Astronics
Make the Most of the Data You Produce
This article first appeared in the Railway-News magazine, Issue 3 2024.
Train subsystems have a lot to tell you – it’s all there in the test data you’re producing, writes Bagali Siddaram, Director of Mass Transit at Astronics.
The rail industry is becoming ever more digitalised, including our specialist area of test, repair and support. This is becoming increasingly software-intensive, as onboard systems as well as trackside and wayside equipment continue to become ‘smarter’.
By ‘instrumenting up’ a train, which involves fitting a wide variety of sensors to these advanced electronics, these subsystems are providing operators with a rich source of information on the health of their equipment.
Do More with What You Already Have
Today transport operators have data from their repair facilities and processes in droves, but not everyone is making the most of this information goldmine.
The data’s there for the taking, it’s just a matter of extracting and making sense of it, and if used correctly, can provide some significant financial and operational benefits.
For example, analysing your data can help you to understand whether fleet-wide maintenance is necessary, as well as better predict when such activities will need to take place.
Predictive Maintenance Solutions
This is made possible through the use of mature artificial intelligence (AI) and machine learning (ML) technologies, which may be simpler – and cheaper – to implement than you might think.
There are two main pathways you can take. The first analyses the large sets of data you’ve already produced and gathered from testing your subsystems. By doing so, it teaches an AI or ML algorithm how to uncover trends, spot anomalies and predict component lifecycles. Plus, as it’s continuously fed more information, it becomes ever more accurate.
