A Shared Control Framework for Mobile Robots with Planning-Level Intention Prediction

TL;DR

This work addresses the challenge of reducing operator workload in mobile-robot shared control without sacrificing safety or efficiency by introducing planning-level intention prediction. It proposes a closed-cone intention domain $\mathcal{D}_I(\mathbf{p},\mathbf{u},h,r)$ to constrain replanning, and jointly learns intention-domain prediction and path replanning via deep reinforcement learning, supported by a Voronoi-based trajectory generator for simulation-only training. Key contributions include the closed-cone domain, an intention-aware two-stage replanning strategy, a hybrid RL framework (H-PPO) for trigger and domain-parameter decisions, and comprehensive simulations plus a real-robot user study showing reduced operator workload and enhanced safety. The approach demonstrates effective sim-to-real transfer and offers a practical path to safer, more autonomous planning-aware shared control in dynamic environments.

Abstract

In mobile robot shared control, effectively understanding human motion intention is critical for seamless human-robot collaboration. This paper presents a novel shared control framework featuring planning-level intention prediction. A path replanning algorithm is designed to adjust the robot's desired trajectory according to inferred human intentions. To represent future motion intentions, we introduce the concept of an intention domain, which serves as a constraint for path replanning. The intention-domain prediction and path replanning problems are jointly formulated as a Markov Decision Process and solved through deep reinforcement learning. In addition, a Voronoi-based human trajectory generation algorithm is developed, allowing the model to be trained entirely in simulation without human participation or demonstration data. Extensive simulations and real-world user studies demonstrate that the proposed method significantly reduces operator workload and enhances safety, without compromising task efficiency compared with existing assistive teleoperation approaches.

Figures (9)

• Figure 1: System configuration.
• Figure 2: The proposed shared control framework. Purple denotes environment and task priors, sky blue indicates robot intentions, green represents human motion intentions, and yellow shows inferred intentions in the form of an intention domain. Dark blue components correspond to modules used for policy training.
• Figure 3: Voronoi-based human trajectory generation algorithm: (a) Extracte the Discretized Voronoi Diagram (DVD) from the occupancy grid map. The start and goal are denoted by the green and red points. (b) Convert DVD into a Voronoi Graph. The sampled points are indicated by yellow nodes. (c) Generate initial trajectory along the Voronoi Graph edges, construct circular corridors and sample via-points within each corridor. (d) Trajectory planning under both corridor and via-point constraints
• Figure 4: Heat map visualization of the distribution of replanning trigger locations, evaluated over 256 human-desired trajectories. The discretized Voronoi diagram is overlaid in green lines.
• Figure 5: Visualization of three subgoal distribution scenarios.