Automation of Aviation Weather Charts using Machine Learning Techniques
2022-26-1231
05/26/2022
- Event
- Content
- In traditional avionics procedures, manual analysis of aviation weather charts is often time consuming and increases the pilot's workload significantly. This manual analysis plays a key role in identifying the weather phenomena, which an airborne weather radar cannot detect such as Clear Air turbulence (CAT). The paper proposes introduction of Machine Learning (ML) algorithms to decode the pictorial representation of Significant Meteorological Information (SIGMETS), and significant weather charts during take-off procedures and textual representation of METER and TAF during in-flight later on. During take-off procedures, ML algorithms are provided with the flight path for the aircraft under consideration, which is about to fly, SIGMET, 14 different routes of aircraft which are about to fly and significant weather charts. CAT on a significant weather chart, turbulence is highlighted by the area surrounded by dashed lines. Area of possible CAT due to close proximity of isotachs can be identified through 300/200 mb chart. The proposal includes the application of significant weather chart as well as 200 mb chart to increase precision of correctness in identifying CAT. The ML algorithm compares the chosen flight path with weather locations that can cause discomfort and examines other aircraft routes and finally proposes a best possible flawless flight path to the pilot. The onboard pilot can then verify the proposed flight path and if it is satisfactory, the proposed flight path is uploaded. Each keyword in METER and TAF is predefined. This manual analysis will be decoded by replicating the pilot's manual analysis through ML Algorithm. The pilot will be dynamically alerted through speakers if the hazardous weather is encountered as an outcome of METER and TAF automation. There is a significant scope to introduce dynamic rerouting facility through the proposed automation of aviation weather charts.
- Citation
- Gangadhar, B., and Ramasamy, S., "Automation of Aviation Weather Charts using Machine Learning Techniques," SAE Technical Paper 2022-26-1231, 2022, .