Learning Autonomous Docking Operation of Fully Actuated Autonomous Surface Vessel from Expert data

TL;DR

This paper introduces a novel imitation learning approach using inverse reinforcement learning to enable fully actuated autonomous surface vessels to perform docking maneuvers based on expert demonstrations, integrating sensor data and vehicle kinematics.

Contribution

It presents a two-stage neural network architecture for learning reward functions from expert data, improving autonomous docking performance in varied environments.

Findings

Effective in simulation for human-like docking behaviors

Generalizes across different environmental configurations

Integrates sensor data and vehicle kinematics in reward learning

Abstract

This paper presents an approach for autonomous docking of a fully actuated autonomous surface vessel using expert demonstration data. We frame the docking problem as an imitation learning task and employ inverse reinforcement learning (IRL) to learn a reward function from expert trajectories. A two-stage neural network architecture is implemented to incorporate both environmental context from sensors and vehicle kinematics into the reward function. The learned reward is then used with a motion planner to generate docking trajectories. Experiments in simulation demonstrate the effectiveness of this approach in producing human-like docking behaviors across different environmental configurations.