We propose using an integrated, multi-layered sensing approach for rapidly assessing runway damage (i.e., craters, spall, debris, and unexploded ordinance [UXO]). Within this approach, a fixed wing UAV will be tasked to fly the entire area of interest. Through edge computing and real-time processing of images the system will assess any specific locations in need of more detailed inspection. Current state-of-the-art UAV remote sensing relies heavily on stitching and creation of a single orthomosiac image prior to further analyses. We propose processing on an image-by-image basis, thereby obtaining actionable information in near real-time with the same effectiveness and accuracy. The results from initial fixed wing UAV sensing flights will direct a multi-rotor UAV with more specialized sensing equipment (i.e. RTK GPS, time of flight (TOF) array ranging and hyperspectral sensing) to more thoroughly inspect regions of interest. Results from the multi-rotor UAV flight will be transmitted to an Autonomous Ground Vehicle (AGV) for further action such as surface profiling further reducing the time to Minimum Airbase Operating Status (MAOS). The fixed wing UAV will allow for a quick overall assessment of the entire runway and surrounding area whereas the multi-rotor UAV and AGV will provide give in-depth information about detected.Runway Inspection,machine learning,UAV,precision ag,beyond line of sight,Multi-Heterogenous Vehicle Control
