Search papers, labs, and topics across Lattice.
This paper introduces a Learning from Haptics (LfH) framework that personalizes safety interventions in haptic human-robot shared control by learning from sparse user demonstrations. By utilizing a differentiable Control Barrier Function (CBF)-based optimization layer, the system adapts safety parameters to align with individual user preferences, moving away from rigid predefined strategies. Experimental results show that the framework effectively reduces the mismatch between user expectations and the haptic feedback provided during teleoperation, enhancing the overall safety and intuitiveness of human-robot interactions.
Personalized haptic feedback can be learned from just a few user demonstrations, transforming how robots adapt to individual safety preferences in real-time.
Haptic feedback provides an implicit channel for communicating safety intentions during human-robot shared control. Existing haptic guidance systems typically employ predefined intervention strategies that cannot accommodate the diverse safety preferences of individual users or application scenarios. To address this limitation, we propose a Learning from Haptics (LfH) framework that learns user-preferred safety interventions from sparse demonstrations, eliminating the need for manual trial-and-error design. Our framework is built on a differentiable Control Barrier Function (CBF)-based optimization layer that automatically adjusts the underlying safety parameters to match the demonstrated haptic responses. Instead of tuning controller parameters directly, users teach the system how they expect it to intervene during teleoperation. The resulting haptic guidance reflects the demonstrated intervention preferences while preserving the intuitive interaction of haptic shared control. Simulation and hardware experiments demonstrate that the proposed framework can learn personalized safety interventions from sparse user input and reduce the mismatch between the generated haptic feedback and the demonstrated preferences.
