Noise and wear reduction for trams with AI-based rail head lubrication
Customer: BERNMOBIL, Switzerland
Project tasks
- Project management
- Concept development
- System integration
- Data analysis
Our approach
The operation of a railway vehicle causes noise and wear on wheels and rails. These issues can be reduced by lubricating the wheel-rail contact point. However, excessive lubrication can pose a safety problem, as it reduces traction forces and increases braking distance, potentially causing ecological pollution.
Together with our partners, we have developed a system as part of a research project supported by the Federal Office of Transport (FOT) that makes it possible to condition the rail head as needed. Sensors installed on the vehicle measure status parameters and feed them to an AI installed on the vehicle. The machine learning model analyses the data, determines the need for conditioning and synchronises this with other vehicles operating on the rail network. Based on this information, each vehicle decides for itself at which point in the rail network conditioning must be initiated by the on-board lubrication system.
Customer benefit
In addition to fully automatic lubrication, the customer has access to comprehensive noise monitoring, which allows sources of noise to be localised at an early stage if they have not already been completely eliminated by the system. Since relevant influencing variables are known through the machine learning model, most noise sources can already be detected predictively and avoided altogether.
Communication between the vehicles creates a form of fleet-wide intelligence, enabling complete coverage of the fleet and the entire route network without having to equip each individual vehicle with expensive measurement technology. Thanks to the measuring system installed, other conditions can be monitored in addition to noise, such as ride comfort, wear between wheel and rail, and the infrastructure.
The customer benefits from our deep understanding of wheel–rail interaction and tram system behaviour, as well as our hands-on technical engagement throughout development.











