Alert, Your Trip Will Be Crowded: Crowding Prediction at Västtrafik

When the Corona pandemic hit the world Västtrafik wanted a precise way of informing travelers on what trips that were likely to be crowded so the travelers could plan their trips in a safe way. Since Västtrafik is working in a large geographical area, and a heterogenous population landscape, creating warnings only based on time of day wasn’t considered to be good enough. We will present a use case of machine learning in production based on data from the automatic passenger counting system. Focus will be on all those details that makes an ML solution useful live in real life; How the results are communicated, how it is maintained and what model is used in the engine. We will also present results on how it was received by the travelers and how it is continuously evaluated and evolving.