This paper presents a novel assembly sequence planning (ASP) procedure utilizing a subassembly based search algorithm (SABLS) for micro-assembly applications involving geometric and other assembly constraints. The breakout local search (BLS) algorithm is adapted to provide sequencing solutions in assemblies with no coherent solutions by converting the final assembly into subassemblies which can be assembled together. This is implemented using custom-made microparts which fit together only in a predefined fashion. Once the ASP is done, the parts are manipulated from a cluttered space to their final positions in the subassemblies using a path-planning algorithm based on rapidly exploring random tree (RRT*), a random-sampling based execution, and micromanipulation motion primitives. The entire system is demonstrated by assembling LEGO® inspired microparts into various configurations which involve subassemblies, showing the versatility of the system.

Assembly sequence planning is an engineering problem that has been of great interest in the manufacturing field at the macro-scale. As more complex assemblies are desired at the micro and nano scales it is no longer feasible for human beings to plan and execute the production of these systems. A promising algorithm that allows optimization of assembly sequence plans that has been developed is called the Breakout Local Search. One drawback of this algorithm is its inability to consider the need for intermediate sub-assemblies to generate feasible solutions. Here an expansion to the BLS algorithm, called the Sub Assembly Generating BLS (SABLS) algorithm, is proposed. The fitness function of this new algorithm is also tailored to the specific constraints and motion primitives for a micromanipulation test-bed allowing for its use in microassembly applications. It is shown that the proposed algorithm is capable of generating optimized solutions that can be assembled with this limited degree of freedom system.