Current developments show that serverless computing is a subject of undergoing intense study in the IT (Informational Technology). As Service oriented architecture or Microservice-Architecture helped in increasing the flexibility and stability of manufacturing plants, it is anticipated that FaaS will help in addressing multiple challenges.
- Making systems more robots by real-time deployment and/or redundancy of software/control functions.
- Make use of 3rd party knowledge and services e.g. in using optimization platforms or machine learning models and divide logical and physical deployment e.g. in deep-learning processes
- Low-risk investments by effortlessly coupling cloud or IT solutions with the factory, e.g. being able to upgrade small chunks of software on a machine/production line and make use of further value-adding software functions independently of the physical deployment of these functions.
This use case provides carefully selected challenges for the next generation of manufacturing lines. It will demonstrate how to transform manufacturing architectures to serverless architecture, it will make use of the FaaS principle and applied semantics to enhance the flexibility and adaptability of modular production lines, demonstrate how FaaS helps in a variety of optimization problems, e.g. in hierarchical task planning and how it helps in improving the robustness of value-adding production services and therefore the overall system.
Involvement of UC partners
Fujitsu is a world class manufacturer in many domains as well as an Edge device manufacturer and provider. As such they will participate in the definition of the use-case scenarios and the evaluation, the business relevance, the technical requirements and possible solution architectures for the smart manufacturing use case. They will also contribute in particular in the development and evaluation of optimization algorithms as PHYSICS nodes (e.g. for component production planning, job shop scheduling or in general combinatorial optimization problems), also providing additional resources where these algorithms are tested and evaluated (e.g. Digital Annealing Unit). Additionally, they will help in setting up the Physics platform or make it compatible to their Fujitsu INTELLIEDE Edge Computing device.
German Research Center for Artificial Intelligence (DFKI), Department of Innovative Factory Systems (IFS), Kaiserslautern, founding member of Technology Initiative SmartFactory-KL e.V, is hosting one of the most advanced manufacturing-independent and flexible Industrie 4.0 production lines maintained together with the Technology Initiative SmartFactory-KL in Kaiserslautern. SmartFactory-KL is a non-profit association, which has the goal to leverage IT-technologies to the factory of the future by providing an industrial grade testbed for testing new technologies. Basic concepts are to treat all academic and industrial members equally and to realize manufacturer-independent solutions. DFKI will be responsible to evolve the current architecture to a FaaS enabled architecture, providing the relevant scenarios together with FuJ, and develop and evaluate relevant algorithms (AI enhanced or stochastic optimization) and the physical and logical distribution. SmartFactory initiative and will cooperate in demonstrating the benefits of PHYSICS and how it could be applied to real world scenarios.