The Whirlpool white goods use case deals with an automatic production line, which produces drums for clothes dryers. The drum production equipment is very complex, highly automated and critical from many perspectives. Currently, preventive and reactive maintenance is implemented and although needed, they exhibit many drawbacks. The reactive maintenance is performed during operation time, causing either a stop of the production or a degradation of it. The preventive maintenance is performed to reduce unwanted breakdowns as much as possible, which can improve the Mean Time Between Failures (MTBF). However, it can be very expensive, impacting the Total Cost of Maintenance (TCM). The Whirlpool use case aims to investigate how Whirlpool can make an effective transition from preventive maintenance to predictive maintenance.
The Whirlpool use case starts with the acquisition of historical and operational data and the analysis of machine failures and maintenance actions to learn from experience and understand maintenance knowledge that relevant data from legacy and operational systems may reveal. The analysis is performed by UPTIME_ANALYZE, combined with the failure modes details detected by UPTIME_FMECA. Accordingly, new additional equipment sensors have been installed into the Whirlpool drum production line, which allows further enrichment of the dataset by UPTIME_SENSE, to continue the evaluation of machine learning algorithms by UPTIME ANALYZE, provide forecasts of future events by UPTIME_DETECT & _PREDICT and build basic rules for establishing health status of the equipment by UPTIME_DECIDE. UPTIME mobile application is available to be used by the factory workers to see generated maintenance recommendations and decide on maintenance actions.
The capability of the UPTIME system to predict future failures of the drum production line and to give indications about prognostic measures will modify the preventive maintenance plan allowing Whirlpool to anticipate planned intervention on components, thus reduce unexpected breakdowns, delay other interventions, and accordingly save money. The MTBF is expected to increase, since some unforeseen breakdowns will be predicted by the system, allowing maintenance to act before the component breaks. Moreover, the Mean Time to Repair (MTTR) is expected to decrease, due to the maintenance activities that will be planned, and thus optimising equipment downtime. All these effects will also positively impact the Total Cost of Maintenance thanks to optimized management of spare parts, technicians scheduling and improved effectiveness.
This items serves as a filter in support of selecting the case and demonstrators associated to the Digital Transformation Pathway Cases Catalogue (see ConnectedFactories Coordination and Support Action - Information sharing and analysis)
Quantity and quality of data need to be ensured from the beginning of the process. It is important to gather more data than needed and to have a high-quality dataset. Machine learning requires large sets of data to yield accurate results. Data collection needs however to be designed before the real need emerges. Moreover, it is important having a common ground to share information and knowledge between data scientists and process experts since in many cases they still don’t talk the same language and it takes significant time and effort from mediators to help them communicate properly.