Flying today is safer than ever, thanks in part to the high standards of communication between the parties. One communication system is called “NOtice To AirMen” (NOTAM) and consists of short text messages broadcast by airspace officials to warn pilots of anticipated events that may disrupt flight en route (closed runways, construction work, closed airspace, etc.).
Unfortunately, thousands of messages are sent every day and their number is increasing very rapidly. One of the reasons for this rapid increase is a lower and lower “safe threshold” that triggers a new NOTAM, sometimes leading to irrelevant NOTAMs. As a result, each pilot has to search and sort through an increasing number of messages, resulting in an increased loss of time and an increased risk of neglecting important messages. In major airlines, a “NOTAM officer” is responsible for pre-screening NOTAMs before issuing them to pilots.
I worked with the data science team of SWISS international airlines to develop an automated NOTAM classifier. The goal was to save time while keeping a high level of safety. It was my capstone project that validated my data science training at Propulsion Academy.
The full code of the project can be found on GitHub. The slides of the final presentation are here, and below is the recorded talk: