Real-Time Detection of Low Altitude Icing Conditions

Our proposal addresses a pain point in the aviation sector, especially within Advanced Air Mobility (AAM) operations: a lack of high-fidelity detection of hazardous icing conditions. Leveraging novel weather sensors, including the Vaisala CL61, Meteomatics Meteodrone, and the Weather Scout UAS, alongside an atmospheric moisture classification ML algorithm, our project will seek to detect conditions conducive to icing and begin work towards an ML-based estimate of icing event likelihood. This will lead to development of a high-resolution short-range icing risk forecast model in Phase II, providing unprecedented safety and operational reliability in aviation, particularly in AAM corridors below 5,000 feet AGL. The technology’s purpose is to respond to client identified pain point for in-cloud icing. During a 2023 weather workshop by the non-profit AeroX, icing was identified as one of the highest impact weather hazards affecting the AAM industry. This finding was based on the risk to safety posed by icing as well as uncertainty in the timing and location of icing. Thus, the ability to safely operate AAM platforms in the national airspace system (NAS), efficiently, requires timely, high-resolution data to detect and avoid hazardous icing conditions. The funding will research sensors that can detect icing potential and develop a framework to produce ML algorithms to produce real-time icing risk assessments for the safe integration and operation of AAM systems in the NAS. The proposed assessments will serve the emerging $68.1B (by 2032) middle-mile drone and air taxi industries to reduce weather-related disruptions, both real and perceived, enhance flight safety, and increase operational efficiency. The objective is when icing products produced for the NAS today depict icing conditions, potentially stopping flight operations without knowing whether icing really exists, the TruWeather system will be able to report, with certainty, either icing does, or does not exist.