RuleFuser: An Evidential Bayes Approach for Rule Injection in Imitation Learned Planners and Predictors for Robustness under Distribution Shifts

TL;DR

RuleFuser is an evidential framework that combines imitation learning and rule-based planners to improve safety and robustness of autonomous driving in out-of-distribution scenarios while maintaining imitation performance.

Contribution

It introduces RuleFuser, a novel evidential approach that effectively fuses IL and rule-based planners for safer and more robust autonomous driving.

Findings

38.43% average safety improvement over IL planner

Maintains high imitation performance in in-distribution scenarios

Enhances safety in out-of-distribution scenarios

Abstract

Modern motion planners for autonomous driving frequently use imitation learning (IL) to draw from expert driving logs. Although IL benefits from its ability to glean nuanced and multi-modal human driving behaviors from large datasets, the resulting planners often struggle with out-of-distribution (OOD) scenarios and with traffic rule compliance. On the other hand, classical rule-based planners, by design, can generate safe traffic rule compliant behaviors while being robust to OOD scenarios, but these planners fail to capture nuances in agent-to-agent interactions and human drivers' intent. RuleFuser, an evidential framework, combines IL planners with classical rule-based planners to draw on the complementary benefits of both, thereby striking a balance between imitation and safety. Our approach, tested on the real-world nuPlan dataset, combines the IL planner's high performance in in-distribution (ID) scenarios with the rule-based planners' enhanced safety in out-of-distribution (OOD) scenarios, achieving a 38.43% average improvement on safety metrics over the IL planner without much detriment to imitation metrics in OOD scenarios.