How to develop an autonomous car end-to-end: Robotic Drive and the mobility revolution

The discussions about the mobility revolution, which we all will experience in the next decade, is strongly driven by the introduction of autonomous driving vehicles. This field is not dominated by traditional car manufactures but allows also new players to enter it – prominent examples are Waymo (a former google project) or the rumors about an Apple Car. But all are facing the same challenge: Creating a robot which looks like a car, can carry some passengers and drive from A to B in self-operated and safe manner.

For this data driven development journey it is essential to combine the core ingredients from especially IT and automotive world. Top of the edge methodologies and technologies need to come together to be able to enable and execute development activities around the globe and on several hundreds of petabytes of data. With the consequences of embedding such code in the car on those high-performance compute platforms and how to verify it so that homologation can be achieved are currently under a heavy transition. In this session we will focus on introducing a corresponding scalable compute platform as well as the virtualization which need to be executed to achieve those goals and overcome the mentioned challenges.

Ulrich’s talk will focus on the virtual validation activities (including closed and open loop simulation), while Mohamed will have a deeper look into the consequences on the ECU’s software (including ECU’s platforms and software development processes).

‍