Aarhus University Seal

MADE REACT

Global manufacturing currently faces challenges from geopolitical tensions to material shortages and a decreasing workforce. These challenges generate opportunities. Danish manufacturing companies are reconsidering strategies, favouring digitally driven, resilient operations while seeking to also strengthen their capabilities to handle High-Mix Low-Volume (HMLV) manufacturing. In MADE REACT, 15 companies and experts from five Danish universities and three RTOs will collaborate to explore AI, digital twins, advanced robotics, and real-time data-driven decision-making. REACT will focus on manufacturing challenges inside the manufacturing ecosystem divided into Factory-, Cell and Process level. A coordinated set of research activities will be initiated within these levels in combination with digital research going across all levels. Demonstration environments around the three levels will be established. These will enable integration of research results in realistic scenarios, and will increase visibility, facilitate dissemination and hence enhance impact. By this, MADE REACT will enable Danish manufacturing to be the most efficient in the world for the Resilient High-Mix Low-Volume (RHMLV) manufacturing of the future.

The project builds on Manufacturing's Academy of Denmark’s (MADE), strong experience and capabilities in bridging academic research to industrial practice by governing all stakeholders from academia over RTOs, technology providers and system integrators to end users. This follows a series of previous projects called MADE SPIR, MADE Digital and MADE FAST.

At AU we closely collaborate with CIM, Novo Nordisk, Grundfos and LEGO and we lead a Work Package called “Digital Integration and Decision Support”, where the goal is that we will:

  1. Establish digital twin capabilities at all levels of the manufacturing system with research as extensions of tools e.g., TensorFlow, CatchAI and DTaaS.
  2. Target optimal presentation of real-time manufacturing data to provide the right information about the production status to different stakeholders to enable them to take action when necessary.
  3. Determine autonomy possibilities by exploring when it is possible to take decisions autonomously at all levels of manufacturing (without human involvement).
  4. Update the operational manufacturing setup with MLOps capabilities to enable the learned models to be deployed in an operational manufacturing setting (including PLCs) in an automated fashion.
  5. Enable real-time fault detection and reporting to detect faults in the manufacturing system at a specific level and reporting to the right stakeholders who will be able to take corrective actions (if necessary).