Instruments & industry / Project story
Predictive Maintenance System
Condition monitoring for industrial gear units, combining sensor data, neural networks, and an application for planned maintenance.
01 / Starting point
About the project
The system captures vibration, temperature, oil, and load data from large industrial gear units. Analysis models detect unusual patterns before damage becomes visible during operation.
02 / Constraints
Requirements
A model finding is useful only when a maintenance team can interpret it and fit it into an operational decision.
03 / Implementation
Implementation
I worked on data acquisition, analysis logic, and the application used by maintenance teams. Dashboards, trends, and configurable alerts make model findings understandable and support maintenance planning.
The platform connects C#, Angular, and Python with cloud services and existing production systems. The key challenge was not only prediction, but its dependable integration into operational workflows.
04 / Current state
Dashboards, trends, and configurable alerts bring the analysis into the application used by maintenance teams. My contribution covered acquisition, analysis logic, and the operator-facing application.
Project images