Automotive Radar perception thoughts towards AV
Takeaways:
– Everybody speaks about vision and lidar systems, but Radar is an indispensable key technology for highly automated and autonomous ADAs
– HW-SW Co-Design is essential to come to the next level
– High Performance radar versus smart radar networks – some thoughts
– Radar perception can benefit performance, if machine learning is applied to it
– Massive AI is entering the Radar-Perception arena.
Abstract:
Over decades automotive driver assistance systems have developed from simple park assistance and ACC to highly complex L3 functions. Among other issues, that has been driven by the dramatic development of radar sensor technology and radar based perception. Radar can “see”, where other sensors can´t. Due to their performance to be packaged nearly invisible, modern cars can still appear as design artwork icons, like the present EQ-S. With the development towards higher automatization the total perception system has to provide much more comprehensive and precise information to the path planning and behavior stages in order to allow proper operation. Based on our experience with the “Bertha Drive” as quasi autonomous vehicle, we got inside what is needed in the future. One outcome of our L4/L5 research projects was, the sensor technology in cooperation with perception algo-concepts have to develop along other lines as compared to simple L2 systems. The key note will line out from a radar point of view what has been achieved so far and what will be needed in the future for L3 to L5 driving.
Dr.-Ing. Juergen Dickmann is head of Radar sensors and radar-based perception for highly automated/autonomous driving, Mercedes-Benz AG. In this role he is responsible for the 6th generation at Mercedes-Benz AG and future radar solutions towards autonomous driving. In the automotive field, he introduced for the first time AI concepts for Radar and exploited Radar usage beyond ranging and detection for comprehensive environment perception. Before that, he managed the technology transfer process of all perception sensing (Radar, Laser, Video, localization etc.) technologies from research into series for former passenger car platforms. He conducted radar research and pre-developments for radar-based Pre-Crash and driver assistant systems for all company platforms (Passenger cars, Busses, Van, Truck) inside DAIMLER Chrysler AG, where he introduced Lidar technology in Chrysler cars. In 1986 he started his career at AEG Research Center, where he did research on III/V-semiconductor processing techniques, mm-Wave device and MMIC-design (up to 120GHz). He achieved several records in high frequency devices and MMICs with III/V semiconductors. He got the academic award “Preis der VDE/ITG” in 1993. The research achievements of him and his group can be found in https://www.researchgate.net/profile/Juergen_Dickmann. He holds various patents.
Speaker(s): Juergen Dickmann,
Agenda:
10:00-11:00 Meeting BGU Autonomous Driving Race Team
11:00- 12:00 BGU Autonomous Platforms (BGU ECE research labs)
12:00-13:00 Automotive Radar perception thoughts towards AV ECE BGU Seminar
13:00-15:00 Autonomous vehicle Cyber security with BGU accelerator team
Virtual: https://events.vtools.ieee.org/m/298697