Soon after graduating in the summer of 2020, I had the pleasure of joining the startup Kopernikus Automotive to work on automated valet parking (AVP). The idea is to enable a car to fully park itself with minimal hardware on board, instead of using external infrastructure to support it. In our case, we use cameras installed throughout the parking lot and deep learning to see the cars we’re driving and everything around them. Then we calculate a trajectory and send it to the car to execute it. It is a very different architecture from all other autonomous cars but one that makes a lot of sense if you’re confining cars to level 4 autonomous anyway.
During my time at Kopernikus, I focused most of my time on path planning and control but as it is with most startups, if you want, you end up doing a bit of everything. In my case, I learned a lot about camera deep learning, MLOps, system design, and everything else you’d imagine goes behind fully autonomous cars. Surprisingly I also helped in the development of the ISO standard for AVP.
My biggest achievement at Kopernikus was at the International Automotive Conference 2021 where I led showcasing our solution and its operability driving vehicles from 5 different OEMs. Apart from developing, refining, and adapting our tech for the demo, I was project planning, managing, liaising APIs with partners and ultimately delivered an exciting peek into AV tech that will go in producting hopefully by 2024.
Oh and I also got to drive a fancy Jag and be a model in the official promo video :)