Project Overview

The QUISS project’s goal was to create data-driven applications using contemporary data science and artificial intelligence methods. These applications were designed to enhance or assist in the planning and upkeep of rail vehicles. To achieve this, a system was established that integrated various data sources, including freight wagon shock sensors, camera bridge images, GPS location information, and assorted inventory data.

Utilizing this data repository, the project was able to establish and experimentally integrate different scenarios for improved and data-centric maintenance.

This project was funded my the Federal Ministry of Transport and Digital Infrastructure (Germany) between 2018 and 2021. It was a collaboration between DB Cargo AG, Inspirient GmbH, and the RWTH Aachen University.

The implementation of these scenarios led to notable enhancements in the availability of rail vehicles. Thanks to the camera bridge image data, it’s now possible to spot vehicle anomalies earlier, allowing for timely maintenance.

Enhancing the freight wagon telemetry data with semantic details has increased the precision of estimated mileage, thereby extending the operational lifespan of the rail vehicles. Additionally, combining GPS data with route network information has improved the accuracy of tracking the vehicles’ routes.

Impact data is also being used to pinpoint high-stress sections of the railway, helping to minimize vehicle damage. In a separate scenario, AI was used to analyze the relationship between vehicle mileage, workshop durations, and previously gathered insights.

The rail transport sector stands to gain significantly from these project outcomes, as they enhance the scheduling and maintenance processes for vehicles and diminish operational disruptions.