Online Intention Prediction via Control-Informed Learning
2026-04-10 • Robotics
RoboticsMachine Learning
AI summaryⓘ
The authors created a system that can guess what an autonomous machine, like a drone, is trying to do while it is working, even if its goals change over time. They treat the machine's goal as a hidden setting in a control problem and update their guesses continuously using new data. Their method uses smart ways to ignore old info and learn efficiently as the machine moves. Tests in computer simulations and with a flying drone showed their approach can predict goals accurately in tricky situations.
intention predictioninverse optimal controlinverse reinforcement learningonline learningshifting horizon strategygradient computationautonomous systemsquadrotor droneadaptive control
Authors
Tianyu Zhou, Zihao Liang, Zehui Lu, Shaoshuai Mou
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 A shifting horizon strategy discounts outdated information, while online control-informed learning enables efficient gradient computation and online parameter updates.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.