Parallel Self-Assembly and Sorting of Polyominoes under Uniform Control Inputs

Automated assembly at micro and nano-scale is essential for manufacturing smaller and inexpensive products at faster rates. Traditional micro-assembly and manipulation techniques involve micro-grippers and manipulators that have limited degrees of freedom when compared to their macro-scale counterparts. A major disadvantage of these techniques is that only one part can be manufactured at a time. Thus, it would be a significant progress if control algorithms are devised that can automatically assemble a huge quantity of small scale components. We present a novel approach towards micro-assembly which employs a large swarm of micro-particles, controlled by a global signal to assemble arbitrary 2D structures. The algorithm automatically generates a micro-factory layout for a given structure, and when this layout is actuated, it manufactures the required number of copies of the shape. We have analyzed the size and time complexity of the micro-factories and present several simulation and hardware results for the assembly task. Since the assembly process is performed in open-loop, the assembled parts could have errors, and in order to detect these errors, we sort them based on their shape.