Delayed-Decision Motion Planning in the Presence of Multiple Predictions

TL;DR

This paper introduces a probabilistic behavior planning framework for automated driving that accounts for multiple possible traffic agent behaviors, using a maximum entropy approach to enable delayed decisions and improve safety.

Contribution

It formalizes a novel behavior planning scheme with multiple future predictions, incorporating maximum entropy principles and model predictive control for safer autonomous driving.

Findings

Delayed decision-making enhances safety in uncertain traffic scenarios.

The proposed method is computationally feasible via quadratic programming.

Validation in simulation and on a mobile robot demonstrates effectiveness.

Abstract

Reliable automated driving technology is challenged by various sources of uncertainties, in particular, behavioral uncertainties of traffic agents. It is common for traffic agents to have intentions that are unknown to others, leaving an automated driving car to reason over multiple possible behaviors. This paper formalizes a behavior planning scheme in the presence of multiple possible futures with corresponding probabilities. We present a maximum entropy formulation and show how, under certain assumptions, this allows delayed decision-making to improve safety. The general formulation is then turned into a model predictive control formulation, which is solved as a quadratic program or a set of quadratic programs. We discuss implementation details for improving computation and verify operation in simulation and on a mobile robot.