#	Pagina
attuale pagina	/open-h2020/projects/221093/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - FlexInt (Flexible Manufacturing using Semantically Interoperable Data Integration)

Teaser

Summary

Every manufacturing enterprise and, in turn, its manufacturing systems, strive to produce at lowest possible cost, to perpetually enhance product quality, and to be able to rapidly respond to changing market demands. Existing manufacturing systems are not designed to meet the required market responsiveness neither at acceptable cost nor at reasonable speed. They are based on rigid production automation architecture designed to produce core products at high-volume and planned capacity in order to be profitable. Consequently, these systems cannot efficiently cope with variations in product demand and the production systems cannot respond properly to changing market demands.
When automation and robotization cannot cope with reshaped way of doing business, what can? Our solution should provide an answer. Namely, FlexInt enables the user of an automation system (for example Production Engineer) to autonomously execute the integration and reconfiguration of a production machine, e.g. industrial robot, without engaging teams of specialists, simply by applying his/her domain knowledge of manufacturing processes and of the machine. The rest is delivered by our Semantically Interoperable Data Integration, a set of digital platform and digital tools (applications), distinguished through their remarkable user experience and operational performance.

Work performed

We developed a prototype that has been tested in a closed and controlled environment of our partner. Key enabling technologies have been tried out. The outcomes have been carefully analyzed and several possible technical implementation strategies proposed. The next step is to upgrade the prototype and try it out in an operational environment at manufacturer site, followed by the system completion and verification in a fully operational environment.

Final results

Within the scope of the project we extended the concept of the solution with a second prototype. This solution allowed us to integrate and feed the IBM Watson unsupervised machine learning solution with the aim to detect anomalies in manufacturing process. Our concept proved of utmost value since it was able to accommodate changes in data models of manufacturing application within an hour.
We have demonstrated that prototype on Hannover Messe 2019 at our own booth as part of IBM solutions. Conducting the discussions with the visitors that are manufacturing decision makers was the second source of customer feedback to our solution.