top of page


Image of Project

GenSquared, a data and analytics company, was working with their client Bombardier to build a data and analytics pipeline for the firm. Bombardier came to them with a challenge - what kind of artificial intelligence could be applied to these airplane maintenance records in order to understand better what is going on? Lacking the deep artificial intelligence expertise, GenSquared in turn came to Bradley Arsenault to prototype deep learning models to these records.

To analyze this dataset, Bombardier had developed a standardized system of Failure Codes that each maintenance record could be classified as. This allowed Bombardier to track the trends of various maintenance issues over time. However, applying these standardized failure codes was a costly and error prone endevour.

Bombardier wanted to see if deep learning technology could be used to automated the classification of failure codes, and sure enough it could. Brad developed a deep-learning based classifier based on cutting edge AI research to automate the classification process, which was subsequently built into the data & analytics pipeline developed by GenSquared.

The ability to get accurate failure statistics over the entirety of the maintenance records revolutionized the maintenance and repair engineering department at Bombardier, leading to a variety of new insights on how to improve the engineering of their C-series airplanes.

bottom of page