Online Intention Prediction via Control-Informed Learning

TL;DR

This paper introduces an online intention prediction method for autonomous systems that adapts to changing goals and unknown dynamics, using inverse optimal control and control-informed learning for real-time updates.

Contribution

It develops a novel online framework combining shifting horizon strategies and control-informed learning to improve intention prediction accuracy in dynamic, uncertain environments.

Findings

Achieves accurate intention prediction under varying noise levels.

Demonstrates effectiveness through hardware experiments on a quadrotor drone.

Outperforms existing methods in adaptive, real-time intention estimation.

Abstract

This paper presents an online intention prediction framework for estimating the goal state of autonomous systems in real time, even when intention is time-varying, and system dynamics or objectives include unknown parameters. The problem is formulated as an inverse optimal control / inverse reinforcement learning task, with the intention treated as a parameter in the objective. A shifting horizon strategy discounts outdated information, while online control-informed learning enables efficient gradient computation and online parameter updates. Simulations under varying noise levels and hardware experiments on a quadrotor drone demonstrate that the proposed approach achieves accurate, adaptive intention prediction in complex environments.