Pilot Reports (PIREPs) are reports submitted by pilots describing in-flight weather conditions that provide crucial weather information to other pilots for pre-flight and in-flight planning. The current PIREP system is antiquated and has been identified as a safety concern, according to a 2017 National Transportation Safety Board (NTSB) Special Investigation Report (NTSB, 2017). The NTSB report and focus groups conducted by the Civil Aerospace Medical Institute (CAMI) Human Factors Division (Kratchounova, 2020) consistently identified the most problematic elements of the PIREP system as it exists today.
Two broad categories of issues that reduce the effectiveness of PIREPs identified in the report included submission and dissemination issues. Specifically, submission deficiencies included pilots providing relatively few PIREPs (particularly during good or as forecasted conditions) and air traffic controllers needing to solicit PIREPs consistently during weather conditions that mandate such service. The dissemination problems are associated with the Air Traffic Control, Flight Service Stations, or company personnel handling PIREPs introducing delays and errors, or even failure, to distribute the information.
The CAMI Human Factors Division conducted a proof-of-concept study between November 4, 2022, and April 30, 2023. The study successfully demonstrated the application of already existing and state-of-the-art off-the-shelf technologies as one potential way to modernize the current PIREP system. More specifically, while using the voice communication system found in most aircraft today (a VHF radio), it explored a system that automatically captured, queried, and transcribed weather information submitted by pilots in flight without talking to an air traffic controller or a flight service specialist.
Researchers collected data within two geographic areas – one in Oklahoma and one in Alaska, where they assigned discrete VHF radio frequencies for the study duration. The study team gathered nearly 5,500 PIREP submissions and retrievals using voice recognition technology and stored and processed them using cloud computing. During flights, the team disseminated PIREPS via VHF radio, while on the ground, they used web-enabled and mobile apps for distribution.
Although operational issues occurred relating to the nature of a low-power radio ground station, challenges with translating aviation-specific language using off-the-shelf processing tools, etc., were implemented; participant feedback was used during the study, allowing for continual improvements in multiple areas.
By the end of the project, ratings of the concept’s feasibility, utility, and usability and the extensive written feedback indicated high satisfaction levels while highlighting areas that needed further improvement. In summary, this research showed that the approach chosen for this study is feasible and demonstrates good utility and usability.
The results from this proof-of-concept study suggest that significant benefits could be derived from conducting additional research to identify and assess the impact of further automating those historically problematic elements. For example, a future study could investigate utilizing ADS-B data to autofill information such as position, altitude, type of aircraft, etc., and allow the pilot to a) focus on reporting weather observations and b) do it more efficiently and effectively. Furthermore, investigating the contribution of this or similar concepts to optimize pilots’ and air traffic controllers’ workload profiles could provide insight into the broader positive effect of minimizing or eliminating the potential of human error in the PIREP submission and dissemination process.
In the context of an integrated system architecture that preserves the human element (i.e., the direct observation of weather conditions made by a pilot in-flight), a potentially viable approach could be to utilize generative artificial intelligence (AI) to transcribe the audio and deconstruct the voice-to-text transcription into plain text. This approach could lead to an optimized human-system integration. If funded, empirical data from the next phase of this research would determine if, as compared to the currently existing system, this concept: a) increases the number and quality of PIREPs, b) optimization of pilot and controller’s workloads, c) improvement in the PIREP system resiliency and d) the ability to impact aviation safety positively.
