Interaction force estimation for tactile sensor arrays: Toward tactile-based interaction control for robotic fingers

2024-11-20

arXiv.org

AI Summary

To address the challenge of real-time, accurate estimation of omnidirectional contact forces during dexterous manipulation with robotic hands, this paper proposes an end-to-end, real-time 3D interaction force estimation method based on a tactile sensor array. Moving beyond conventional one-dimensional or decoupled modeling approaches, we develop a data-driven calibration framework explicitly accounting for material deformation and sensor nonlinearities, enabling tight integration of tactile perception and force control in a closed-loop architecture. Leveraging the uSkin flexible tactile array, the Allegro dexterous hand, and a ROS-based real-time system, our deep learning–based regression model achieves sub-millisecond 3D force estimation with a mean error of less than 0.12 N. Experimental results demonstrate substantial improvements in grasp adaptability and manipulation stability, successfully enabling delicate object grasping and dynamic contact force tracking.

Technology Category

• Intelligent Robots: Manipulation
• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Interaction Techniques and Devices

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Accurate estimation of interaction forces is crucial for achieving fine, dexterous control in robotic systems. Although tactile sensor arrays offer rich sensing capabilities, their effective use has been limited by challenges such as calibration complexities, nonlinearities, and deformation. In this paper, we tackle these issues by presenting a novel method for obtaining 3D force estimation using tactile sensor arrays. Unlike existing approaches that focus on specific or decoupled force components, our method estimates full 3D interaction forces across an array of distributed sensors, providing comprehensive real-time feedback. Through systematic data collection and model training, our approach overcomes the limitations of prior methods, achieving accurate and reliable tactile-based force estimation. Besides, we integrate this estimation in a real-time control loop, enabling implicit, stable force regulation that is critical for precise robotic manipulation. Experimental validation on the Allegro robot hand with uSkin sensors demonstrates the effectiveness of our approach in real-time control, and its ability to enhance the robot's adaptability and dexterity.

Problem

• Accurate force estimation on complex robotic hand surfaces.
• Integration of tactile sensing into reactive control loops.
• Data-efficient calibration for real-time force feedback.

Innovation

• Data-efficient calibration for force estimation
• Online closed-loop control for force tracking
• Admittance control loop at 100Hz precision