Component of the "Smart Factory Elements" model
Predictive Manufacturing or Prediction in general is about predicting events and results on the basis of data and analyses. At the same time, Predictive Manufacturing is one of the components of the control loop of the "Smart Factory Elements" model. To remind ourselves: In this control loop, production is planned according to specifications of different sources (Planning & Scheduling) and this plan is then implemented or executed (Execution). The data collected is analyzed (Analytics) in order to make predictions (Prediction) which, together with other findings, can then be integrated in the planning. The Industrial Internet of Things supports this cycle by collecting and providing data.
Predicting events with AI
Prediction applications use large amounts of recorded data and methods of the artificial intelligence (AI) to predict events and behaviors. The procedure is very similar to that of a human expert. A good forecast is based on experiences gained and interrelations detected. The main difference is that AI can process more data in less time..
Predictive quality
In addition to the omnipresent Predictive Maintenance, MPDV offers Predictive Quality, the first application to predict product quality using recorded real-time data. The predictions are based on a model, which can be created via machine learning and is then executed by the Manufacturing Execution System (MES) HYDRA. Other Prediction applications are being developed.
Outlook
The MES HYDRA already meets most of the requirements that the "Smart Factory Elements" model places on a modern IT tool for Predictive Manufacturing – for example by calculating material availability. With the increased use of artificial intelligence and machine learning, MPDV will gradually extend existing and develop new applications.
(approx. 2,100 characters)