DDBot: Differentiable Physics-based Digging Robot for Unknown Granular Materials

2025-10-20

AI Summary

Precise, small-scale excavation of granular materials (e.g., sand) with unknown physical properties remains challenging due to the complexity of granular dynamics and the scarcity of real-world training data. Method: This paper introduces the first differentiable robotic excavation framework tailored for granular manipulation. It integrates GPU-accelerated differentiable physics simulation, automatic differentiation, and task-oriented demonstrations to enable zero-shot system identification and skill optimization—without requiring pre-training on real-world data. Key components include a differentiable skill-to-action mapping, gradient clipping, and line-search-based optimization. Contribution/Results: The framework achieves system identification and policy synthesis within 5–20 minutes and enables direct deployment in real-world settings. Experiments demonstrate high-precision zero-shot transfer in laboratory environments, significantly outperforming state-of-the-art methods in excavation accuracy and generalization.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Automating the manipulation of granular materials poses significant challenges due to complex contact dynamics, unpredictable material properties, and intricate system states. Existing approaches often fail to achieve efficiency and accuracy in such tasks. To fill the research gap, this paper studies the small-scale and high-precision granular material digging task with unknown physical properties. A new framework, named differentiable digging robot (DDBot), is proposed to manipulate granular materials, including sand and soil. Specifically, we equip DDBot with a differentiable physics-based simulator, tailored for granular material manipulation, powered by GPU-accelerated parallel computing and automatic differentiation. DDBot can perform efficient differentiable system identification and high-precision digging skill optimisation for unknown granular materials, which is enabled by a differentiable skill-to-action mapping, a task-oriented demonstration method, gradient clipping and line search-based gradient descent. Experimental results show that DDBot can efficiently (converge within 5 to 20 minutes) identify unknown granular material dynamics and optimise digging skills, with high-precision results in zero-shot real-world deployments, highlighting its practicality. Benchmark results against state-of-the-art baselines also confirm the robustness and efficiency of DDBot in such digging tasks.

Problem

• Automating granular material manipulation with complex dynamics
• Achieving efficiency and accuracy in unknown material digging
• Developing differentiable physics-based robot for high-precision tasks

Innovation

• Differentiable physics simulator for granular materials
• GPU-accelerated parallel computing with automatic differentiation
• Differentiable system identification and skill optimization