Graph Isomorphisms and Homeomorphisms for Part and Assembly Matching: Distributed System and Provenance with Blockchain Technology

Abstract

In recent times, the democratization of manufacturing through various developments in the industry, such as 3-D printing and crowdsourcing, has led to increasing levels of innovation and a greater ability for smaller organizations and even individual entrepreneurs to participate in the manufacturing process. One problem to be solved with the advent of the democratization of manufacturing is how a manufacturer can determine if a required part is already being manufactured by another manufacture exactly or with sufficient similarity without having to manually search through existing parts.

In this thesis, we present algorithms for searching existing parts and assemblies for a pattern specification, and we propose some distance measures for determining how similar two matched part or assembly trees are based on node attributes. Furthermore, we present a distributed system that can be used to search parts and assemblies created by other manufacturers, make purchases after receiving results with respective distances, and track provenance of parts used in assemblies by third-party manufacturers using blockchain technology, all without requiring a centralized authority or database.