Production System Opportunistic Maintenance Planning

This tool has the objective to provide (i) Condition-Based Maintenance capabilities to manufacturing system users, (ii) combined with the ability to select the best time windows (opportunities) to carry on preventive maintenance without interfering with the achievement of the production plans. The tool is composed of three major modules: 1. Data gathering from sensorial data: this module receives in input data from the multiple sensors implemented at shop floor level to observe the degradation behaviour of the critical components in the system. In RobustPlaNet, a subset of signals and the related sensors has been selected for the interested use case. Then, these signals are elaborated and associated to states of the monitored machines. Machine degradation modelling: This module takes in input the association of signal patterns to degradation states and initial sample data obtained from the shop floor. It provides in output a Markovian state-transition model that represents the dynamics of the machine while visiting its different degradation states [16] 2. Manufacturing system model: this module integrates the developed resource degradation models into a production system model (including finite capacity buffers, part routings, cycle times). It provides in output estimates of the main system KPIs (throughput, lead time, WIP) under specific maintenance policies. 3. Opportunistic Maintenance Policy Optimization: based on the previous model, this tool applies optimization algorithms to select the best possible opportunistic maintenance policy that look both at the degradation state of the specific machine and the state of the remaining resources in the system. The output of this tool will be the derivation of a specific opportunistic maintenance policy to be implemented at each monitored machine for improved robustness of the system.