Title: Resilient Autonomous Driving

Speaker: Federico Lucchetti, CritiX, University of Luxembourg

Date: January 18, 14h00

Where: FCUL, room C6.3.27

Abstract: Autonomous driver assistance systems (ADAS) have been progressively pushed to extremes. Today, increasingly sophisticated algorithms, such as deep neural networks, assume responsibility for critical driving functionality, including operating the vehicle at various levels of autonomy.
In this Navtalk, we share our findings in retrofitting application-agnostic resilience mechanisms into existing hardware and software-stack for autonomous driving. The goal is to ultimately decrease the vulnerability of autonomously driving vehicles to accidental faults and attacks, allowing them to absorb and tolerate both. It is demonstrated the replication and rejuvenation features on the driving stack’s control module and discussed how the resilience can be extended both downwards to the hardware level as well as upwards to the prediction and planning modules.

Bio: Federico Lucchetti is a physicist with a PhD in Neuroscience from the Free University of Brussels. He’s now working as a postdoctoral researcher at the University of Luxembourg in the CritiX research lab with Prof. Marcus Voelp. His research interests revolve around the topic of resilient computing applied to autonomous driving with a special focus on retrofitting common fault and intrusion tolerant techniques to artificial neural net-dependent modules.