CGRO Collective Growth Corporation

We won the BMW program with Innovizone, our solid-state LiDAR, which is a very high-end LiDAR and cost effective. We froze that design 3.5 years ago. Since then, we made a lot of progress in our technology, to bring now, InnovizTwo, which provides a 70% cost reduction. This enables adopting a LiDAR, a high end LiDAR, also in L2+ platforms. L2+ price point is much lower than L3, and in order to be able to absorb such a technology, it needs to be very cost effective. Now, with InnovizTwo, we believe that we will be able to penetrate the L2+ and allow carmakers to launch earlier with a much more simple platform, which allows the car to drive itself, yet the driver holds the wheel and pays attention. That allows the car maker to collect data and validate through crowdsourcing – testing their software in real time, in real life conditions in different countries. That not only saves the cost of the LiDAR for the platform, but also the cost of validation, which is very high. We see ourselves penetrating through that into the market. Being an incumbent in the L2+, when L3 occurs in several years, we will already be part of many programs.

In order to be a strong player in automotive, and also specifically in LiDAR, it is not enough to have very high performance or low pricing point, it is also required to have a Tier 1 partner. A carmaker like BMW would never source a technology directly through a Tier 2 because for the sense of sensitivity in the supply chain, they are required to have a Tier 1 with knowledge and experience in industrialization or functional safety components. In this case, we have four Tier 1s that allows us to compete on different programs. Sometimes we compete on the same program with different Tier 1s. BMW is helping us as a technical leader to have our LiDAR an automotive grade, while other carmakers see BMW as setting the direction in the market. The perception layer is also a very important piece of the puzzle, as carmakers want to have a seamless integration of a sensor in their platform and they require the software layer that translates the raw data into meaningful information, and that’s part of what we provide.

Our technology was built from the ground up in order to bring the best technology at the right price point and the right reliability. For that, we developed three different technologies altogether combined with a lot of optics and software. We designed our own MEMs, detector, and a very complex mixed signal ASIC, which are all available today in this device. On the right, you can see the perception layer – the bounding boxes that provides classification of cars, trucks, pedestrians, motorcycles. We provide shape, orientation, and occlusion status, absolute and relative velocity, and acceleration.

I’ll show you a few videos. All of these videos are available on our YouTube channel, so possibly they will jitter because of the bandwidth of the internet, but as you can see the LiDAR has very high resolution. This video was taken at 15 FPS. I always advise investors to see a LiDAR in real life while working because some LiDAR companies like to show images, but it is very important to see the LiDAR at the right working point where it’s going to be used in the car. This video shows our perception software. This is real-time processing of the 3D information that allows detection and classification of the different vehicles. This is probably the most mature software available for solid-state LiDAR.

You can see here a video taken from a LiDAR that was mounted in the grille of the car. It looks as if it comes from a drone, but that’s only because the data is 3D. It allows us to change the pointing view of the sensor, but the sensor is just mounted on the grille. You can see how easy it is to see the different objects, seeing the different people, pedestrians, and people on bicycles or motorcycles. This is the most advanced LiDAR today available and it’s going to be in mass market next year.

The computer vision is a growing asset. We are collecting millions of objects in order to strengthen algorithms using our own database and our partners - Tier1 partners. Our IP is built in a structure layer. Starting from a very unique concept that is patent granted, down to the different components that we design ourselves such as the MEMS, such as the ASIC, and detector. A lot of IPs are buried in the silicone up to the software layer of the computer vision. This is me holding the LiDAR and this is the entire product, so this includes all of the optics, the laser, the detectors and the processing power of the 3D. There’s no additional box, and as you can see it’s very, very small and could be embedded seamlessly in the car in different locations. It has 115 degrees over 25 degrees field of view, which provides a very good understanding of the scene. Automotive grade is really not just a buzz word. I would say that autonomous driving is magic, but in real life, most of the work is just making it very, very safe and the quality very high. It requires a lot of testing in different weather conditions, thermal shock, water sealing, mechanical shocks. This requires a lot of details to take care of in order to bring the LiDAR to the right level of quality when replacing the driver.

We are very honored to work with BMW and Magna that are helping us get there. Not only that the technology needs to be automotive grade, but also the company. We have been working the last three years to bring up all of the qualification that we need to meet – all of the certification, such as VDA, which goes to the way that we manage suppliers, Automotive SPICE, which goes to how we write the code.

This slide talks about the way we split the market. We’ve split it between L2 and L3, and the reason is price. The pricing barrier for L2 is to have a LiDAR which is below $500. in this space, our competition is very low-end LiDARS with very low resolution. InnovizTwo, which is targeting a very low-cost, provides a path not only to allow very good and safe L2+ platforms, but also a path for the carmaker to develop on top of it, the software that eventually reaches L3. Our competition in the L3 and L4 are products that are far from being mature, don’t meet the resolution, very big, and really most of our competition are start-ups that don’t have Tier 1 relationships, which allows them to compete on business.

On the manufacturing front, we have different partners and different business models. We work with Jabil today. We have a line in Germany that produces these LiDARs, which we provide to different partners.

Within the automotive space, our partnership with the Tier 1 works in the fact that we provide a business model based on components. We sell our unique components, which are the MEMS, the detector, and the ASIC, and the Tier 1 is responsible for the high-volume manufacturing. We are now ramping up a high-volume manufacturing with our partner to serve BMW and others.

Let me hand over to Oren to present our product portfolio. Thank you.

Thank you, Omer. Hi, I’m Oren Rosenzweig, Chief Business Officer and Co-Founder at Innoviz.

As you understood from Omer’s presentation, we’ve been very focused on the passenger car market since we started the company in 2016 and this is how we managed to strike four Tier 1 partnerships and this is how we won the first and largest L3 series production award for a LiDAR in passenger cars, and it’s really through the work of our global sales teams that are working very, very closely on a daily basis with the passenger car OEMs on their RFIs and RFQs, and the work of our product team which is getting guidance on LiDAR specifications to enable L2+ and L3 passenger cars from our Tier 1 partners, and of course from our lead customer, BMW. But it’s really not just about the product performance, it’s also about cost, especially in the passenger car market and this is true for our current product, InnovizOne, and also for the next generation products, InnovizTwo, and Innoviz SLR.

Besides the passenger car market, where we’ve got a leading position in, we also have very strong business traction in other industries. So, you can find InnovizOne on the shuttles and robotaxis of the leading companies that develop their L4 fleets. We have partnered with Shaanxi trucks, one of the leading truck companies in China, earlier this year, and we’re also working with other leaders of the trucking space, and our LiDARs are also used by many of the companies developing delivery robots and autonomous drones. We’re also working with a variety of heavy machinery companies for mining applications, construction, and agriculture.

The way we approached modeling revenue is a completely bottom-up approach. We look at a funnel of about 200 companies across all these industries that our team works with today, especially over 20 passenger car OEMs across U.S., Europe, China, and Japan. We’re in daily, close discussions about the LiDAR that they’d like to put in the cars that they’re developing. About 25 of the 200 accounts are in late stages of awarding series production contracts and the win rate that we model will depend on the stage of those accounts. We want to be able to beat our projections, so we took a very conservative approach to modeling volumes as well, so we use take rates of about 3% maximum. Therefore, what we see here is a very big upside, so for example, an award such as BMW at a standard 15% take rate at the peak year of the program, would represent more than $700M in booking and $220M in revenue in 2025, which would be about 38% of 2025 projections.

Another big upside we see here is around the software, so in our model we only included the one-time sale of perception software that we sell together with the LiDAR, but from more recent discussions with OEMs, they would also like to get updates to the software to support corner cases that they encounter after they roll out their vehicles, or when they want to support additional locations. This will be offered as a subscription model that, as I said, we have not included in our model yet.

I will hand it over to Eldar Cegla, CFO, to go through the financials.