Development and Testing of a Software Stack for an Autonomous Racing Vehicle

Abstract

Autonomous racing aims to replicate the human racecar driver with software and sensors. As in traditional motorsports, Autonomous Racing Vehicles (ARVs) are pushed to their dynamic limits in multi-agent scenarios at high speeds (>=100mph). This Operational Design Domain (ODD) presents unique challenges across the autonomy stack. The Indy Autonomous Challenge (IAC) is an international competition aiming to advance autonomous vehicle development through ARV competitions. To compete in the IAC, we developed a full autonomy stack from scratch.

We present our design philosophy and strategy when developing the stack from scratch while under the constraints of the competition, including the ODD, rules, timeline, and limited field testing time. In particular, we present our contributions to the design, integration, and testing of the Perception and State Estimation systems and lessons learned from testing and deploying these systems on a real, full-sized system (the Dallara AV-21 platform). Finally, we demonstrate how our design process enabled rapid adaptation and development of the stack, which was shown capable of overtaking an opponent ARV at speeds exceeding 150mph.