Air Canada is well known as an industry pioneer and that drive extends well beyond flying millions of passengers across the globe. The airline is increasingly focused on developing its in-house artificial intelligence expertise and recently a team from Air Canada’s AI team took home top honours at the IATA Aviation Datathon in Athens, Greece.
A team of talented Air Canada data science professionals took up the challenge of developing a safety incident classification system based on safety incident report data.
The process of classifying safety incident reports is largely a manual and time-consuming process. The manual nature of this process can lead to errors and inconsistencies. The goal of the challenge Air Canada undertook was to leverage machine learning approaches in natural language processing to develop an accurate, automated and cost-effective incident report classification solution.
The team’s solution was deployable, required minimal data processing, leveraged machine learning and impressed the judges over all other solutions.
“The IATA Datathon win is not just a testament of a strong solution to the challenge presented, but it also speaks to the team’s energy, drive, and cohesiveness. A global IATA award is always a remarkable feat, but when you consider that the team has been working together for a short period of time, it makes it even more impressive and validates that we are on the right path for our AI journey,” said Catherine Dyer, Senior Vice President and Chief Information Officer at Air Canada.
The IATA event was focused on addressing industry challenges and priorities, and proposing solutions leveraging data science capabilities in areas where data is already available and accessible. The objective was to raise awareness and trigger innovative ways airlines and the wider aviation industry can extract value from data – something Air Canada recognizes as a key priority.
The team’s innovation at the IATA event will lead to important advancements at Air Canada, with uses that go well beyond incident report classification to include customer service applications such as service request classification and social media sentiment analysis
