Improving efficiency and safety in aviation industry using big data analytics (phase II)

Aviation industry uses flight data recorders (FDR) to monitor a high number of parameters during each flight it operates. It is expected that analyzing this data will provide useful information to airlines for improving flight safety and efficiency. However, this analysis is a challenging task in itself because the amount of accumulated data is enormous and also because it is diverse. To overcome these difficulties, data is first preprocessed (or cleaned) and only significant parameters are kept. Then, neural networks are used to model the relation between these parameters and find optimal values. Finally, clustering is used to group similar flights together and the properties of each cluster are analysed to explain how they influence flight safety and efficiency.

Intern: 
Jérémie Villeneuve
Faculty Supervisor: 
Richard Labib
Project Year: 
2015
Province: 
Quebec
Partner: 
Discipline: 
Program: