Fraunhofer Institute for Production Technology IPTResearch Project "AI-NET-ANIARA"
Digital transformation through the use of AI in the management of network infrastructures for production and sensor technology
The 5G mobile technology is ready for the industrial use. Deployment is starting in several markets around the world. 5G offers unprecedented capacity, low latency and significantly improved support for a wide range of applications. To realize its full potential, critical applications are no longer deployed in a centralized cloud environment, but in distributed locations at the edge of the network. This creates the conditions for a wave of innovations - even more so in interaction with other key technologies such as edge computing and artificial intelligence.
The goal in the project was to enable companies to use high-performance services that are operated at the network edge. Technologies around network edges can be used in applications for smart cities, robotics, public safety, sensors and manufacturing. The part of "AI-NET-ANIARA" performed by the German consortium partners focuses on the application fields around sensors and manufacturing technologies.
5G, edge computing and artificial intelligence for automated network management
To create service and application platforms, the partners of the "AI-NET-ANIARA" project combined three technologies: 5G, edge computing and artificial intelligence. European industry holds a good position in the roll-out of 5G networks, transportation and industrial applications, but needs to strengthen its position in cloud technologies, data center technology, virtualization and artificial intelligence to be a leader in these technology fields. The Fraunhofer IPT provided its expertise in four use cases of the research project "AI-NET-ANIARA":
- Multisensory platform for air exchangers: The goal of the first use case was to develop a mobile multisensory platform for measuring and transmitting air parameters (e.g. temperature, flow, pollution) of industrial air exchange systems using LoRa and/or 5G infrastructure.
- Activity state detection / plausibility check: The second use case aimed at reading, checking and analyzing existing clustering patterns in time series data of energy carriers (compressed air, heat, electricity and process gas) as well as detecting the status of edge devices behind a measuring node.
- Predictive maintenance: The goal of the third use case was to develop early warnings of potential hazards to air exchange systems based on equipment vibration data.
- Incremental manufacturing: The fourth use case was about integrating digital infrastructures and individualizing the process chain while improving the quality of the collected data and efficiency through the use of intelligent data processing using machine learning.
AI technologies for network operators support edge automation
The partners on the research project "AI-NET-ANIARA" developed advanced AI technologies for network operators that specifically target predictive and proactive edge automation and the various services that the network edge will support. The algorithms were intended to be compatible with and leverage the distributed edge infrastructure to increase efficiency, reliability and scalability. The research project "AI-NET-ANIARA" also expected to produce new technologies for "self-driving" network management guaranteeing the performance and availability of services and reduce complexity for end users.
Structure of the overall project "AI-NET" and overview of the participating countries of the respective subprojects "AI-NET-ANTILLAS", "AI-NET-PROTECT" and "AI-NET-ANIARA".
"AI-NET" is an EU research program to accelerate the digital transformation in Europe through the automation of intelligent networks. The overarching structure covers topics such as the automation of telecommunication networks ("AI-NET-ANTILLAS"), the security of network services ("AI-NET-PROTECT") and the automation of network edge infrastructures and applications with artificial intelligence ("AI-NET-ANIARA").
Project Coordination
- Ericsson AB (EAB), Schweden
Project Consortium
- Arctoslabs AB, Sweden
- Chalmers University of Technology (CTH), Sweden
- Delta Electronics, Sweden
- Enoc System AB, Sweden
- Fraunhofer Institute for Production Technology IPT, Germany
- Fraunhofer Institute for Surface Engineering and Thin Films IST, Germany
- HAL Robotics, United Kingdom
- IconPro GmbH, Germany
- King‘s College London, United Kingdom
- Konica-Minolta, United Kingdom
- Logical Clocks AB, Sweden
- Opel Automobile GmbH, Germany
- Qamcom Research and Technology AB, Sweden
- RI.SE Research Institutes of Sweden AB, Sweden
- Royal Institute of Technology, KTH (Kungliga Tekniska Högskolan), Sweden
- Systemair AB, Sweden
- Technische Universität Braunschweig, Germany
- Univrses AB, Sweden