How does making an (older) machine 'smart' work in practice?

The presentation by Bas de Koningh (Routeco) and Timo Hamoen (PlantMatiQ) on the Mechanical engineering event 2019 gives visitors practical tips on what is needed to make an (older) machine 'smart'. And: what do you do with the retrieval, storage and processing of the data? In 2018, PlantMatiQ successfully converted an (older) machine to the latest state of the art. Routeco Nederland provided the hardware engineering for this.

The next step is the 'smart' part: retrieving, storing and processing data. PlantMatiQ is developing software that retrieves the data with a router and collects it in the cloud. This data can then be used for preventive maintenance, for example. Together, PlantMatiQ and Routeco use this practical example to show what is needed to make an (older) machine 'smart' and what possibilities this then offers.

Predictive maintenance

The potential benefits of predictive maintenance are increasingly being discussed in conference program of the Machinebouw event 2019. By predictive maintenance By applying it at crucial locations, higher availability can be achieved at lower maintenance costs. Higher availability of machines at lower maintenance costs, who wouldn't want that?

Wiljan Vos by Legs Industrial Automation explains it to the audience in his lecture: by predictive maintenance By applying it at crucial locations, higher availability can be achieved at lower maintenance costs. New sensor and communication technology ensures more extensive applications and thus more insight into the reliability of assets. By means of historical data and analysis thereof, it is predicted when maintenance needs to be carried out. Beenen Industrial Automation presents on behalf of Phoenix Contact how various applications in this area can be realised economically with the help of a smart control platform.

Predictive Maintenance through Automated Machine Learning

In the presentation of Marcel Tuit (Weidmüller Benelux) clearly shows what the first question should be in every 'Industrial Internet of Things' project: what do I want to achieve? More efficient machines with more or a higher quality output? Less downtime? A new revenue model in which I offer the output of my machine as a service? Only when this is clear in mind does it make sense to give substance to predictive maintenance, for example.

The software from Weidmüller could then be helpful for this. Via an intuitive interface it is possible to enrich the data by entering irregularities, for example. Machine builders themselves often know best when and how much the temperature or pressure deviates from the norm, or which vibration pattern indicates a certain type of wear. The more data points are added, the higher the probability percentage of the various algorithms. This is explained in practice using a practical case.

The sixth edition of the Machinebouw event will take place on Tuesday, December 3, 2019 at 1931 Conference Center in Den Bosch. Register to visit the event. This is completely free (including coffee, lunch and networking drinks) provided you belong to the target group.