Extremum Seeking Controlled Wiggling for Tactile Insertion

2024-10-03

arXiv.org

AI Summary

Robotic insertion into unknown locks under purely tactile feedback remains challenging due to the absence of visual cues, precise lock models, or prior knowledge of key-lock geometry. Method: We propose a model-free, parameter-free “jiggle-and-perceive” control framework inspired by human manipulation strategies. It employs extremum seeking control (ESC) to apply online stochastic perturbations in end-effector pose space, leveraging high-frequency (13 Hz) GelSight Mini tactile feedback to simultaneously maximize insertion depth and minimize strain—without requiring dynamics modeling or lock-specific calibration. Contribution/Results: The framework generalizes across diverse key-lock pairs without task-specific tuning. In experiments, it achieves 71% success rate (120 trials) from random initialization, improving to 84% (240 trials) with single-degree-of-freedom perturbation optimization. Average insertion time reduces from 262 seconds to 147 seconds, demonstrating robustness and efficiency in unstructured, model-agnostic tactile insertion.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

When humans perform insertion tasks such as inserting a cup into a cupboard, routing a cable, or key insertion, they wiggle the object and observe the process through tactile and proprioceptive feedback. While recent advances in tactile sensors have resulted in tactile-based approaches, there has not been a generalized formulation based on wiggling similar to human behavior. Thus, we propose an extremum-seeking control law that can insert four keys into four types of locks without control parameter tuning despite significant variation in lock type. The resulting model-free formulation wiggles the end effector pose to maximize insertion depth while minimizing strain as measured by a GelSight Mini tactile sensor that grasps a key. The algorithm achieves a 71% success rate over 120 randomly initialized trials with uncertainty in both translation and orientation. Over 240 deterministically initialized trials, where only one translation or rotation parameter is perturbed, 84% of trials succeeded. Given tactile feedback at 13 Hz, the mean insertion time for these groups of trials are 262 and 147 seconds respectively.

Problem

• Developing robotic wiggling control for tactile-guided insertion tasks
• Maximizing insertion depth while minimizing strain via extremum seeking
• Solving complex geometry insertions without contact modeling or learning

Innovation

• Extremum seeking control wiggles end effector pose
• Maximizes insertion depth while minimizing tactile strain
• Uses GelSight Mini sensor for closed loop feedback