2021 Sept Demo Day - Sebastian Bauer

SEBASTIAN BAUER: Yes. Hello, everybody. I'm Sebastian, the CEO of Ubicept. We are a computer vision software company pre-seed with five full-time employees right now.

And we actually make use of each individual photon-- so each photon, that's the smallest unit of light-- and that allows us to see in challenging environments. And we call it Ubicept, ubiquitous perception.

Imaging fast motion, for example, is super challenging. The same also holds for scenes with low light and scenes with large brightness variation. And in fact, that is so challenging that usually there's a dedicated camera out there for each of these application scenarios, but that means you have to have these cameras.

And for example, it's really hard to image fast motion in low light. If you have a low light camera, then brightness variations are a challenge because in bright light when you have headlights, for example, your camera saturates really quick. And if you have a high dynamic range camera that copes with brightness variations, it's really challenging to deal with fast motion because they get motion blur.

We have actually talked to over 100 customers in different verticals and got that confirmation that these problems are indeed customer pain points. And our idea, really, is to provide one single system that can solve all of these problems and generate higher efficiency, more safety, and even can unlock new application scenarios that have been unheard of before.

This is our secret sauce. Each white dot here is actually a single detected photon from a single photon sensor. And that process happens thousands of times every second, and we, again, make use of these photons and derive the whole story from that.

And this is an application scenario here that's a very dark office building where the lights are off. Pretty much the only light sources are at exit signs. We have a single photon sensor here with our software running on that. This is for comparison-- a low light camera that costs about $800 and claims to be sensitive down to six millilux. Also, thermal cameras are, of course, competitors, whereas a conventional DSLR camera doesn't see anything.

And if you look closely, you see that, especially in the presence of motion in the dark, like the door frames, these vertical edges, you get a lot of motion artifacts. And what's also surprising, that even surprised ourselves, the resolution of the sensor is much higher than ours, but still we provide a better image quality.

The same also holds-- in the very same scenario-- sorry, in the very same scenario, again, with the same cameras. So we have a very high dynamic range, which means we can resolve the picture frame here. And for comparison, this other existing sensor is not able to resolve that dynamic range in a single video frame.

That actually leads over to our applications for customers. So for vehicles, autonomous trucks, autonomous cars, for example, and even for humans, it's very challenging to see in a scenario. We have headlights, but because of eye safety constraints, they only have a certain range, and we can't see what's beyond that.

And, actually, with our high dynamic range and the ability to resolve motion, we can increase the range of cameras. The same holds in a surveillance scenario. So well-lit regions, we have no problems imaging in, but you see these dark regions where it's really hard to see anything.

In a defense space, we provide low light imaging capabilities and even night vision, which is especially beneficial in the presence of motion. And a huge advantage is that this is passive sensing, which means we don't have any light source or lasers or anything because these active signals give away the on position, which is certainly not desired.

And for drones, what we envision here is the ability to fly faster in low light because these camera systems can do motion compensation. Ubicept's camera systems can do motion compensation and fly faster, and therefore, reduce the strain on the battery.

We are currently active in these verticals on the left with a special focus on mobility and surveillance, but we're open to exploring other application scenarios as well. So there we are looking to connect with companies, specifically, perception teams that have interest in testing our devices, very happy to stop by and present them in person with the goal to come up with further engagement. And these companies could, hopefully, incorporate our technology into their products.

With that, thank you very much, and let's open for a Q&A.

MARCUS DAHLLOF: Great. Thank you, Sebastian. Maybe tell us a little bit more about your roadmap, your plans for the next 12 months. What are you looking to get done? What are you looking to test and validate?

SEBASTIAN BAUER: Right. So the first step, really, is to come up with demo systems. We've seen these videos so far, which is great, but we need to test them in more relevant customer scenarios. So really put them on a car and show this extended range. That's one thing.

In the first three months, let's say, at the same time, also engage with more customers to get that customer pool to, basically, drive forward and really apply a lot of customer discovery there. And pretty much put it into a strategy to be able to serve these different markets and capturing the global maximum there instead of being pigeonholed into one market. But happy to really stop by and show these things in person.

MARCUS DAHLLOF: Can you tell us a little bit more about the size of the system and some of the specifications, and also, try to tie that with the different use cases? Are there limitations in your specs or sort of the operating parameters that define what is a better use case or not?

SEBASTIAN BAUER: That's a very good question. In terms of parameters, the power consumption is about 20 watts. The camera itself will be 10 by 10 by 5 centimeters, so it's kind of like a conventional digital camera, not too big. And we are aiming at reducing these requirements obviously, so.

Initially, what we think is super interesting application-wise are stationary cameras, for example, because they don't have strict requirements, and also probably trucks, autonomous trucks, is a good application scenario because also price also is an issue, obviously, but that's much less of a concern with trucks than with cars. And we'll get there, and it's really the goal to reduce size, weight, power, and cost.

MARCUS DAHLLOF: Got it. Thank you, Sebastian.