The research is part of the Industry 4.0 approach, which aims to develop Control, Remote Monitoring and Remote Maintenance Centers implementing various tools, techniques, technologies and processes in order to provide effective after-sales service to customers.
The project will also allow the industrial members of the consortium to gain in productivity and differentiation from their competitors as well as to develop new business models based on the servicization of equipment and/or PaaS (Product-as- a-Service) offers.
Data Science ⊕
Domaine: Manufacturing ⊕
Innovation theme: Artificial Intelligences ⊕
Asset: TSorage ⊕
The objective of this project is to allow, in the long term, the member companies of the consortium to be able to create and operate advanced control centres to support more efficient after-sales service activities, or even new business models of type "service".
To achieve this, this project will push innovation in distributed artificial intelligence technologies with preservation of data privacy, as well as augmented reality techniques.
Augmented reality research will be carried out by the "Research Center in Information Systems Engineering (PReCISE)" at the University of Namur.
Federated learning research will be carried out by the "Big Data and Machine Learning (BDML)" at the University of Mons and will be integrated into software libraries by CETIC.
The industrial members of the consortium will provide practical cases for validation, as well as the business expertise necessary for the critical analysis of the results.
The expected result of this project is the validation in an industrial environment of research results in federated learning and in augmented reality responding to the challenges of the industrial world. To this end, industrial cloud "data hub" infrastructures will have to be adapted and edge devices compatible with remote AI outside the cloud will have to be developed. Research results will take the form of recommendations, algorithms, heuristics, and tools based on these algorithms. The industrial members of the consortium should then be able to build advanced control centres on the basis of these results.
The innovation driven by this research project resides in that federated learning and augmented reality techniques require the resolution of a number of technological limitations in order to effectively integrated into an industrial environment.
For federated learning, technological challenges include the federated and robust modelling of heterogeneous signals coming from different sites, as well as the processing of data flows in a dynamic and heterogeneous environment.
For augmented reality, the challenge remains to improve current immersive analysis, gestural interaction, and remote collaboration.