Z-BRE4K introduced a novel design for a predictive maintenance (PdM) capable of making accurate predictions for the future states of the components/machines/systems by the employment of intelligent simulators forecasting the generation of failures, estimating the remaining useful life RUL and triggering respective remedy actions.
During the project lifetime, Z-BRE4K solution passed through three stages of readiness levels (TRL5 to TRL7) at three end users’ industrial environments. In general, lessons learned can be summarized as:
Live data are gathered by sensors and other systems;
Data from individual systems are incorporated in a distributed system;
Quality and maintenance measurements are available;
Manual maintenance schedules are replaced with PdM procedures and schedules;
Maintenance experts are supported by gathered data and predictions to improve their know-how on the maintenance domain;
PdM accuracy and performance are established;
Productivity is improved;
Specifically, for SACMI-CDS, Z-BRE4K system gathered sensor data by adding additional sensors and condition monitoring solution where the system is distributed in the PdM that determines the RUL and machine failure. Main impact has been obtained on plant productivity and component’s management stating that it is of high importance to collaborate, not only with mechanical engineering and maintenance related professionals, but also with different technical background experts that together can improve multi-tasking, combining shopfloor and office-related activities.
The solution implemented in GESTAMP is an integrated system that exploits information on the shopfloor while connects MES and quality control system and sends data to the PdM to create predictions. The solution positively impacted at reducing breakdowns, machine downtimes, optimizing the working conditions on the shopfloor based on PdM that supported better understanding of GESTAMP´s reflection and readiness to apply PdM solution to its plants while new mitigation actions related to process flaws and defects identification were developed.
Finally, the solution on PHILIPS combines all separate data and gathers to the PdM of the RUL where the PdM outcome is sent to the decision support which creates a suggestion for the production managers or the tool workshop operators. It was interesting to evidence the impact where PdM helps to improve the uptime of their tools in the live production while reducing tooling costs, man-hours and unnecessary tooling parts stock. Lessons learned: “listening” and understanding the machines is the key for success as well as close contact between technology providers and experts where data integration/architecture and machine learning are both very important projects.
