Review of online Predictive Maintenance knowledge days

Data is the basis for predictive maintenance, but a proper maintenance strategy and software are necessary to make this maintenance method successful. During the two-day online sessions At Predictive Maintenance, an event of the Industrial Automation sector, the subject was approached from various angles.

Erik Daalder & Robin Daalder, Schneider Electric, kicked off the ball on Tuesday. In their presentation, they provided a nice insight into how asset performance management can not only achieve cost reductions, but can also contribute to the sustainability objectives of companies. Kees Lambregts of Novotek explained in his webinar a use case of a large water board that can prevent malfunctions in its pumps using predictive maintenance. The emphasis was on the question of which data is most useful to analyze.

Jos Geers of SAMSON REGELTECHNIEK spoke under the title 'Begin, learn and grow'. He discussed the most important trends in the process industry and that plant managers with the right software gain insight into trends, concrete root causes and an early insight into possible deviations.  Jules Oudmans (UReason) gave a lecture entitled 'from asset data to work order'.  He gave a closed loop demo and told his listeners: “listen to the data your assets generate”.

The first theme day was concluded by Wiek Wijands (Kensan). He provided a webinar on behalf of Weindmüller and discussed a windmill case. He showed how a predictive warning system based on machine learning looks at deviations and possible disruptions in them. Wijnands emphasized that this requires relatively little data.

Day two

On Wednesday, Michiel Dondorp and Anish Bisnajak of Croonwolter&dros were the first to speak. They outlined the importance of a good data scientist and the challenges that companies face when implementing predictive maintenance. Diederick Nab, WAGO Netherlands, then addressed the question of what role cloud services can play in a successful implementation of predictive maintenance.

Marcel Kelder (Yokogawa) provided insight in his webinar on how asset information can be made more transparent using machine learning. The lecture by Omar Thiam, Wonderware, also covered the benefits of artificial intelligence and the different models that AI uses, such as supervised and unsupervised machine learning. He also indicated how AI can be used to discover patterns in order to prevent downtime in process installations.

During the session of Emerson, by Sabine Spapen & David Miny, vibration analysis technology in rotating machinery was discussed. By capturing the right data again, the right maintenance can be performed at a strategic moment. The last lecture of the knowledge days, by Koen Leeflang & Robert Vincente of Festo, dealt with how predictive maintenance can also be applied in the world of compressed air and pneumatics.