Problem

During the first COVID-19 lockdown in April 2020, as travelers gradually returned to public transportation under new restrictions, maintaining safe distancing on trains became a major concern.

Uncertainty about train capacity and potential crowding at stations made many travelers feel unsafe and hesitant to use the trains, significantly contributing to a decline in passenger numbers.

Solution

Each S-train is equipped with weight sensors that measure the train's load. Previously, DSB had only used this data internally to calculate braking distances.

We leveraged this data with machine learning to estimate passenger numbers based on the train's weight, providing a reasonable estimate of how many people were on board.

In just three weeks, we developed and launched a mobile site that displays this information in real-time for passengers. They can use it at the station to decide whether to wait for the next train or follow their preferred train's expected capacity. Additionally, the site provides forecasts to help passengers plan the best times to travel.