The Smart Locomotive is based on the proprietary platform Clover that is designed to process and analyze large volumes of industrial data. Based on this platform, a solution titled Ctrl@Maintenance was created—an intelligent system for risk-oriented asset management and maintenance and repair planning. The system comprises a repository of rules and mathematical models (MathExperience).
The system already analyzes:
-
59 workshops
- over 8,000 locomotive sections
- 1.5 TB of data every month
How does it work?
All trains are equipped with sensors and microprocessor control systems that register locomotive equipment status data.
What for and how are data processed?
Every time a customized locomotive enters the workshop, telemetric data recorded on the way are read from the onboard system. Then, the operator uploads those data to The Smart Locomotive and the program starts converting and processing information automatically. After that, algorithms for identifying inoperable sensors initialize. On average, most failures are identified a month before they occur.
Results:
- 3x lower breakdown maintenance costs
- 12% less locomotive downtime
-
locomotive diagnostics take 20 minutes rather than 4 hours
Technology stack
Predictive analytics, artificial intelligence, neural networks, Internet of Things, big data analysis.