What clouds have to do with autonomous driving Engineer for Environment Sensing

More than 100,000 points start flashing. Invisible laser beams bounce from object to object. The surroundings are scanned and captured in a matter of seconds. What sounds like a science fiction episode is actually a Lidar scanner. It is working in high gear when it comes to the development of fully automated vehicles. Andras Tüzkö has a degree in computer science and is an engineer for environment sensing. He provides an insight into the world of autonomous driving and the Formula Student Germany, where he was able to put his knowledge to practical use for the first time.

Mr. Tüzkö, please introduce yourself briefly.
I am 28 years old, come from Hungary and started at Daimler in the fall of 2017. At present, I am working on fully automated and self-driving vehicles in a joint project of Daimler and Bosch. I am teaching our cars to understand their surroundings based on sensors. Before that, I earned my master's degree in computer science at the KIT Karlsruhe with an emphasis on "machine learning."

How did you end up at Daimler?
It was a total coincidence, through the Formula Student Germany, a design competition for students. When I participated in 2017, it was the first year students were also allowed to build a self-driving car. I was very happy. Because this was my chance to help work on a car as a computer scientist. Daimler sponsored our KIT Driverless Team and supported the vehicle construction with mentors from different specialist units. When I sat in the stands watching our vehicle on the racetrack, a colleague from Daimler suddenly turned around and asked "do you need a job?" We then had a speed job interview there in the stands.

That sounds strange. Just like that?
Yes, he must have learned that I looked after the sensors of the self-driving car. That piqued his curiosity. Today I am very glad that he reached out to me on the spur of the moment. Of course, it didn't go entirely without formalities. I officially applied at Daimler afterwards and had a proper job interview.

How exactly does the FSG work? What were the challenges in the Driverless Team?
The teams work on their vehicle for almost exactly a year and are sponsored by companies. In addition, some companies also provide mentors, that is to say experts in their field, who help the students with the build of the vehicle. Daimler has been a partner of the FSG since 2008 and is therefore strongly involved. The challenge for us in the Driverless Team specifically was that we had to start completely from scratch. Because there were no comparable vehicles from previous years.

You tell us that you looked after the sensors. What exactly does this mean?
The route is marked with pylons and the camera system I developed identified these pylons and evaluated them appropriately. This means I had to program the software to be able to discern the different colors of the pylons from the camera image and evaluate them. In other words, yellow meant to drive left, blue right, for example. Today I do this in a somewhat more complex way in my job as an engineer for environment sensing.

What do you like about the FSG? Why should you participate as a student or mentor?
I love the challenge. I was able to learn a lot when I took part myself. Not only for my own specialist knowledge, the camera system, but also in areas that I didn't know anything about before. For example, I also helped in my team with the development of the emergency braking system. Those are things I couldn't have done at all without the FSG. You also learn how to work in a large team. That is why I am participating this year as a mentor from Daimler. I want to share my knowledge, pass on my experience and support the students.

You are working on a joint project with Bosch these days. What exactly is behind this?
We want to develop an autonomous vehicle that will allow us to offer a fully automated and driverless ridesharing service. It basically involves a car2go that comes to me when I need it. We share an office with the colleagues and work on the concept. It is a highly exciting project!

What are your duties there?

I am on the Lidar team and we develop different algorithms for the environment detection of the sensors. This works using so-called Lidar scanners. They send out laser beams that are harmless to people, which are reflected by a surface. The sensors measure the time it takes for the beams to make it back. A surface can be a car driving ahead, the guardrail or a stop light, for example. The result of this laser beam measurement is called a point cloud. It's like a map on which you can see the objects and their distance.

That sounds highly complex.

Yes, this so-called point cloud usually consists of more than 100,000 points that must be identified. Specifically, this means that our team programs whether an individual point is hitting a vehicle, the road or a building just then. This is how we are slowly moving closer to attaining automatic identification.

Please tell us a little bit about your team. What is its composition?

We are a young team, all engineers with a broad range of degrees. From computer science and electrical engineering to physics and mathematics, we have it all. Some are more in charge of the hardware, others of the software. Of course, the focus is on the latter when it comes to autonomous driving.

More information about FSG is available here.

The Formula Student Germany (FSG) is an international design competition for students. For Daimler, the FSG is an ideal platform to engage with future automotive engineers. The competition fosters commitment, entrepreneurial spirit, independent work and the courage to innovate in the students – traits that are crucial for shaping the mobility of the future. Daimler not only provides monetary support for the teams, but also provides employees from different areas as mentors, who pass on knowledge on the one hand and on the other foster the talents of the future at the same time. Daimler will again sponsor several teams next season.

FSG info sidebar

We use cookies

We want to make our website more user-friendly and continuously improve it. If you continue to use the website, you agree to the use of cookies.