With Industry 4.0, manufacturing companies are adopting IoT devices, leading to large volumes of data generated from the shop floor. Such data are extremely valuable to design and operate manufacturing systems, but they must be aggregated, cleaned, and analyzed to be exploited. In addition, the amount of manufacturing software systems is growing in diversity, which significantly increases the complexity of integrating and synchronizing these systems in efficient manufacturing pipelines. ASSISTANT will provide a secure and intelligent data fabric as an autonomous data management tool to manage, protect, and make data available throughout the full life cycle of the systems. The data fabric will (i) automate secure collection and cleaning of data, (ii) integrate digital twins, (iii) identify/allocate suitable resources for data storage processing (including training of machine learning and AI), and (iv) make data available for exploitation in AI reasoning systems. It will provide a new level of interoperability in AI for manufacturing systems and will be designed to be standardizable, technology neutral, and adaptable to a wide range of flexible manufacturing scenarios.

The objectives of Work Package 6 are :

  • To develop uniform data models and a standardizable specification for a secure and intelligent data fabric capable of instrumenting and integrating the ASSISTANT AI systems.
  • To design and define template blueprints for domain models to be used in the ASSISTANT digital twins.
  • To provide prototype implementations of the data fabric to be used by the ASSISTANT systems.
  • To establish a simulation-based platform for modeling and experimentation with computation / data orchestration and enactment strategies (i.e. simulation support for the data fabric AI).
  • To develop mechanisms for characterization, analysis, and prediction of data fabric workloads.
  • To provide foundational AI tools for intelligent (planning and scheduling based) and secure the orchestration of the ASSISTANT data fabric systems.

Leader WP 6: P-O Ă–stberg, PhD

LEAD BENEFICIARY