Tracing Energy Flow: Learning Tactile-based Grasping Force Control to Prevent Slippage in Dynamic Object Interaction

2025-12-24

AI Summary

Slip during dynamic grasping remains challenging due to unknown object properties, multiple rolling contacts, and unreliable external sensing. Method: This paper proposes a purely haptic-driven, real-time grasp force control strategy. It introduces a physics-inspired “energy reservoir” abstraction, where slip stability is determined by the instantaneous mismatch between finger input power and object stored energy—eliminating reliance on vision, prior knowledge, or external sensors. The approach integrates physical modeling, tactile-signal-driven model learning, and differentiable real-time optimization into an end-to-end energy-dynamics-based control framework. Contribution/Results: Simulation and real-robot experiments demonstrate that the method achieves significant slip reduction and extended stable grasp duration after only minutes of online learning. It exhibits strong generalization across diverse object–motion combinations, enabling robust, adaptive manipulation without external perception.

Technology Category

• Intelligent Robots: Manipulation
• Computer Vision: Low Level & Physics-based Vision
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Regulating grasping force to reduce slippage during dynamic object interaction remains a fundamental challenge in robotic manipulation, especially when objects are manipulated by multiple rolling contacts, have unknown properties (such as mass or surface conditions), and when external sensing is unreliable. In contrast, humans can quickly regulate grasping force by touch, even without visual cues. Inspired by this ability, we aim to enable robotic hands to rapidly explore objects and learn tactile-driven grasping force control under motion and limited sensing. We propose a physics-informed energy abstraction that models the object as a virtual energy container. The inconsistency between the fingers' applied power and the object's retained energy provides a physically grounded signal for inferring slip-aware stability. Building on this abstraction, we employ model-based learning and planning to efficiently model energy dynamics from tactile sensing and perform real-time grasping force optimization. Experiments in both simulation and hardware demonstrate that our method can learn grasping force control from scratch within minutes, effectively reduce slippage, and extend grasp duration across diverse motion-object pairs, all without relying on external sensing or prior object knowledge.

Problem

• Learning tactile-driven grasping force control to prevent slippage
• Modeling object as energy container to infer slip-aware stability
• Optimizing grasping force without external sensing or prior knowledge

Innovation

• Physics-informed energy abstraction models object dynamics
• Model-based learning optimizes grasping force from tactile sensing
• Real-time planning prevents slippage without external sensors