EASY
Energy-efficient analysis and control processes in the dynamic edge-cloud continuum for industrial production
Short description
EASY is developing an edge/cloud system for industrial production in which AI-based analysis and control processes can be executed in a distributed manner and on site wherever possible. This approach significantly enhances productivity and resource efficiency across the entire production process. |
EASY is developing a user-friendly system for industrial production that gives companies flexible and sovereign control over whether and how they process their data locally ("on the edge") or centrally ("in the cloud"). This “edge-cloud continuum” offers corresponding services to this end, which automatically optimise the energy requirements and data usage of the applications used. EASY is thus creating the technological basis for control processes that are easy to implement, both within automation and in the context of the data economy. Companies of all sizes can use EASY to increase their productivity and the resource efficiency of their production processes, while also securely implementing autonomous data utilisation.
Market perspectives and product claims
Companies deploying EASY decide themselves which production-sensitive data they choose to retain within the company and which data they transfer to the cloud. In this way, EASY removes a current major obstacle to the use of production data for process optimisation. Furthermore, EASY is pioneering a new, energy-saving and therefore more climate-friendly data processing technology through dynamic service distribution. The project thus contributes to the implementation of core sustainability goals such as economic growth, promoting innovation, and energy and resource efficiency (UN Sustainable Development Goals 8, 9 and 12).
By strictly following open standards such as GAIA-X and the Industrie 4.0 administration shell, EASY supports an ecosystem in which not only providers of industrial analytics applications, but also conventional production companies can participate. Market entry is facilitated in particular for small and medium-sized companies with typically limited IT resources. In addition, EASY relies on a zero-trust architecture that incorporates the communication influences of the overall system into the security architecture, thus achieving a new level of data security. The more companies that use EASY, the easier it will be to prevent attacks on companies, while also reducing the damage caused if attacks do occur.
Challenge and innovation
The overarching mission of Industrie 4.0 is to achieve significantly increased efficiency and greater flexibility in production by fully digitalising and networking all industrial processes and systems. This objective requires the convergence of operational technology (OT) and information technology (IT). Only through this convergence of operational and information technology can the direct control and comprehensive monitoring of all production levels be achieved, along with the complete integration of value chains. EASY's innovative approach centres not on implementing IT/OT convergence via a central cloud, but on creating decentralised data processing through the edge-cloud continuum. AI-based analysis and control services can be executed within the continuum in local edge and cloud environments alike. Execution is dynamic: The continuum can therefore automatically balance local and global computing capacities via measurement and analysis procedures for resource consumption and data transfer.
The project will therefore need to develop innovative measurement and analysis methods for resource consumption and data transfer as well as tools for efficiently balancing local and global computing capacity. EASY is investigating which architecture for distributed learning processes offers the best resource efficiency. With regard to flexible planning and decentralised monitoring, AI planning processes based on digital twins are being implemented for resilient control processes at the edge level.
Use cases
The methods and systems developed as part of the EASY project will be evaluated using selected applications in a variety of production environments with regard to economic and ecological criteria and demonstrated in multiple reference applications (e.g. in the SmartFactories KL and OWL as well as in the ARENA2036 research factory in Stuttgart; three further demonstrator applications are being developed at BOSCH, ArtiMinds and Coboworx). The results of the evaluation serve as a corrective measure in the technical development of the edge-cloud continuum. The use cases will demonstrate a complete and dynamic edge-cloud continuum that complies with data sovereignty requirements and is directly transferable for manufacturing companies.
Consortium
Empolis Information Management GmbH (consortium leader), Deutsche Forschungszentrum für Künstliche Intelligenz GmbH (DFKI), Robert Bosch GmbH, Industrial Automation branch INA of Fraunhofer IOSB, Trier University of Applied Sciences – Environmental Campus Birkenfeld, ArtiMinds Robotics GmbH, coboworx GmbH, Salzburg Research (Austria)
Benefits
Current situation | Future vision |
The complete digitalisation and networking of all industrial processes and systems often fails because of unresolved data protection issues and a lack of trust in cloud providers. | With the edge-continuum, EASY enables the local execution of AI-based applications, thus ensuring genuine data sovereignty for manufacturing companies. |
Cloud-based approaches to data processing are associated with high latency times and data transfer costs that do not meet the requirements of Industrie 4.0. Opportunities for managing production data therefore often remain unused. | With EASY, companies in the manufacturing sector can reap the benefits of both edge and cloud computing efficiently while conserving resources. |
Decarbonisation of the industrial sector lags behind its potential due to difficulties associated with cloud-based processing of manufacturing data. | The edge-cloud continuum establishes the technical prerequisites for user-friendly control and analysis applications that optimise manufacturing processes in a resource- and energy-efficient manner. |
Proprietary systems present high barriers to market entry for providers of industrial analysis systems. | The EASY continuum ensures semantic interoperability across multiple manufacturers and multiple components. Smaller AI start-ups can therefore also develop new generic analysis services and offer these to manufacturing companies via the marketplace. |