SBIR/STTR Award attributes
Reliability and availability are critical to the availability and operation of U.S. Navy aircraft. Time based maintenance have been the mainstay for most naval assets, however a shift toward evidence driven predictive maintenance using sensors and product level models could improve efficiency and reduce total ownership costs. One of the biggest threats to aircraft availability is corrosion, which serves as one of the leading maintenance costs for military aircraft. New sensing technologies are emerging that provide maintainers with local environmental measurements and other parameters that relate to the propensity for atmospheric corrosion to develop in compartments and locations that are difficult to access or visually inspect. This data is currently being recorded and downloaded for offline analysis by subject matter experts, resulting in a largely manual process. Luna proposes to develop a digital twin that serves as a virtual model for airframe corrosion using laboratory measurements and aircraft data that has been collected through previous instrumentation efforts using Luna’s embedded corrosion monitoring sensors on Navy and Air Force aircraft. These models will combine feature selection / extraction techniques with machine learning algorithms to automate the analysis process and generate a model for the corrosion state based on onboard measurements. This virtual representation of the airframe corrosion hotspots will provide maintainers more timely, actionable information and metrics that will improve predictive maintenance strategies. Data features and metrics will also be used to identify corrosion trends across multiple aircraft when compiled in the maintenance and logistics information systems to identify drivers that accelerate corrosion and impact fleetwide availability for different naval aircraft.
