Hello, welcome to today's webcast presentation with Terranet, who's presenting the Q2 report for 2023. With us presenting, we have the CEO, Magnus Andersson, and CTO, Nihat Küçük. If you have any questions for Magnus and Nihat, please use the form that is located to the right. With that said, I'll hand over the word to you guys. Please, go ahead.
Thank you very much. I'm very excited to be here with you today, and thank you very much for, for joining us, for this webcast. I also especially would like to, to, welcome some of our new shareholders as well. This is a very exciting day today. My name is Magnus Andersson. I'm the CEO of Terranet, and as mentioned, I also have my CTO with us, Nihat, today. Let's move on. Agenda today, first, I do would like to talk a little bit about Terranet, who we are and what we do. Just as an introduction, especially for some of the new listeners. We move on and give you an update on the financial for the second quarter this year.
Also, of course, give you some highlights in the business, what has happened, during the second quarter. Both in terms of business development, but then also, talk about product development and how we're doing there. Last but not least, also give you some insights on what is coming, what is the road ahead. This webcast will be available on Terranet's website, and on the YouTube channel afterwards. Let's get going.
We are born in a forward motion. We are taught to keep on moving. If you think about it, sometimes the most important thing is knowing how to stop. In the blink of an eye. As cities grow bigger, urban traffic becomes denser and more complicated. Each year, 1.3 million people die in traffic, More than 40% of fatalities occur in urban areas. We believe that a diverse set of sensors, combined with intelligent and precise data fusion, are crucial to achieve zero fatalities. When lives are at stake, every meter matters, every split second counts.
That was just a very short introduction, why we exist, and why we feel that Terranet really can add some value to the mobility sector. We heard about that it's a lot of people still dying in traffic every year, and that needs to reach zero. UN has set up goals, European Union as well, and we believe that we can help achieving that. We shape the future of urban road safety. What are we doing? The product that we're trying to develop here, it's BlincVision, and we believe that we can be the most fastest and most accurate anti-collision system for automotives, especially in urban environments. That's Terranet and what we do. In terms of the company, we are based in Lund, in South Sweden. A very good center for competence within vision, embedded and software.
Also we have an office in the automotive center of Germany, Stuttgart. We do believe, as I introduced, that we have a groundbreaking technology for ADAS and autonomous vehicles. Our target customers are OEMs, automotive manufacturers, and also for them, then Tier one suppliers. We are listed on Nasdaq First North Premier Growth Market, and we are continuously growing our team, and now we are around 15 people. Of course, we also work with strategic partners around the world, both on the business development side and also for supplying components and competence to our product. Let's move on and talk a bit about the financial for the second quarter. It was a very busy quarter in terms of financing.
issue, and even in the tough market climate, we were very happy that we actually reached 70% subscription from existing shareholders, but also from some new shareholders joining Terranet. Also, very happy about that members of the board also participated, and management as well, including myself. Our largest owner actually made of a capital invested above their share of ownership. Very good news. The rights issue ended up with SEK 52.8 million before issuing costs.
As part of this financing exercise in Q2, we did also refinance our loan to Formue Nord, a loan that we took earlier to finance our investment in holoride. Now in this refinancing exercise, we actually also amortized half of its loan, so that's also very good. If we move on and looking at the financial outcome, revenues, as you know, we are a company developing products, so they are quite small, mainly from Vinnova project. In terms of costs, they have increased compared to previous year, but that is also because we have increased the speed of development and starting actually to develop and produce our product. Also increasing actually our marketing costs to reach out to the market.
Therefore, consequently, that also affect the cash burn, so slightly higher than last year. In terms of the cash balance, by end of the quarter, it was around SEK 22 million, but actually all of the rights issue capital has not come in by end of the quarter. In the beginning of July, we actually had a cash balance of SEK 35 million. That's the financing part. If we then move into the key highlights of the business, we told you before that we received our prototype of a laser scanner, so we successfully tested that in our lab environment in quarter two. Together with other components, it showed very good results. We could demonstrate a range between five and 30 meter, which was what our objectives had been set.
This, of course, is very important step to actually complete the prototype that we later on wanna put on, on the vehicle. In parallel, we also make some good progress on the AI side, and also working with setting up a virtual simulation system. By doing simulations, we can speed up the product development, getting a product fast to the market, but also a product which is more 100%, so to speak. That was very good achievements in the product development. In terms of the business development side, we continue to participate in many conferences and events around the world, and with different purposes, I would say.
One, of course, to meet up with our customers, OEMs and Tier 1, but also to confirm that we are that competitive that we believe we are, that our product is really unique, and the value we can add is something that is really needed on the market. We attended several events in the quarter, and that was confirmed. There is still a great interest for BlincVision and Terranet. Some of the events we attended is also good. We were not only attending, but we actually were invited to speak in panels and on stage. We actually participated in a panel together with our customers in ICA Summit in Frankfurt. We were also here in Sweden, in Gothenburg, and there we actually had a booth.
We also spoke on an ADAS conference down in Stuttgart, end of June as well. It shows that there is really an interest for what we are doing. By end of the summer, also, we were over in Detroit, so we also are not only visible in Europe, but also in North America, and also were invited to participate in a panel. While in Europe, we had had some panels discussing a lot of ADAS, so driver assistance, but in North America, we also participate in a panel for autonomous vehicles. We see there is an interest for our product in many different areas.
Another thing I we should mention as well, which we also are very proud of, that we are part of a Business Sweden's Autotech program in this year. Meaning that together with us and, and some other, smaller companies, startups, in the automotive sector, we are approaching the industries together. That has several benefits for us also to, to talk to other startups, some who have entered the market further than us, and we can share learnings and successes with them, very useful, and contacts. Also being part of a Sweden government program, where they also introduce us to, to key stakeholders in the industry. We get a wider reach, but also meeting more of the right people.
For example, now, in these conferences, I had many discussions with Tier 1 suppliers like Continental, ZF, Magna, Aptiv, Valeo, Bosch, and Denso. They're all very interested, and yes, they see the anti-collision potential, but they also actually see other use cases for us, where our solution could be used. That's also very exciting. Of course, we also speak with OEMs. Since we are in Sweden, of course, we speak with Volvo, but also in Germany, with Mercedes-Benz, Audi, Volkswagen, and others. Yeah, we're speaking with many people in the industry, which is very important for us. Also, one key event I should mention as well that we participated with a different kind of aspect, was Vision Zero conference in Stockholm, NYU. Actually, the Minister of Infrastructure from Sweden was at the event.
Sweden is seen as in the forefront of saving lives in, in traffic, and we still are. There was a lot of international participants and a lot of regulators and other there. Very important for us to be visible, to, to work closely with the regulators to see how they're constricting some of the targets to ensure that products like Terranet are being used. Very useful. Last but not least, we did also launch a media campaign, a three-week media campaign during quarter two, to reach the wider tech and business segments. Very good visibility for us and a good reach. That was some of the business highlights for the second quarter. Very busy again, very good progress, both from the product development, but also the business development side. Let's now move on to the next agenda point, and Nihat will give you some more details on our product development.
Thank you, Magnus. Good morning, everybody. My name is Nihat. I'm responsible for product strategy and product development at Terranet. I'll give you a brief update on the status of our development. Our development activities are running according to plan without delays. We made good progress in all areas, bringing us steadily closer to our prototype. We're able to recruit additional engineers, which is good for our development to pick up more speed. On this slide, on the first slide, I'm going to show you an update on the scanning device, which is our emitting unit. After a custom-made scanner prototype had arrived in Lund last quarter, we started the device testing in a controlled lab environment.
We operated our laser in a safe environment to experimentally explore the best combination of parameters like the laser color, the wavelength, the laser power, pulse length, scanning patterns, et cetera. The setup included the two synchronized event cameras we have, capturing the scanner's sweeping patterns and forwarding the raw data to our processing unit. Since our scanner serves as a vital component in the entire system, we think that the testing phase marked a significant milestone towards our final prototype. The scanner has been able to demonstrate excellent performance in illuminating objects up to a range of 30 meters in front of a vehicle while maintaining, of course, international eye safety standards. The next slide, I'll describe the status of our processing unit.
After our stereo camera system has captured the reflected laser beams, along with a lot of noise, of course, the incoming data needs to be filtered due to ambient light. Getting rid of irrelevant data needs to be eliminated, trajectories need to be reconstructed, and finally, to get a 3D scene, a triangulated 3D scene. These main tasks define the most relevant requirements that we have, you know, for our processing unit. Far, we have not had the opportunity to test or optimize our algorithms for beams arriving from five fast-moving lasers in parallel. This is currently being executed in our lab under artificial lighting conditions. The processor's output is a fast, asynchronous data stream of 4D splines, which are basically a sequence of vertices connected by lines.
This is a totally different output than you know from cameras, lidar or radar systems. In this and next quarter, we still plan to heavily optimize those algorithms, and we also plan to port them on accelerated, dedicated hardware, for example, down to the level of FPGAs, in order to further increase the performance. We have to ensure that everything is processed in real-time, very quickly, with ultra-low latency. That data, which I'm going to show in the next slide, is forwarded to the recognizer. The recognizer is the unit that is going to understand what's happening in front of the vehicle. We continuously improve our event-driven object recognition software to ensure that BlincVision achieve our goals with the lowest possible latency. Therefore, we can't simply use off-the-shelf object perception algorithms because they are frame-based.
They're optimized for regular cameras or lidar systems. We started implementing our own so-called event-based algorithms, which means sub-frame resolution. Understanding and processing data in between frames, because our entire system is asynchronous and not frame-based, and therefore we can achieve a higher end-to-end latency than other systems. Training our AI software stack for object perception requires large amounts of virtual road users in various traffic scenarios. That's what everybody knows, right? You need to have good data when you train your system. Our engineers currently validate our perception stack under a vast range of different conditions using, of course, simulation platforms. We're trying different simulation platforms, work with them in parallel, and we'll continue this work until end of this year.
The primary goal is still to reduce the time required for the full object classification and tracking the entire cycle down to the order of 15-20 milliseconds. That's, that's our goal. Of course, before we take a test vehicle on the road, we want to analyze different critical driving situations, different Euro NCAP situations in that test environment, in the virtual environment. By using those professional end-to-end simulation platforms, we can both validate the performance of our system, also we will accelerate our hardware and software development in general. Yes, thanks for your attention. I'm heading back to Magnus.
Thank you very much, Nihat. Very exciting on our product development side. If we now look ahead at what's coming, what are we spending our time on today, tomorrow, and this coming quarter? As Nihat mentioned, we are focusing a lot on the AI capabilities. Yeah? In the simulation side, but also starting implementing this part of the product. Very interesting, very exciting, and a very important part to also keep up the speed on identifying and adding value to, to the traffic safety. Secondly, of course, we want pre-production or a prototype ready as soon as possible, okay? We're working up in the lab, and then, of course, the next step after that is to put it out on, on a vehicle.
We put a lot of focus on that and continue to do so in the next quarter. I mean, in the end, our customers want to come here and visit us, and even in the lab, to see our product. That's why this is also important. In terms of business development, we will continue our partnership with Visit Sweden. Actually now in September, we are again invited to speak at a conference in California. We will do that together with a delegation with Visit Sweden and including a road trip in California. That's very interesting for us. We also have other conferences we need to attend in Europe, of course, during September as well. Lastly, continue expanding the team.
Nihat stated that we added some excellent engineers to our team, and we continue to do that. Also, in terms of other functions, we also are strengthening up the team, so we're continuing that effort. That's our focus areas for the, for the coming quarter. Thank you very much for listening today. Stay up-to-date with us. Of course, look at our website for press releases and news updates. You can, of course, also send us an email on investor relations if you have any, any questions. Lastly, of course, we also try to be quite visible on social media, especially on LinkedIn, so please ensure you, you follow us there. Let's move on to the Q&A session and see if we have any, any questions from, from the audience.
Thank you very much, Magnus and Nihat, for that presentation, and now we'll jump, like I said, into the Q&A section. We'll start with the first question here: How long does the funds you receive from the rights issue last?
Yeah, thank you very much for that question. A very, very good question. Of course, depending on the different paths we take, moving forward, the choices in development and the speed we do this, that will affect how long the cash will last, right? Also, we have an outcome of warrant series in TO6 and TO7, in one in December and then in June next year. That, the outcome of those will also affect how long the cash will last from the rights issue.
Thank you. We'll take the next question here. Many competitors worldwide working on lidar and radar with their own object perception. What would you say is unique with your solution when it comes to object perception?
Yeah, Nihat, do you want to start?
Yeah, let me, let me take that one. Yes, our objective is to design an approach that reduces the processing latency as much as possible with maintaining highest performance possible. That means that we have to go also different ways when it comes to the object perception. For example, we apply a novel asynchronous event processing paradigms that generalize the standard, so-called graph neural networks process. We are processing events as they come in. We're not collecting them to frames and then process them. If we did that, we wouldn't have any advantage over the so-called frame-based approaches.
This is why BlincVision scanning and capturing totally differs from regular RGB cameras or time-of-flight concepts others use. We have to have those tailored data models and also our own neural networks and the different training. T his is different and needs to be different because the concept is different and will make us, you know, from terms of in terms of performance and also latency, top-notch.
Okay, thank you. And you mentioned in your presentation today that you wanted to develop your own AI. Does it make it more complicated, and could it possibly cause any delays?
Yeah, that's also a good question, which basically relates to the previous one somehow. We don't expect any delays there because we have those algorithms in place. The neural networks need to be trained, whether you use, whether you use conventional models or not, and our model needs to be trained in different way. Why? Because we see differently, right? The pictures, more or less, that we see is totally different from Lidar or from cameras. That means that is not additional work that will cause delays. More complicated, I would say we're using the latest algorithms that are developed, or at least papers that are developed by many different universities across the world, and we take those algorithms and derive algorithms, and finally, code and neural network models out of that.
Clearly not, it's not going to make things more complicated, or cause delays. We knew this from the beginning, and this is in plan since the beginning, so I can actually relieve people. No, it's not gonna be more complicated.
Thank you, Nihat, for that answer. We'll move on to the next question here. Autonomous cars in the future are autonomous cars in volumes are far in the future. How would you generate sales before that?
I mean, I, I can answer that. A good question, and of course, autonomous vehicles is discussed a lot in, in, in media and so on. We are not only focusing on autonomous cars, we are also focusing on cars where you have a driver driving a car, so advanced driver assistance system, okay? These cars are all on the road today, so there is already a need for our product on the road today. Actually, if it was on the road today, we probably would have very much less accidents, huh? Autonomous vehicles, yeah, there is a lot of things happening there. They are on the streets, especially in US, you can see it, but we also have many challenges, so it's gonna take some time. I think we can also help them solve these challenges.
Of course, there is an interest from autonomous vehicle OEMs from, for us as well. As I said, we can add value on cars on the street today, which have a driver. Actually, that's probably more of our focus area at the moment. Yeah, I think we will generate sales as soon as the product is ready.
Thank you, Magnus, for that answer, and we'll take the next question here. You said that your cash burn is, is increasing. Will we see the same pace going forward?
I would say so. Thank you. I mean, that's also another good question. I think, yes, of course, we are now speeding up development to get the product ready, so we will spend more money, both on product development, but also on marketing activities. Yes, it will continue to increase in the near future. Yes.
Okay, we'll take one last question here. What can investors expect of Terranet in the remainder of 2023?
Yeah, that's a good question. What I hope and what I'd like to promise is good progress, huh? Yes, we will continue to achieve a goal set up in our plan, both on product development and on the business development side. Soon, in the near future, we will see a prototype ready on in the lab, and then later on, out on a vehicle.
Okay. Thank you very much, Magnus and Nihad, for that presentation. Thanks to all of you who tuned in for Terranet's Q2 presentation. I hope you have a great rest of the day, and until next time, thank you, guys, and goodbye.
Thank you very much.