Periodic Reporting for period 2 - AIMS5.0 (Artificial Intelligence in Manufacturing leading to Sustainability and Industry5.0)
Période du rapport: 2024-05-01 au 2025-04-30
AIMS5.0 a collaborative Innovation Action aims at strengthening European digital sovereignty in comprehensively sustainable production, by adopting, extending and implementing AI-enabled hardware and software components and systems across the whole industrial value chain to further increase the overall efficiency.
Vulnerability of existing supply chains in crisis shows the need for shorter supply chains and for keeping production in Europe. AI enabled fabs will be given more output and higher sustainability, which makes them more competitive on a global scale. New technologies from IoT and based on semantic web ontologies, ML and AI will help to enable the transformation from Industry4.0 to Industry5.0 to create human-centric workplace conditions and to enable the transformation of European industry to climate-friendly production. Above all, sustainability and resilience will be improved.
In essence, AIMS5.0 will deliver:
- AI-enabled electronic hardware components & systems for sustainable production
- AI tools, methods & algorithms for sustainable industrial processes
- SoS-based architectures & micro-services for AI-supported sustainable production
- Semantic modelling & data integration for an open access productive sustainability platform
- Acceptance, trust & ethics for explainable industrial AI leading to human-centered sustainable manufacturing
Vulnerability of existing supply chains in crisis shows the need for shorter supply chains and for keeping production in Europe. AI enabled fabs will be given more output and higher sustainability, which makes them more competitive on a global scale. New technologies from IoT and based on semantic web ontologies, ML and AI will help to enable the transformation from Industry4.0 to Industry5.0 to create human-centric workplace conditions and to enable the transformation of European industry to climate-friendly production. Above all, sustainability and resilience will be improved.
In essence, AIMS5.0 will deliver:
- AI-enabled electronic hardware components & systems for sustainable production
- AI tools, methods & algorithms for sustainable industrial processes
- SoS-based architectures & micro-services for AI-supported sustainable production
- Semantic modelling & data integration for an open access productive sustainability platform
- Acceptance, trust & ethics for explainable industrial AI leading to human-centered sustainable manufacturing
During the second reporting period significant progress toward the achievement of the objectives (O1-O6) were made:
O1 - A functional prototype of the Smart Voting-Wristband Arrowhead application device has been designed and manufactured, integrating a Bosch BMI270 motion sensor for accurate thumb gesture detection. The system architecture has been built using the Eclipse Arrowhead Framework, with secure communication established between the wristbands and a central server. A control panel has been developed using the Python Streamlit library for easy creation and configuration of voting sessions and real-time voting results.
O2 - There is a significant advancement of the AI toolbox. The toolbox now includes some well-defined functionality regarding design templates, examples, and reusable tools, describes best practices, that help the designers to enhance reliability, includes prompt evaluation to test industrial proof-of-concepts. The algorithms, tools and methods are identified for the use cases, information is collected about them containing their descriptions, plans and connections to UCs. They show solid foundation on the SOTA and significant advancement beyond.
O3 - The definition of Eclipse Arrowhead v5.0 and emerging implementation of v5.0 in both Java and Golang with new documentation partly based on the Arrowhead DSL provides an easier to install and use AI for automation integration technology. Particularly the implementation of security is simplified compared to v4.6 of Eclipse Arrowhead. Considerable contributions to other fundamentals like composability, real time and deployment has been made. Many not directly visible to most users. Most of these contributions provides improved robustness, usability and engineering efficiency.
O4 - WP4 has taken the review feedback into consideration to include more of the project use case partners involvement into the digital reference model and OAP. The simulation-based optimization approach for production planning of a single wafer fab has been designed and initially tested. Comparison of the planning UC with the concept of the ecological operating curve from UC 18 as well as the sustainability lobe of the Digital Reference was studied. Current structures in the Digital Reference Models regarding the battery management system of an electric vehicle were evaluated, suggested extensions for a product carbon footprint have been derived.
O5 - Various advancements were achieved across multiple facets of autonomous systems and AI-driven technologies, demonstrating the potential for transformative impacts in industrial settings. The continuous iterations of requirements engineering, application, and evaluation of developed solutions in lab environments and real production use cases ensured robust validation and refinement of AI techniques across different industrial domains. Intensive testing, validation and the influence on sustainability KPIs will be shown in the subsequent reporting period. For the majority of use cases an ontology was worked out in the second year. Many results of the use cases have a very positive influence on the working conditions (objective 6).
O6 - A self-assessment tool was designed, which is a chatbot built using a Retrieval-Augmented Generation (RAG) architecture. The chatbot helps users assess their AI systems against the EU AI Act by offering personalized compliance support through an intuitive conversational interface. The Quantitative Effect and Technology Acceptance Model (QETAM) was used to test the user acceptance of this tool.
O1 - A functional prototype of the Smart Voting-Wristband Arrowhead application device has been designed and manufactured, integrating a Bosch BMI270 motion sensor for accurate thumb gesture detection. The system architecture has been built using the Eclipse Arrowhead Framework, with secure communication established between the wristbands and a central server. A control panel has been developed using the Python Streamlit library for easy creation and configuration of voting sessions and real-time voting results.
O2 - There is a significant advancement of the AI toolbox. The toolbox now includes some well-defined functionality regarding design templates, examples, and reusable tools, describes best practices, that help the designers to enhance reliability, includes prompt evaluation to test industrial proof-of-concepts. The algorithms, tools and methods are identified for the use cases, information is collected about them containing their descriptions, plans and connections to UCs. They show solid foundation on the SOTA and significant advancement beyond.
O3 - The definition of Eclipse Arrowhead v5.0 and emerging implementation of v5.0 in both Java and Golang with new documentation partly based on the Arrowhead DSL provides an easier to install and use AI for automation integration technology. Particularly the implementation of security is simplified compared to v4.6 of Eclipse Arrowhead. Considerable contributions to other fundamentals like composability, real time and deployment has been made. Many not directly visible to most users. Most of these contributions provides improved robustness, usability and engineering efficiency.
O4 - WP4 has taken the review feedback into consideration to include more of the project use case partners involvement into the digital reference model and OAP. The simulation-based optimization approach for production planning of a single wafer fab has been designed and initially tested. Comparison of the planning UC with the concept of the ecological operating curve from UC 18 as well as the sustainability lobe of the Digital Reference was studied. Current structures in the Digital Reference Models regarding the battery management system of an electric vehicle were evaluated, suggested extensions for a product carbon footprint have been derived.
O5 - Various advancements were achieved across multiple facets of autonomous systems and AI-driven technologies, demonstrating the potential for transformative impacts in industrial settings. The continuous iterations of requirements engineering, application, and evaluation of developed solutions in lab environments and real production use cases ensured robust validation and refinement of AI techniques across different industrial domains. Intensive testing, validation and the influence on sustainability KPIs will be shown in the subsequent reporting period. For the majority of use cases an ontology was worked out in the second year. Many results of the use cases have a very positive influence on the working conditions (objective 6).
O6 - A self-assessment tool was designed, which is a chatbot built using a Retrieval-Augmented Generation (RAG) architecture. The chatbot helps users assess their AI systems against the EU AI Act by offering personalized compliance support through an intuitive conversational interface. The Quantitative Effect and Technology Acceptance Model (QETAM) was used to test the user acceptance of this tool.
AIMS5.0 will result in lower manufacturing costs, increased product quality through AI-enabled innovation, decreased time-to-market and increased user acceptance of versatile technology offerings. They will foster a sustainable development, in an economical, ecological and societal sense and act as enablers for the Green Deal and push the industry towards Industry5.0.
The innovations will leverage the experience of the 53 partners, such as renowned OEMs, Tier-1 and Tier-2 suppliers, technology and application large enterprises and SMEs, supported by academic research specialists in fields like AI, industrial hardware and software, decision making and management algorithms.
Specific outcomes of the project are
- 20% faster time to market,
- Participation of disabled people in the factory environment > 5% (in relation to the total number of employees employed in production),
- AI based MES capability > 10 %,
- Increased user awareness and trust by 10%,
- Subsequent reduction of environmental footprint for wafer transport, handling and storage > 20 %,
- 50% reduction of time for monitoring industrial equipment.
The innovations will leverage the experience of the 53 partners, such as renowned OEMs, Tier-1 and Tier-2 suppliers, technology and application large enterprises and SMEs, supported by academic research specialists in fields like AI, industrial hardware and software, decision making and management algorithms.
Specific outcomes of the project are
- 20% faster time to market,
- Participation of disabled people in the factory environment > 5% (in relation to the total number of employees employed in production),
- AI based MES capability > 10 %,
- Increased user awareness and trust by 10%,
- Subsequent reduction of environmental footprint for wafer transport, handling and storage > 20 %,
- 50% reduction of time for monitoring industrial equipment.