3D Scan Enabled Machine Learning for Reverse Engineering and Parametric Modeling (3SEML)

The Department of Defense (DOD) has faced significant challenges in the field regarding Diminishing Manufacturing Sources and Material Shortages (DMSMS). These issues cost upwards of $750,000,000 per year and cause breakdowns in military effectiveness and readiness. While there are several mitigation strategies for these issues, most of these options require significant upfront financial investment and introduce further logistical complications. This cost and the challenges of supply chain, military readiness, and DMSMS mitigation demand a solution that is cost effective, robust, and field-ready to allow deployed personnel to quickly respond to part shortages and sustainability needs with onsite reverse engineering and immediate manufacture of replacement parts.To address this need, XO Armor has developed technology capable of fast on-site precision 3D scanning, modeling, and immediate manufacturing (3D printing) in both austere and forward operating environments. The XO system is a fully offline 3D printing package that includes an easy-to-use handheld 3D scanner, reliable and fast commercially available 3D printers, and a portable computer with user friendly 3D manipulation software. Using our preexisting base technology, XO Armor proposes to create the 3D Scan Enabled Machine Learning for Reverse Engineering and Parametric Modeling (3SEML), which will involve software developments to empower the user with fast on-site reverse engineering and immediate manufacture of non-electronic parts via 3D printing. The project will follow three distinct phases: 1) Converting precision 3D scans from the XO Armor 3D scanner to user-adjustable parametric models. 2) Expanding the capability of the software to read-in technical drawings to user-adjustable parametric models. 3) Convert the workflows established in the first two phases to integrate with a machine learning (ML) database to automatically determine component details from a 3D surface scan using a database of technical drawings and prior component designs. The proposed software will be developed to allow a non-technical individual to create and adjust a 3D model to subsequently additively manufacture to meet US Army needs.The results from Phase I will inform the design team about the advantages and shortcomings related to an automated process that can incorporate 3D scan, technical drawings, and discern components using ML. Success in Phase I will inform a Phase II roadmap to deliver a fully functional AI solution for reverse engineering. The Phase I work also serves as a foundation for a field deployable solution to overcome the challenges associated with contested logistics. All components of Phase I will be designed with the requirement that a non-technical user operates the system in a field-deployed situation. The Phase I technical objectives serve as a foundation for a truly autonomous AI solution that can revolutionize reverse engineering and logistical support in military environments.