The German operator DB Cargo has introduced an artificial intelligence-based system to plan the spare parts needed for the maintenance of Class 77 locomotives, with the aim of reducing vehicle downtime.
The project uses data analysis models to anticipate spare part requirements and avoid situations where locomotives are unavailable due to a lack of components.
AI Model for Spare Parts Forecasting
The system was developed at the DB Cargo Railport Darmstadt logistics center and combines historical data on parts consumption with information regarding:
- locomotive mileage;
- maintenance intervals;
- the context of maintenance work in depots.
The fleet includes approximately 60 diesel locomotives used on non-electrified lines. Since the locomotives were built in Canada, some spare parts may have delivery times of weeks or even months, making traditional inventory planning difficult.
Example: Forecast for Oil Pumps
An example presented by DB Cargo is the case of oil pumps for Class 77 locomotives.
Traditional forecasting methods did not indicate a need for new parts, while the AI-based model estimated the requirement at five units, very close to the actual consumption of six parts.
Given delivery times of approximately 500 days, such a difference can determine whether a locomotive remains out of service or can return to operation quickly.
Optimized inventory planning
According to DB Cargo, the system allows for differentiation between expensive parts with long delivery times—which must be kept in stock—and easily procured components, which can be ordered as needed.
The company states that the use of data-driven analytics helps reduce maintenance bottlenecks and increase locomotive availability.
Share on:
