AVO: Amortized Value Optimization for Contact Mode Switching in Multi-Finger Manipulation

2025-10-08

AI Summary

In dexterous manipulation, conventional trajectory optimization suffers from suboptimal local minima and high computational cost due to fragmented subtask optimization during transitions among multiple contact modes (e.g., rolling, sliding, adhesion). Method: This paper proposes a learning-augmented trajectory optimization framework that models future task performance as a learnable value function; its gradient is backpropagated into the current optimization to enable cross-mode cooperative planning. Unlike traditional piecewise optimization, our method jointly optimizes for both immediate execution and favorable post-transition states under a unified objective. Results: Evaluations in simulation and real-world grasping-and-rotating experiments demonstrate that our approach achieves significantly higher task success rates and trajectory quality using only 50% of the computational resources required by state-of-the-art methods, validating its dual advantages in efficiency and robustness.

Technology Category

• Search and Optimization: Learning to Search
• Intelligent Robots: Learning & Optimization for ROB
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: The sharing economy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Dexterous manipulation tasks often require switching between different contact modes, such as rolling, sliding, sticking, or non-contact contact modes. When formulating dexterous manipulation tasks as a trajectory optimization problem, a common approach is to decompose these tasks into sub-tasks for each contact mode, which are each solved independently. Optimizing each sub-task independently can limit performance, as optimizing contact points, contact forces, or other variables without information about future sub-tasks can place the system in a state from which it is challenging to make progress on subsequent sub-tasks. Further, optimizing these sub-tasks is very computationally expensive. To address these challenges, we propose Amortized Value Optimization (AVO), which introduces a learned value function that predicts the total future task performance. By incorporating this value function into the cost of the trajectory optimization at each planning step, the value function gradients guide the optimizer toward states that minimize the cost in future sub-tasks. This effectively bridges separately optimized sub-tasks, and accelerates the optimization by reducing the amount of online computation needed. We validate AVO on a screwdriver grasping and turning task in both simulation and real world experiments, and show improved performance even with 50% less computational budget compared to trajectory optimization without the value function.

Problem

• Optimizing multi-finger manipulation with contact mode switching
• Addressing computational expense in trajectory optimization sub-tasks
• Improving performance across sequential manipulation sub-tasks

Innovation

• Amortized Value Optimization guides trajectory optimization with learned value function
• Value function bridges separately optimized contact mode sub-tasks
• Method reduces online computation while improving manipulation performance