Optimizing geometry monitoring with the help of AI and IIoT

Case Study, 12.04.2022

In 2012, Oc’via and SNCF signed a public-private partnership (PPP) for the construction and maintenance of the new Nîmes and Montpellier rail bypass (called CNM)—France’s first mixed line (high-speed trains and cargos).

Oc’via Maintenance, Oc’via’s service provider, has been appointed to monitor, maintain (preventively and correctively) and renew the CNM’s installations until 2037. The company is involved in all the assets of the 80 km line: track, catenary, signalling, telecommunications, buildings, and engineering structures.

The Challenge

Oc’via Maintenance utilizes an inspection wagon, the WIN, which monitors platforms and railway installations thanks to its various sensors. This machine allows monitoring of track geometry to ensure traffic safety and passenger comfort.

Oc’via Maintenance, therefore, aims to regularly operate on the CNM, monitor any existing defects, examine their development, and detect (and treat) new defects as quickly as possible.

This approach, however, has two major limitations: the cost (frequent recordings allow optimal monitoring of the defects but generate significant costs) and the inability to identify the root cause of the defects (the wagon only measures values and does not provide additional context on causal factors).

Example of a geometric defect on the CNM

The Solution

Oc’via Maintenance’s strong dedication to digitalization led it to explore a partnership with KONUX. One of the projects of this partnership had the goal to complement Oc’via Maintenance’s current regime with continuous track monitoring provided by The KONUX Predictive Maintenance System. Oc’via Maintenance equipped the tracks with the KONUX IIoT devices in two of its most problematic areas.

To tackle this complex problem efficiently, Oc’via Maintenance utilized the data from the IIoT devices to extract insights that they can immediately act upon. By analyzing the correlation between the sudden changes in tilt data and several contextual events, KONUX was able to detect the events resulting in the geometry defects that occurred between the two WIN train passages.

The Results

By taking the “quick-win” approach to get actionable insights, Oc’via Maintenance was able to have immediate and tangible benefits in detecting and diagnosing geometry irregularities.

Oc’via Maintenance had an initial hypothesis that the track geometry issue originated from rising groundwater levels. By cross-referencing the KONUX data with various metrics such as precipitation, temperature, and potential traffic changes, Oc’via Maintenance was able to factually support their hypothesis.

The estimation based on this project indicates that if this approach can efficiently identify the root cause of geometry defects and lead to its resolution, Oc’via Maintenance could save around three tampings per area per year. Also, the adoption of continuous monitoring can support track quality without the need to increase the frequency of WIN. Based on this experience, KONUX has launched a research project to further develop the use case for geometry health monitoring. This use case will enable evaluating the quality and durability of maintenance actions not only for improving vertical displacement but also for geometry issues. Furthermore, this initiative will continuously provide Oc’via Maintenance with complementary information regarding geometrical defects and enable them to act at the right time.

Would you like to learn more about the case? Get in touch with our team here.