Digital pipeline for data orchestration and reconfigurable production systems
The traditional manufacturing methodology relies on manual optimization loops in which part of the information is manually exchanged from the different manufacturing steps coming from different formats, sizes, and types (data at rest/static data and data in motion) which requires different computing platforms and results in information silos. Another handicap is the lack of standardization and definition of the data modeling resulting in non-interoperable and non-reconfigurable systems that are impossible to orchestrate manufacturing processes and resources.
Digital technologies can make a significant contribution to making production systems more flexible for modular production. Reconfigurability and interoperability between automated industrial systems and Cyber-Physical Systems (CPS) can be achieved by implementing formal definitions following Industrie4.0 recommendations and implementing data models or Digital Twins based on Asset Administration Shell (AAS). This AAS modeling contains machine-readable and self-description metamodels providing interoperability capabilities to the different assets involved in the production systems (e.g., robots, software modules, perception systems, …). Converging will take advantage of the knowledge from previous projects such as DIMOFAC in which modularity and reconfigurability based on AAS is the main objective.
For that Converging will provide a modular software-oriented architecture (SoA) able to interconnect all production entities (Big Data pipeline) for real-time capturing (Digital Twin), storing (Data at rest) and processing (Data in motion) to support autonomous and collaborative behavior with minimal user intervention achieving Plug and Produce capabilities thank to the common data modeling.
Figure 1. Converging Digital Thread based on I4.0 architecture
Special focus will be on the Digital Thread or digital pipeline for seamless data flow across the product lifecycle for interconnecting heterogeneous data. When it comes to multi-disciplinary production systems, data traceability for tracing the interplay between different domains (e.g., design, digital twins, production, quality, …) is one of the most important points to be considered, as working on the right data could significantly reduce the integration time and costs. Previous EU projects such as PENELOPE, have implemented this concept, but have not focused on collaborative robotics applications and have not considered all types of data (e.g., data in motion).
In Converging we will go beyond previous deployments, and the consortium will work on a tool able to orchestrate the production entities thanks to the information that is pushed and propagated down the Digital Thread and the use of a common language (AAS data modeling). Information from the production entities (e.g., robots or humans), task planner tools (e.g., optimized tasks sequence), perception sensors (e.g., 2D and 3D vision systems), and quality systems (e.g., Ultrasound equipment and data) will be considered and tested in 4 different use cases in 4 different sectors –automotive, white goods, aeronautics, and additive manufacturing– and it will be available for testing/training during the Open Pilot lines demonstration at TECNALIA, IPK, LMS, and AIMEN facilities. So, stay tuned!