital transformation model, which I'll walk through in the next slide.
Here, what I really want to kind of talk about in these five pillars is the first is our scalable digital core, right? What does it actually mean, right? Our unified data and integration layer that gives us consistency across the network that also allows us to deploy these capabilities globally. Second is our decision algorithms. Elliot's going to talk about in detail about these algorithms in a minute here. But these engines make real-time choices around labor dispatching, pallet routing, and dock allocation, right? They reduce idle time and help us get more throughput from the labor that we already have. This forms the core of LinOS, and we're going to be talking about this in the subsequent slides. Third is customer experience.
Both Greg and Jeff alluded to how it's important for our customers to have a unified experience given our network and how varied and wide it is. Lineage Link, which is another digital enabled platform, allows our customers to get real-time facility, inventory, orders, appointments, and shipments. And this platform reduces friction and makes us easier to do business with. Fourth is energy. Energy is another large expense for Lineage after labor. We're an industrial consumer of power. We've been deploying AI-driven controls that help optimize refrigeration and power usage, lowering energy consumption and operating costs across our network. Think about energy kind of costs like a thermostat in your home, like if you had a smart thermostat at your home. And basically, our industrial platform is exactly like that, but for a giant industrial freezer.
Finally, the future of automation is really, from a LinOS perspective, is our warehouse execution layer. We're going to be talking about this in great amount of details. This is the software layer that orchestrates people, equipment, and automation in a single consistent way. Together, these pillars help us run a more productive and predictable and resilient network as these technologies continue to roll across our network. All right, LinOS. LinOS is the execution layer that brings our investments together, right? It is a warehouse decision engine that has been built over many years. LinOS really makes thousands of decisions each day, assigning work and routing pallets and sequencing all of the tasks in real time. These are practical operational decisions that have a measurable impact on productivity.
Somehow, these algorithms work together in practice and kind of make, because of the spatial awareness, it makes execution really, really simple within the four walls of the warehouse. Let's talk about how these algorithms work together. LinOS is built around a set of algorithms that each manage specific part of warehouse execution. They assign trucks to the right dock doors, route pallets to keep them flowing, choose optimal travel path, schedule replenishments, and dispatch operators. These decisions happen continuously and at speed and consistency that's difficult to match manually. Not entirely a novel concept, but the impact is straightforward. Less forklift travel, higher units per hour, faster truck turn times, and steadier labor utilization. The decision automation is what is driving productivity improvements that we are seeing at LinOS sites today.
So I get often asked this question, what is the difference between a warehouse management system and LinOS or warehouse execution system? If you don't live in it, it's very hard to imagine. But in a traditional WMS or a warehouse management system platform, it's primarily an inventory and a transaction system, right? Think about that as a system of record, which also enforces workflows and a rules-based decision for interleaving, right? It's highly deterministic. It still relies merely on human planners to decide the sequence of work, and it's kind of unidirectional most of the time. LinOS is different. It sits above the WMS or the warehouse management system as a real-time decision engine. It uses live data combined with spatial awareness to inventory, labor, dock optimization, dock conditions, equipment availability, and continuously optimizes micro decisions like task assignments, replenishments, sequencing, routing, and trip maximization.
LinOS is not just a feature within the WMS, right? It's a separate execution layer, as you will see, which is built specifically for high-velocity kind of warehousing and distribution center operations, and then if we go to the next slide, thanks, Alex. LinOS orchestrates both humans and robotics, right? LinOS gives us a very clear picture of what's happening in the warehouse, inside the warehouse, and where the warehouse operators are, what tasks are they performing, and what they're working on, and on what is the most efficient route to actually take, and it kind of helps reduce travel time. The situational awareness, when combined with real-time tasking, which drives productivity, right? Think about this as a kind of difference between a radio taxi and Uber, right? That is the difference. We built LinOS to be software first.
We're creating value now while building an execution platform that can adapt as automation technology continues to mature. I'm going to turn it over to Elliot. We'll actually walk you through how these decisions and algorithms are made in real time in our warehouse.
Super. Thank you, sir.
Thanks, Alex.
You go to the next slide. Okay. LinOS started with our near decade-long journey designing, building, and operating automated greenfields. Those greenfields use cranes instead of forklifts. They use layer pickers instead of case pickers, and they use conveyance instead of other forms of manual material handling. It's good to see some familiar faces who've been to the Olathe warehouse, for example. Now, we had to not only physically design and build those structures, build those robotic systems, but we also had to develop software that could control them. That software had to do three major things. First, it had to control and manage the dock in novel ways due to the increase in height of the building that automation facilitated. Second, it had to synchronize the activity between the humans and robots.
And then third, it had to, as Sudarshan indicated, automate the decisions made in the course of operating the warehouse. And so go to the next slide. What you're seeing here is a three-dimensional simulation of an automated warehouse. That simulation is not only physically accurate to model the performance of the robotic components, but it's also making the decisions that LinOS has to make in the course of executing automation. So among those decisions, I'm going to read them off clockwise from the right. Where does inventory land in the storage racks? Which pathway does inventory travel through the robotic layout? What schedule do we expose to which transportation carriers and which customers? In which sequence and orientation do we put pallets into trucks when we load them? Which door should each truck utilize? Which inventory do we allocate to a given order?
Which equipment is pulling which pallets and in which order? Each of those decisions very often has a solution space that's larger than the number of elementary particles in the solar system. Now, we make those decisions using a variety of methodologies. Those methodologies include convolutional and recurrent neural networks, include branch-and-bound combinatorial solvers, include convex optimization, and natural language processing. They also include other forms of mathematical optimization, machine learning, and artificial intelligence. Now, the objectives of these decisions are to maximize the efficiency of our humans and equipment subject to the needs of our customers, such as turn times and drop- dead times. Now, every one of the seven decisions that I just read out is identical to or has a direct analog to a decision made in a conventional warehouse, which is where LinOS Manual comes in.
Applying those software methodologies to our conventional warehouses, there's four phases of software rollout here. The first is high-reach operations, which is already operational and we'll detail today. That phase dispatches our highest skilled and the largest fraction of our labor. These are the operators that pull pallets into and out of storage racks, potentially cantilevered five or more feet and 40 feet in the air. Then they bring those pallets to and from the warehouse docks. Second phase is value-added services. Services such as blast freezing, stretch wrapping, placarding, date labeling, and other ancillaries that we perform in addition to storage and handling. LinOS will direct the performance of those services and record whenever they are performed, confirming that we build comprehensively for the services that we provide and thereby increasing our revenue. Third phase is case picking activity.
That's the physically taxing exercise of manually removing cases of product from pallets and using them to build mixed-SKU outbounds. Think bouncing around at Costco to go pick an order for your family, but that outbound order weighs more than one ton. Then the fourth phase is dock management. Movement of inventory between trailers and the dock, including receiving pallets. Loading pallets absent LinOS requires a particularly steep learning curve as operators have to eyeball weight and balance considerations, stop sequences, maximize capacity, and perform safety checks. All of that will be automated under LinOS. So to understand the current state pre-LinOS, I'm going to show you what it looks like. It starts with a whole lot of paper. WMS will print out orders, and then those orders will be placed in stacks. Those stacks will then be distributed to physical cubbies, which represent appointments in time.
Once that appointment is there and the truck has arrived, the order gets sent out to the warehouse floor, correlated, order checked. I'm going to introduce you to my colleague here, José Boloña. José here is a high-reach operator, one of our facilities in Chicagoland. He physically traveled to the supervisor station and picked up that stack of paper representing the order that he's about to go perform. Now, in addition to that administrative burden, the bigger constraint is that working stack of paper by stack of paper implies working on a single order at a time. An order is either inbound or it's outbound. It's not both. So as José will bring in a pallet, place it in a rack, and then he will drive back out again with nothing on his forks.
Now, the constraint of each team member working on one order at a time implies the dynamic here that you see on the left. Pallet goes in, and then the driver returns with empty forks. One round trip performed one task. Now, the principal thesis of LinOS Manual is to remove the constraint and dispatch the forklift so that it brings a pallet in on one inbound order and then utilizes the return trip to bring out a pallet for a different outbound order. One round trip, two pallet moves instead of one. Now, systems off the shelf can plan interleaving in advance, but they're not dynamic to changes in the warehouse. Trucks could come in not in the order that we expect. Customer drop- dead times could change, and the systems also often lack the spatial awareness, which Sudarshan mentioned.
Prior to LinOS, the baseline interleaving percentage in our facility was zero. But under LinOS, we are reliably interleaving approximately 40% of our moves. So start at zero, have a discontinuous improvement up to 40%. Now, achieving this level of interleaving requires not just LinOS within the four walls of the warehouse, but LinOS scheduling of truck appointments to bring inbound trucks to docks at the building at the same time as outbound trucks. It requires LinOS dock door selection to put the inbound truck as close as practicable to the outbound truck. But the net result is we're interleaving approximately 40% of our moves in the first sites. Now, we're looking here at an aerial view of a warehouse. The solid lines represent a utilized forklift, a forklift with a load performing a value-added task. The dotted lines represent switching distance.
You can see here the LinOS system moving around this particular team member. Now, the system is also computing the paths of the travel, which you can see in the animation. The mapping of this site and the paths is thanks to our use of LiDAR scanners, sensors that send out thousands of laser beams per second to generate a sub-millimeter accurate three-dimensional map of the warehouse. Then using that map, we computed the travel time and distance from every location in the warehouse to every other location in the warehouse. Now, in a production warehouse, there's not just one team member to manage, but multiple team members moving freight in and out. The multiplicity of those team members means that the LinOS system has a choice of which team member to assign to which pallet movement.
LinOS makes that choice by minimizing the total travel time and particularly that dotted line switching distance. So from this visualization, you can see the system functioning across all three of its DCs. One, it's fully digital dispatch, dynamic to the changing warehouse conditions. Two, it's interleaving inbound movements with outbound movements to utilize both ends of the operator round trip. And then three, it's choosing which operator makes which move in order to minimize that switching distance. The net result of all these DCs together are a frankly astounding increase in the productivity on our high-reach function. So in this particular warehouse, one of our initial sites, we got a 50% increase approximately in that forklift, in that high-reach UPH. Now, there's a huge step function increase here represented by LinOS being switched on.
But also, since LinOS was switched on, you see a slow drift upwards, a continuous improvement. And that continuous improvement has come and will continue to come via four additional dynamics. So first, our operators are getting better familiarity with the system. They're getting much more used to it. Second, we're doing a better job of scheduling trucks so that inbound trucks are coming in approximately the same time as outbound trucks. Then third, we're getting better at staffing the site to better match the new and decreased labor demands thanks to the efficiency of LinOS. And then fourth, we're tuning and improving these algorithms constantly. So to give you the after, let's meet Mauro Pulido, who is a high-reach operator at a different site that's now operational under LinOS. So he picked up a pallet per the instruction of the system.
He's driving in, scans the rack for confirmation that the pallet's landing at the right place. He puts it in. You can see the skill of the high-reach operators here. Then instantly gets another task from the system. He's supposed to go one aisle over in the same room of the warehouse, pick up an outbound pallet, and it's going to send him back to the dock. So this is his return trip, which is now utilized as Mauro returns to the dock. He puts away that pallet in the dock lane, and then the system immediately gives him another task, which is to go pick up the next inbound pallet that's proximate to where he just dropped off. Picks it up, system gives him the location.
He drives in, scans the rack for confirmation, puts the pallet away, again, cantilevered way high up in the air, then wash, rinse, and repeat. He's got a next task ready to go.
You notice in the second video that Elliot walked through, there was no paper. There was no time going to find a supervisor. It was all directed work on the handheld.
Go for it.
Thanks, Elliot.
Yep. We'll turn it over to Jeff Rivera, our COO.
All right, so it's early, but we have piloted at 11 sites so far this year, and in the middle bucket there, you'll see HRO productivity, so HRO stands for High-Reach Operator. That's Mauro, who's on the truck in the video, so for his department across these locations, we've seen on average a 30% lift in high-reach productivity, and then when looking at total labor cost for throughput pallet across these buildings, we're seeing overall cost reduction of 5%.
Including indirect.
Yeah, that's including that salaried, indirect, and direct total labor, so we've learned a lot, and with any major transformation with new technology, we've experienced some starts and stops. It's major change management. It's difficult, and delivering innovation is never easy. What we have found is that at our smaller buildings, being able to work through the change management and be able to get our salaried and hourly workforce trained is pretty quick, and in our larger buildings, our larger distribution centers, it has been more challenging, more change management, harder to get to third shift on a Saturday and Sunday night, and because of that, it's taken longer, so we continue to learn and adjust our training accordingly. I'm going to walk through one example here, and I could talk about this all day long. It's pretty cool.
This actually is our first building that we turned on in Q, I think it was Q2 this year, a smaller building in the Chicagoland area. What was really great about this building is within two weeks, it's really surprised us that within two weeks, the building had really latched on and was performing well. On the upper left here, that's the high-reach operator productivity. Again, Mauro's department in this building saw a 35% improvement in units per hour. The overall site productivity, so total pallets per hour across the entire building, saw a 27% lift, which is higher than our overall average, but
it's 20. Yeah,
It's 25, sorry, 25% lift. Then you can see the trend on the right there that shows kind of the week-over-week pre-versus-post transition after LinOS and really that significant lift.
What was also neat about this building is that about six weeks after the transition, because it improved dock flow and improved reach truck productivity, we were able to add another customer into the building without adding any additional labor, and I heard feedback from the team that without having the tech, we would have really struggled with dock flow, plus the benefits of productivity. In addition, in this building, when I was standing on the floor observing, we had a very senior tenured team member drive by in a reach truck and unsolicited yelled out, "This system is freaking awesome," and for me to hear positive feedback is rare, so to hear that from one of our senior guys was great, and as you saw a little bit in the video, it's just a better user experience for our drivers. They don't have to get off the reach.
They just want to do their job and not have to get off and on equipment and deal with paperwork so having an easy-to-use handheld and user experience has been much better for them overall and then we've also found that it's helped a lot with new employee onboarding, that they get up to speed on the system much faster. Dad, I'll hand it over to you, Rob.
Great. Thanks, Jeff.
Job done. Thanks, Jeff.
Great. So Greg closed his section with this slide to show the value levers that we have in hand to combat some of the near-term challenges we've been seeing over the past couple of years. Today's presentation spent more time on productivity and the digital enablement tools, which frankly, as Greg mentioned, underlie all these value drivers. But Greg also mentioned that we have other initiatives in place to ultimately create value. And I just want to spend a minute on a couple of those other ones. Customer excellence, you see there, capital allocation, and network effects. Now, we expect to do deep dives like we're doing here on each one of these in time and also give out KPIs and targets that ultimately allow you to map what we're seeing inside, and so give us time to do that.
But ultimately, we expect to come back and be held accountable for all these levers. But just to spend kind of a brief moment on the ones that we didn't talk about earlier with some of the speakers, the first one would be customer excellence. This starts with providing the best service in the industry. We distinguish ourselves with our customers with responsiveness and adaptability and really try to react to what our customers' needs are. Whether they're looking for full automation solutions with dedicated sites or integrated offerings with our transportation, we really try to serve them where they need us most. And their reward to us is their loyalty. We actually have about an over 30-year tenure with our customers. So it really proves that we have long-term relationships with our customers.
The second slice or lever that you can see there in terms of value drivers is around capital allocation. So that's all about compounding free cash flow and trying to reinvest that properly. I think most people know that presently we're very focused on building out some large greenfields that we have for large customers that have agreed to take most, if not all, of the space in those greenfields. So that's really where our allocation of capital is focused. But we'll continue to look for site expansions in tuck-ins and M&A as we've done in the past. And I did add balance sheet capacity because at certain times like now, one of the most compelling strategies can simply be to stockpile cash and wait for opportunities to get richer for us. Thirdly, you can see network effects.
We believe our unique scale can be a powerful force in many of the markets where we operate. We can offer customers the opportunity to grow and contract within a market, given that we have multiple locations or even across multiple geographies. We can grow with our global footprint. Our scale also allows us to idle capacity within a market, a unique advantage versus the smaller peers that might only have one or two sites that can't be shuttered or idled in a location, and then finally, in this slice, we think we can get some substantial leverage in our admin and our procurement spending areas as we mature and ultimately come more and more under one banner.
And then the final area of this slide that really we spent the bulk of the day on, just to reiterate some of the points some of the other speakers talked about today, hopefully you took away, is that this is all about getting real-time data and using it to drive lean operations. We also see LinOS and automation as tools to make better decisions and improve operational responsiveness and drive efficiency. So here, we're trying to display some of the internal KPIs we're watching to see how our initial rollouts are going. We've talked about a couple of these in various speakers' presentations. The four key areas of KPIs that really reveal to me the significant impact we're seeing in our operations and our profits include first, HRO, High-Reach Productivity. This is the activity, again, of team members moving product up and down the aisles.
There can be significant wasted time that we found in terms of operators only storing products or only retrieving products. And hopefully, you're convinced today that LinOS is really changing all of that. We're going to see in a couple of slides that HRO is our most significant cost of direct labor. And in early test cases, we're seeing units per hour, which is how we track this one, up 20% to 30%. Second, in terms of site productivity, we track pallets and volume that comes in and out of the warehouse, and we track the time and cost for every movement. And we see a very positive economic impact, but also our customer response times are really improving dramatically. And this is widening our lead as the industry's front runner in customer responsiveness. We track this one with PPH. And so Jeff gave a couple of those.
We're seeing a double-digit improvement across the pilots in a pallets per hour basis. And then thirdly, direct labor costs. We're finding that warehouse managers are finding ways to reduce shift labor and using employee attrition to decrease direct labor at some of those pilot sites over time. And then finally, total labor savings, which includes the indirect labor areas such as maintenance and site management. And what we're seeing is that with better data and asset utilization, the deployment of technology is causing some great gains there too. So all of this really rolls up on the layer at the very bottom in green, which is labor per throughput pallet. This is the statistic that we encourage you all to really track us on over time. You can do this calculation from an outside basis, and we're going to see that statistic continually be driven down by this team.
Okay. So the base case estimate that we see in terms of the savings playing out for us and the annualized EBITDA impact is $110 million, I put in the lower right-hand corner. Let me build you to the elements of that. The possible enhanced profit is really going to come from two different elements, a cost element and a revenue element. On the element of labor savings, first is HRO. As we mentioned a couple of times, this is our largest bucket of cost. That's 25% of our direct labor. And we see the ability to potentially reduce savings or to gain savings about 20% to 25% over time. In the case pick area, which is 10% of our labor costs, we see 5% to 15% savings. Dock, both the inbound side and the outbound side, represents about 20% of our costs.
And likewise, here we see 5% to 15% savings as possible. Then finally, a large bucket of indirect costs, which includes the maintenance and other site management costs. This is large at 45%. We see a slight improvement in several different categories. Ultimately, we believe that category will have a cost savings of about 5%. That's the cost elements. On the revenue side, the elements of revenue enhancement is really around what this technology will allow us to do in terms of tracking activities that we perform on behalf of the customer and make sure that we're properly billing for those services rendered. There's some 300 billing codes for the accessorial services that we offer within the warehouse. So our technology can catch errors or underbilling for these helpful services that don't always make it onto customer bills.
So we see about a 5% uplift in that part of the equation. So again, that all sums to a base case estimate of $110 million of annualized EBITDA in the next three to five years.
I mentioned that's 250 buildings.
That's right. If you go back to that slide, I think I mentioned on the bottom, that's going to be where we see the most material opportunity to be in about 250 of our sites where we're going to see a rollout have a direct impact.
So there would be upside in the incremental sites. Certainly .
Okay. To put that savings in context and try to give you a sense for what that costed to achieve that great outcome, let me spend a minute on the ROIC of this project. So we invested about $250 million over the past decade to build out all the LinOS initiatives that we've been talking about. And we expect to complete that project with an incremental $200 million investment over the next five years. Now, to be clear, as I noted on the right-hand side of this chart, much of that initial investment went to standing up the automation capabilities for the dedicated automation sites. But we've leveraged that technology and expanded the flexibility to allow for conventional site deployment. And so it's those gains on the conventional site areas that form the basis of our business case.
And we've excluded the full site automation, the full automation sites and the benefits that we see. We've sort of put that off to the side.
Yeah. I think that's important to point out that $200 million investment fueled the development of LinOS and all of our automated sites to where we've deployed it. And we announced the largest cold storage deal in history with Tyson earlier this year. And we wouldn't have won that without LinOS. And none of those benefits are included in this ROIC.
Yeah. Good point, Greg.
So through that lens, again, we envision a base case target of $110 million of EBITDA uplift at our conventional sites. And that maps to a 24% ROIC as you see here. Now, we hope we've convinced you that that underwritten base case is readily achievable given the early KPI readouts that we've been sharing with you today at the pilot sites. And we'll share in a couple of slides that we actually see even extra layers outside of site labor that provide what we call a halo and might ultimately drive upside of that $110 million estimate. Now, here's a view of how we see our past and future rollout of these key technology features. As Elliot and Sudarshan talked in their slides, we first had to develop the proprietary software products as we prepared for a full automation offering.
You can see that on the top there. Next, we decided that we could yield much of those automation benefits in sites by deploying those same technologies into digitizing our conventional warehouses, so we're now in that lower box where we're rolling out those features in a calculated manner across our conventional warehouses. We'll first, as we're doing now, deploy HRO, High-Reach Operations modules, then on to value-added activities, then to case pick, and then across the dock, and over the next three to five years, we plan to finish the product development on all those areas and deploy it across, again, as Greg said, the 250 warehouses that are likely to benefit most from the digitization, and to be clear, we are already piloting the HRO features in those 11 sites.
We'll be judiciously moving out on other phases and with additional sites in a measured fashion to ensure that we walk before we run as we have found that addressing inevitable challenges, as Jeff talked about in the early days of a rollout, will ultimately allow us to go faster and have a cleaner ramp in the future. So we're right in the midst of trying to assess that timeline. We plan to regularly report out on how each phase is going over the next few quarters and years as we increase conviction in this timeline, but wanted to give you a snapshot at this point. Okay. As I referenced earlier, while we're certainly focused on that underwritten side and on day-to-day execution in the LinOS strategy, we do see several knock-on call options that will impact us potentially positively in future years.
These outer layers are not part of the business case, again, that we presented a couple of slides ago, but could lend themselves to meaningful extra upside in the medium term. Some examples in that non-blue halo layer outside the core underwritten layer that I'm most excited about is the reduced support costs around hiring and training and even admin and technology costs. I'm really excited to see what we can do with lower maintenance expenses, reduced CapEx of forklift equipment at the warehouse, and better energy efficiency as we learn how to run our facilities more optimally. And then even further out, I would say, on the long-term halo and trying to talk you through what's exciting there, we really see opportunities to gain market share from the extra services and innovation we can bring to our customers.
We see the opportunity to have more take-up of our integrated service or transportation offerings around our GIS, which is our integrated solutions business segment, and then finally, we even see on the lower right a unique synergy in future M&A processes. Here, we can simply bring the technology we have uniquely been able to build, given our scale and investment capabilities, and apply it to more paper-based, smaller tuck-in opportunities.
I think one other thing to point out on that slide is we're right now focused on these 11 sites and learning everything that we can learn through this major transformation. This is a wholescale change with how we run our buildings. And that's where we'll be focused on for the next couple of years. But I grew up in ambient and dry warehousing for my first 15 years. And there's nothing about this technology that says that it couldn't be applied to every other warehouse in the world long-term. And so we're focused on the front, what's going to impact Lineage today, right now, but this has there's definitely an opportunity to expand beyond cold storage in the future.
Great. In summary, while in early days of implementation, a significant upside awaits as we scale these investments across our portfolio. I think despite challenging times, we do have levers to support growth and deepen our customer relationships. We'll focus on warehouse productivity through people, process, and technology, as Jeff described. Our scale and digital infrastructure is a distinct competitive advantage. We are the clear automation leader. We'll use those same proprietary technologies to further streamline operations, lower our long-term cost structure, and serve our customers even better than we are today at our conventional sites. Finally, this is only one aspect of our competitive differentiation. We continue to invest in many areas that have proven to make us successful in the past.
And then just on a personal note, I joined Lineage not too long ago because I believe that challenging times pose an opportunity for great teams with clear competitive advantages to grow even stronger as we are able to focus on what we're good at, double down in focused areas, work as a united team, and make the hard decisions together. I believe we have the ability to turn this challenging supply and demand dynamic into a catalyst to accelerate our maturity as a very recently stood-up public company and drive an even further lead over our peers. And so I expect you to hear a lot more and more about that message about methodical execution as we turn the page in 2025 and we look toward 2026. And we'll lay out those plans for you guys as well as the years ahead.
And so with that, I want to thank you for your time. This really concludes the formal remarks. And I'll hand it back to Kevin.
Thanks, Rob.