Enabling Predictive Maintenance of Railway Switches
The customer
The company is the biggest rail builder and one of the largest construction and maintenance service providers. Founded in 1995, the organization has offices in six cities and employs 2,500+ people.
The need
When the customer turned to Altoros, it had a legacy system for railway switch life cycle management. With no analytic capabilities, it was impossible to monitor parameters key to switch maintenance.
Collaborating with Altoros, the customer wanted to enable data-driven analytics and ensure predictive maintenance.
The challenges
Under the project, the team at Altoros had to address the following issues:
- It was almost impossible to identify a conventional name of a particular switch, as there was no naming standard across different railway traffic control systems.
- With scarce data scattered across multiple sources, it was still crucial to calculate cargo weight carried by freight passing through railway switches.
- The system needed to be aware of train routes and which train was splitted where.
The solution
Engineers at Altoros enabled the system to accommodate all the variations of conventional names coming from multiple sources and identify the switch those names were associated with.By integrating a data mining tool that scrapes through 12,000 text documents and extracts the relevant data, our developers ensured proper cargo weight calculations.Based on the information received from the governmental traffic management system, experts at Altoros made it possible to track each train passing through the switches and their routes. Using ArcGIS, the team at Altoros developed a module responsible for extracting train data (type, number, etc.) from available sources. With QGIS, our engineers visualized this data in a dashboard. Furthermore, developers at Altoros delivered a dashboard view of a switch maintenance status with a detailed report of key parameters monitored.
The outcome
Cooperating with Altoros, the customer developed the solution to manage railway switch life cycle. Through enabled analytics, the system ensures predictive maintenance across 2,420 railway switches in 4 regional areas.
Technology stack
Platform
Microsoft Azure
Programming language
R
Frameworks and tools
Microsoft.Maps.SpatialMath, ArcGIS, QGIS, PostGIS
Databases
PostgreSQL
Contact us
Jan-Terje Nordlien
Daglig leder
jan-terje@altoros.no+47 21 92 93 00Altoros Norge AS
Org.nr.: 894 684 992
Tordenskiolds gate 2,
0160 Oslo