To Develop a Machine Learning Assisted Pilot Debrief and Assessment System.
Description
- Pilot debrief is based on memory of instructor and his ability to identify mistakes. Instructor takes assistance for tools like Flight Data Recorder (FDR), video playbacks and other tools for preparation of pilot debrief. This activity is time consuming and subjective. In the existing debriefing systems, human bias and personality traits which vary from instructor to instructor may transpire leading to loss of objectivity in pilot assessment.
- An automated system to understand policies, instructions and flight manoeuvres expectation and classifications based on comparison with actual pilot ideal missions to enable rapid analysis using machine learning and preparation of personalised debrief of individual pilots is required to be developed.