• The Swiss Federal Railways (SBB) would like to further reduce the effects of disruptions to infrastructure and rolling stock and ensure smooth operations 

• The machine learning solution developed with Zühlke enables SBB to become aware of possible punctuality risks in the operating network at an early stage 

• This enables SBB to increase customer satisfaction and save costs 

SBB wants to detect the impact of operational malfunctions more quickly and thus optimise train punctuality. A machine learning solution developed by Zühlke is helping achieve this.

About 1.25 million passengers use the Swiss railway network every day. The Swiss Federal Railways (SBB) want to further reduce the impact of malfunctions in infrastructure and vehicles in order to ensure smooth operation. Big data is now being used to optimise staff planning, infrastructure maintenance and train punctuality. Zühlke has joined forces with SBB specialists to develop machine learning algorithms and deploy them on a big data platform. 

Zühlke data scientists are supporting SBB specialists on site. Together, the team is drafting a proof of concept and developing the architecture of a machine learning solution for operative implementation. This connects to data such as trouble records and weather forecasts and is then used to train the machine learning algorithms. The result? An effective early warning system that never stops improving. 

Thanks to Zühlke’s machine learning solution, SBB is made aware of potential network delays in the upcoming weeks. This allows them to take the required measures early on in order to optimise punctuality. As a result, SBB can increase customer satisfaction and save costs.