Although aviation technology continues to advance, producing more modern and reliable aircraft, there remains one unpredictable factor that can cause accidents: Human Errors.
Aviation is one of the safest modes of transportation. However, increasing air traffic in civil aviation has added complexity to safety management, requiring strategies to evolve alongside growing operational demands. While aviation accidents can be the result of an assortment of factors, approximately 70% are attributed to human error.
Project approach:
Within DEMOQUAS, the team from Aristotle University of Thessaloniki (AUTH) are developing advanced tools to contribute to aviation safety through Artificial intelligence and Machine Learning techniques, in collaboration with Egnatia Aviation (EGA). More specifically:
- A Bayessian-Network-based framework to proactively predict and diagnose pilot performance. The model simulates factors that can impact the capacity of a pilot, to enable the identification of influential parameters and allow for proper training and mitigation actions.
- A Natural Language Processing model to analyze aviation safety reports, to rapidly handle large amounts of data for effective report classification and cause detection identification.
Through these models, DEMOQUAS can quantify the uncertainties arising from human activity, enhance aviation safety, and contribute to accident prevention.
