We propose to develop an enterprise software system specifically designed to support visualization, annotation, and analytics of aircraft inspection data. The system provides a range of tools and automation between raw captured data and the back-end repository. The tools provide a convenient way for the human operator to stay in the loop, visualize the collected data in powerful ways, provide human knowledge by annotating the data, and flag corrective actions. The annotated dataset produced by our proposed system is a necessary training set for creation of certifiable automated data analytics algorithms. Such datasets are completely missing today; our system will capture them over time as Air Force personnel codify their knowledge into it. The emerging data collection systems, such as autonomous drones, capture geometric metadata, which provides where an image was taken from and what the field of view was on the inspected asset. These powerful datasets can be fully leveraged within our proposed system to exploit additional dimensions of captured data and enable anomaly tracking per location. The proposed system is seen as bridging the present gap immediately, while evolving into a highly automated and intelligent tool over time.
