FastGrasp: Learning-based Whole-body Control method for Fast Dexterous Grasping with Mobile Manipulators

2026-04-14

AI Summary

This work proposes FastGrasp, a framework addressing the challenges of impact stability, whole-body real-time coordination, and cross-object generalization in high-speed grasping with mobile manipulators. FastGrasp integrates point cloud–driven grasp generation, whole-body dynamic control, and tactile feedback within a novel two-stage reinforcement learning strategy: it first employs a conditional variational autoencoder to generate diverse, feasible grasp poses, then dynamically refines these poses using tactile information. Experimental results demonstrate that FastGrasp efficiently accomplishes rapid grasping tasks across multiple object categories in both simulation and real-world environments, exhibiting strong robustness and effective sim-to-real transfer capability.

Technology Category

• Intelligent Robots: Manipulation
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Computer Vision: Generative Adversarial Networks (GANs) for Vision

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Fast grasping is critical for mobile robots in logistics, manufacturing, and service applications. Existing methods face fundamental challenges in impact stabilization under high-speed motion, real-time whole-body coordination, and generalization across diverse objects and scenarios, limited by fixed bases, simple grippers, or slow tactile response capabilities. We propose \textbf{FastGrasp}, a learning-based framework that integrates grasp guidance, whole-body control, and tactile feedback for mobile fast grasping. Our two-stage reinforcement learning strategy first generates diverse grasp candidates via conditional variational autoencoder conditioned on object point clouds, then executes coordinated movements of mobile base, arm, and hand guided by optimal grasp selection. Tactile sensing enables real-time grasp adjustments to handle impact effects and object variations. Extensive experiments demonstrate superior grasping performance in both simulation and real-world scenarios, achieving robust manipulation across diverse object geometries through effective sim-to-real transfer.

Problem

• fast grasping
• impact stabilization
• whole-body coordination
• generalization
• mobile manipulators

Innovation

• learning-based whole-body control
• fast dexterous grasping
• tactile feedback
• conditional variational autoencoder
• sim-to-real transfer