Mr. von Birgelen, what is the Improve research project about?

Alexander von Birgelen: In our Improve team, we have been conducting intensive research into Industry 4.0 since the start of the project, covering the entire research spectrum - from data acquisition, machine learning, simulation, optimization, diagnostics and user interface optimization. We have developed solutions in all areas that are now being tested in industry.

We are a total of 13 international project partners from industry, research and IT. The Ostwestfalen-Lippe University of Applied Sciences, where I work, is coordinating the project.

Which area have you been working on?

Our team has been working intensively on a new condition monitoring tool that could replace traditional, manual condition monitoring. At the moment, condition monitoring is usually carried out by experienced employees. However, accurate predictions of wear and required maintenance work are difficult and are usually only based on empirical values.

In order to minimize this source of error and make accurate predictions, we use artificial intelligence methods. The condition monitoring tool developed is based on a machine learning framework. The methods make it possible to collect data from industrial plants and obtain models of plant behavior from this very data. For example, the models can detect and localize anomalies or predict the condition of certain components of the industrial plant.

What are the advantages over previous condition monitoring tools?

Thanks to the self-learning methods and data-based prediction of required maintenance work, our model offers enormous potential for innovation and improvement. It is modular and can be adapted to a wide variety of industrial plants. As the prediction is based on a self-learning model, there is also no need for manual modeling.

Is the tool already being tested? What successes can be seen?

Yes, the tool is being tested both by us at SmartFactoryOWL at inIT at Ostwestfalen-Lippe University of Applied Sciences and by our industrial partners Reifenhäuser Reicofil and our Italian partner OCME. The tests have not yet been completed, but the results of the data obtained so far are very promising. We are currently working together to implement the tool in the partners' industrial plants.

What are you showing at the Hannover Messe 2018?

We are presenting one of our demonstrators from the SmartFactoryOWL, which we can use to show various models and learning processes of the tool. In addition to the topic of artificial intelligence in automation, we will be presenting other technologies from the fields of information fusion, industrial internet, cyber security and assembly assistance systems. You can find us at the it's OWL joint stand in Hall 16, Stand A04.

What is the roadmap for the future?

Our Improve project is now about continuing the successful tests to date and perfecting the model so that partners in industry can adapt and use it for themselves. Here I would like to differentiate between scientific and industrial use. Our technologies are the basis for further research in the field of Industry 4.0 on the one hand and the basis for technology transfer and successful implementation in industry on the other.