Exp-Force: Experience-Conditioned Pre-Grasp Force Selection with Vision-Language Models

TL;DR

Exp-Force is a framework that predicts minimal safe grasping force using prior experiences and vision-language models, improving accuracy and reliability in robotic manipulation without complex contact models.

Contribution

It introduces a novel experience-conditioned approach that leverages prior grasping experiences with vision-language models for force prediction, eliminating the need for analytical contact models.

Findings

Achieves a 72% reduction in MAE over zero-shot inference.

Increases appropriate force selection rate from 63% to 87% in real-world tests.

Demonstrates reliable, generalizable pre-grasp force prediction.

Abstract

Accurate pre-contact grasp force selection is critical for safe and reliable robotic manipulation. Adaptive controllers regulate force after contact but still require a reasonable initial estimate. Starting a grasp with too little force requires reactive adjustment, while starting a grasp with too high a force risks damaging fragile objects. This trade-off is particularly challenging for compliant grippers, whose contact mechanics are difficult to model analytically. We propose Exp-Force, an experience-conditioned framework that predicts the minimum feasible grasping force from a single RGB image. The method retrieves a small set of relevant prior grasping experiences and conditions a vision-language model on these examples for in-context inference, without analytic contact models or manually designed heuristics. On 129 object instances, ExpForce achieves a best-case MAE of 0.43 N, reducing error by 72% over zero-shot inference. In real-world tests on 30 unseen objects, it improves appropriate force selection rate from 63% to 87%. These results demonstrate that Exp-Force enables reliable and generalizable pre-grasp force selection by leveraging prior interaction experiences. http://expforcesubmission.github.io/Exp-Force-Website/