Traditional maintenance concepts rely on a “fixed interval approach”, which takes into account a significant safety margin. As a consequence, maintenance is almost always taking place too early or too late, which makes it probably one of the most inefficient industrial activities and most critical at the same time
The CIMPLO-project aims at developing a cross-industry predictive maintenance optimization platform, which addresses the real-world requirements for dynamic, scalable multiple-criteria maintenance scheduling. Although the system’s capabilities will be demonstrated on our industrial partner’s application cases, namely aircraft engines (KLM) and passenger cars (Honda), it will be developed as a cross-industry platform for generic applications in predictive, multi-objective, dynamic maintenance scheduling. The goal is to provide industry with a tool for dynamically optimizing maintenance activities so as to:
- Reduction of turnaround time for maintenance.
- Reduction of maintenance cost.
- Improvement of maintenance staff satisfaction due to smoother, less stressful handling of dynamic change requests.
- Increased safety and improved spare parts management.
Three main components (data infrastructure, modeling engine, optimization engine) are developed to form the CIMPLO project. Through onboard sensor equipment, the amount of data available from equipment operations is extremely large nowadays, such that appropriate infrastructure for big data collection, preprocessing, and analytics is required. Next, algorithms for nonlinear, multivariate predictive modeling are developed to predict the amount of time remaining until maintenance is required for particular asset components. Moreover, to achieve the full business advantages in terms of safety and financial savings, the CIMPLO project proposes to combine predictive modeling for maintenance with dynamic multi-objective scheduling for maintenance, such that maintenance events and the required assets can be dynamically (re-)scheduled in a cost- and time-optimal manner.