ECSEL-project DAIS, Distributed Artificial Intelligence Systems

The three Norwegian partners Step Solutions, Expert Analytics, and Jotne Connect have participated in the project Distributed Artificial Intelligence Systems (DAIS). Throughout the project, these companies have collaborated to develop an industrial demonstrator at the Norwegian gypsum board manufacturer NORGIPS. The demonstrator combines data collection, control system integration, analysis with Artificial Intelligence, and Product Lifecycle Management to optimize a critical step in gypsum production. This combination of technologies has been successful, and the partners are exploring new opportunities to further mature this solution for other industries as well. Participation in the DAIS project has also involved close collaboration with a wide range of other European companies, where sharing of analysis results and technological solutions has been important. The DAIS project has been an important initiative to both strengthen each company's development of technological solutions and to combine these into a holistic solution to optimize the process industry in line with Industry 4.0 principles.

Effects: -Increased competence in Europe for the integration of AI in edge computation aimed at data analysis for industrial use cases -Establishment of capability in the use of AI for industrial use cases, in particular for data harvesting, monitoring, maintenance and analytics aimed towards process industry Impact: - Increased Norwegian and European competence in the practical application of AI for industrial purposes - Optimization of the production process of plasterboards, aimed towards the raw material processing - Market growth and expansion of staff members for the norwegian companies involved in DAIS

The DAIS project aims to strengthen Europe's position within distributed artificial intelligence (AI). At the present, data analysis mainly takes place in cloud-based systems and thus at a separate location to where the data is produced and collected. This leads to AI being applied outside of the systems it is meant to improve. The solution is to place AI at the level where the data is generated, such that analysis, support for deciscions and process control takes place as close as possible to the monitored activity. The impact of this is that processes can be improved locally, automatically and in real time. Depending on the state of the market, this can have a disruptive effect and be a major competitional advantage if applied. The technology shall be based on open standards and be developed within the framework of existing industrial requirements both for hardware and software. The Norwegian use case will collect, process and analyze audio data from a running water turbine in order to optimize start-stop cycling of the machinery. The analysis will be done by AI connecting to a product lifecycle management software framework. The output from the analysis will be fed back to hardware connected to the system in order to facilitate optimized regulation.