Phase II year
2019
(last award dollars: 2022)
Phase II Amount
$4,282,513
Aircraft maintainers are seeking predictive machine learning methods to enhance aircraft maintenance readiness. Currently, maintainers must generate both handwritten and electronic records during their daily activities, ranging from manual entry of forms to transcribing the results into the databases. Valuable context to MRO operations is provided within this written documentation as well as maintenance manuals, and other documentation related to either the overall design, operation and maintenance of the aircraft. Today, these type of knowledge is consumed by human operators but can provide no assistance to predictive MRO initiatives without those humans. This Direct to Phase II SBIR proposes a software solution that can take human-readable documentation such as documents (whether PDF, scanned, or office formats) and maintenance records and extract a knowledge graph (a linked, graph-based representation entities and the relationships between them) from these documents in an automated way. This KG makes the information available to human maintainers as well as downstream systems such as UAV-based aircraft inspection.
