Smart maintenance with the ITK Data Lab Framework

The challenge: Enable smart maintenance for train door monitoring systems

A previously implemented remote monitoring tool for train door systems retrieves diagnostic messages. Based on this data, evaluations, data extraction and visualizations are created. The smart assessment of millions of historical data points goes to make maintenance and scheduling more efficient and provide the underpinning for predictive maintenance.

• What potential can be exploited with the diagnostic data?
• What use cases can it serve?
• Is the available diagnostic data suitable for predictive maintenance?
• Would predictive maintenance require further diagnostic or maintenance-related information?
• How does the potential benefit compare to the required effort?

The solution: Fast results delivered by the ITK Data Lab

The ITK Data Lab served to quickly pinpoint and specify the potential in the data pipeline. The team conducted an exploratory analysis and formulated hypotheses. It takes the business impact into the equation to identify several use cases, two being the defect detection tool and predictive maintenance. 

Use case: A defect detection tool
Using data sourced from train doors markedly increased the defect detection rate and the doors’ uptime. The ability to automatically detect defects makes scheduling maintenance easier. This primarily increased the availability of the doors. In addition, the automatic detection of defects improved the predictability of maintenance for operators.

The added value: Ability to quickly identify and efficiently tap data potential

The pre-developed Data Lab uses artificial intelligence for initial analyses to rapidly pinpoint potential on the path to smart maintenance. Its modular structure and highly automated routines deliver fast results with short runtimes and at low cost. In addition, work was carried out according to a tried-and-tested structured procedure, so that value is added as quickly as possible.