State-Based Disassembly Planning

TL;DR

This paper addresses the inefficiency and limited autonomy of physics-based disassembly planners that rely on exhaustive simulations and manual motion-mode selection. It introduces State-Based Disassembly Planning (SBDP), a framework that prioritizes translational motion, stores intermediate motion states, and uses Directional Blocking Graphs (DBGs) with state information to guide search. DBG-guided SBDP achieves state-of-the-art performance on a large benchmark, solving 4135 of 4196 problems (98.55%) and reducing search space and planning time relative to prior methods, while enabling mixed translational and rotational actions within a single plan. The work significantly advances autonomous, scalable disassembly planning and offers pathways for integration with learning-based guidance and assembly-by-disassembly tasks in industrial settings.

Abstract

It has been shown recently that physics-based simulation significantly enhances the disassembly capabilities of real-world assemblies with diverse 3D shapes and stringent motion constraints. However, the efficiency suffers when tackling intricate disassembly tasks that require numerous simulations and increased simulation time. In this work, we propose a State-Based Disassembly Planning (SBDP) approach, prioritizing physics-based simulation with translational motion over rotational motion to facilitate autonomy, reducing dependency on human input, while storing intermediate motion states to improve search scalability. We introduce two novel evaluation functions derived from new Directional Blocking Graphs (DBGs) enriched with state information to scale up the search. Our experiments show that SBDP with new evaluation functions and DBGs constraints outperforms the state-of-the-art in disassembly planning in terms of success rate and computational efficiency over benchmark datasets consisting of thousands of physically valid industrial assemblies.

Figures (10)

• Figure 1: An illustration of DBGs static analysis (left) with respect to a bolt $p_b$ and a washer $p_w$ along six directions; a demonstration of potential blocking assessment (right).
• Figure 2: DBGs examples with respect to construction through static analysis (grey) and updates due to the collision state (blue) and the disassembled state (green) in a problem $\mathcal{P}$ that consists of a bolt $p_b$, a washer $p_w$, and a pin $p_p$.
• Figure 3: A flow chart of DBG-guided SBDP. DBGs constructions and updates are highlighted with the same colors used in Figure \ref{['fig:DBG_in_SBDP']} and usages are highlighted with yellow in both translational planning (left) and rotational planning (right). TP and RP are acronyms of translational planning and rotational planning, respectively.
• Figure 4: An example disassembly planning process with DBGs in SBDP for a problem $\mathcal{P}$ that consists of a bolt $p_b$, a cover $p_c$, and a pin $p_p$. The colored DBGs represent the same constructions and updates approaches as depicted in Figure \ref{['fig:DBG_in_SBDP']}.
• Figure 5: Disassembly benchmark examples from the small, medium, and large categories.