Preliminary Results on Generating Assembly Sequences for Shape Display

Recently there has been much interest in developing shape display systems, i.e. systems that can generate, based on a virtual representation such as a CAD file, a physical shape that a user can touch. The approach considered here is to automatically assemble a shape from simple, identical building blocks using a robot. This paper describes preliminary research results on generating assembly sequences for this purpose, using a rule-based approach in order to deal with the huge search space due to the large number of components. While many of the details of the rule-based approach are discussed elsewhere, this paper focuses on analyzing and comparing the use of two different methods, the assembly method and the disassembly method in the z-direction, for assembly sequence generation.