Redundant Perception and State Estimation for Reliable Autonomous Racing

TL;DR

This paper develops a redundant perception and state estimation system for autonomous racing cars, combining sensor modalities and probabilistic failure detection to improve reliability during high-speed operations.

Contribution

It introduces a novel multi-sensor perception framework with real-time fusion and failure detection tailored for autonomous racing environments.

Findings

Achieved lateral accelerations up to 1.7G

Reached top speeds of 90 km/h in real-world tests

Demonstrated robustness to sensor failures

Abstract

In autonomous racing, vehicles operate close to the limits of handling and a sensor failure can have critical consequences. To limit the impact of such failures, this paper presents the redundant perception and state estimation approaches developed for an autonomous race car. Redundancy in perception is achieved by estimating the color and position of the track delimiting objects using two sensor modalities independently. Specifically, learning-based approaches are used to generate color and pose estimates, from LiDAR and camera data respectively. The redundant perception inputs are fused by a particle filter based SLAM algorithm that operates in real-time. Velocity is estimated using slip dynamics, with reliability being ensured through a probabilistic failure detection algorithm. The sub-modules are extensively evaluated in real-world racing conditions using the autonomous race car "gotthard driverless", achieving lateral accelerations up to 1.7G and a top speed of 90km/h.