SBIR/STTR Award attributes
Many maintenance and supply activities for nuclear weapons (and elsewhere across AF/DoD and wider government) are often still using paper-based, manual products - for debriefing, work order, item replacement, maintenance data management etc. In order to improve supply and maintenance processes (incl. supply chain risk analysis), the Air Force and others need a software tool to automate and optimize these activities. As per the solicitation, the desired digital tool needs to be able to: (1) allow on-site users to view collected and integrated component information after scanning the UII code and to interact with the backend system; (2) gather and integrate significant amounts of infrastructure data continuously from many sources for supply and maintenance (focused towards the authorization orders and the warfighter); (3) communicate information across weapon platforms and information systems (for both government and contractors); (4) enable data visualization; (5) provide (predictive) data analytics functionality, and; (5) allow users to query and organize data as required to make data-driven decisions based on their role and function across the weapon system. We propose “TAV-SCRAMS”, a software solution consisting of COTS scanner(s), COTS tablet(s) (e.g., Android based), and an Artificial Intelligence & Machine Learning (AI/ML) based SCRM backend. The solution builds upon and extends our SCRAMS supply chain risk software platform and other technology components as a foundation to carry out the proposed R&D. The solution includes specific features for job-site users and for off-site users, minimizing the risk of human error, improving visibility of operations end-to-end, improved cross-organization efficiency etc. AI/ML is a central feature of TAV-SCRAMS in the background processing of enriched data associated to materials, products, and processes including analyzing risks and performing predictive analysis. Furthermore, using AI/ML, TAV-SCRAMS will intelligently and continuously integrate data from various targeted sources and learn the data in supervised, unsupervised, and reinforced AI/ML manners (with human oversight) to manage data-to-data and data-to-users connections, validate data integrity, and communicate results. The proposed system will also apply AI/ML extensively for predictive analysis, risk assessment, and automation features (data ingestion, data integrity, etc.). Our Phase I feasibility study results include risk assessment, predictive analysis with ML, data integration from multiple sources, SCRM, graph DB, scanner and tablet, fine-grained access control etc. In particular, we have developed SCRAMS (Supply Chain Risk Analysis Management Solution, a software product that provides organizations with visibility into their supply chains and automatically identifies supply chain risks (SCR). In the proposed effort we will research and develop a working prototype, and validate the working prototype through experimentation and validation.
