The use of big data in maintenance
TechValley is a partnership between innovative North Holland machine builders, including Total Productivity.
TechValley focuses on 11 concrete innovation projects of participating companies. One of these is predictive maintenance.
With predictive maintenance, companies perform maintenance at exactly the right time. Predictive maintenance increases the reliability of infrastructure and production resources and reduces the costs of maintenance.
Or to put it more simply: Better maintenance and fewer malfunctions at lower costs.
Forecasts based on data analysis
Maintenance forecasts are based on logging and analyzing data over a longer period of time (big data). Partly for this reason, you cannot implement this maintenance approach overnight.
Joery Brown, who is a member of this project on behalf of Total Productivity, will set up a model that can be used for any machine through practical research at one of our partners.
Data collection and analysis
In this project, Predictive Maintenance and increasing the Overall Equipment Effectiveness (OEE) are examined in 4 areas. We use the following data for this:
- Failure data (which is already generated).
- Real-time data from the sensors/actuators, such as motor current, force, temperature and power.
- Real-time condition monitoring. Warnings are given based on a pre-set critical level.
- PdM 4.0. Continuous real-time monitoring where alerts are based on predictive techniques, such as data consistency analysis.
At the moment we are still in the data collection phase. This project is expected to take approximately 3 years to complete.