Project Partners

Coordinates the entire project and oversees day-to-day operations, financial tracking, and risk management. This partner leads the creation of the safety-certified workstation adaptation agent, utilizing quantised Long Short-Term Memory (LSTM) networks to foresee short-term worker overload or task deviations. HIS integrates these technical insights to automatically adjust manufacturing speeds, thereby preserving worker energy and preventing injuries without sacrificing plant efficiency.

Is responsible for the engineering requirements, use case specifications, and the multi-agent system orchestration logic. LMS builds the alternative scenario simulation engine featuring advanced multi-objective optimization to help factory managers select ideal augmentation strategies. Furthermore, LMS deploys a post-hoc eXplainable AI (XAI) layer that integrates SHAP, LIME, and LRP techniques. This layer converts complex automated choices into transparent, color-coded heat maps for XR overlays and structured audit trails on dashboards, ensuring full compliance with the EU AI Act’s stringent transparency guidelines.

Contributes core Social Sciences and Humanities (SSH) expertise to anchor the project in cognitive psychology and user experience design. UU formalises the multidimensional Human-Value Scorecard to evaluate worker autonomy, competence, and trust using real-time sensor streams. Additionally, UU designs the edge context microservice and pilots the immersive, hands-free Unity-based XR smart glasses interfaces that present spatial callouts and safe gaze-aligned guidance directly to operators.

Leads the ingestion pipelines to extract structured data from unstructured factory text, standard operating procedures, and technical manuals. TUDa constructs plant-specific knowledge graphs and aligns them with Asset Administration Shell (AAS) semantics. This architecture grounds “FactoryGPT”—the project’s multi-modal conversational large language model assistant—allowing it to supply context-aware, low-latency training and troubleshooting dialogues to operators with minimal hallucinations.

Drives the technical co-design and implementation of the OMNI-SENSE suite’s multi-modal hardware configuration. UNIVPM integrates medical-grade wearable bracelets and smart headbands that capture real-time biometric metrics, including heart rate variability, skin temperature, and electrodermal activity. These signals feed the core system to systematically classify physical strain and cognitive workload across diverse demographics.

Develops miniaturised, close-range 3D vision systems and wearable artificial fingertip tactile sensors embedded with accelerometers and load cells. USIT digitises delicate, traditionally manual evaluation workflows, such as micro-level flush and gap measurement. This hardware links physical human actions directly with digital quality systems, enabling instant corrective notifications through interactive consumer interfaces.

Designs and delivers the core digital backbone of the project by deploying a secure, containerised multi-agent system platform built upon Apache Kafka and the Asset Administration Shell (AAS) V3 metamodel. NETCOM builds OAuth 2.0 authorization frameworks, signed Docker images, and field-level encryption to satisfy rigid audit standards. Netcompany also translates automated agent metrics into centralised, web-based management dashboards to ensure seamless supervisor oversight and system-wide transparency.

Creates a self-calibrating, modular stereovision sensor grid solution to track operator movements, spatial positioning, and 3D kinematics on the shop floor. This vision grid achieves depth accuracy superior to 1 cm across large industrial work areas. VIS implements automated, on-device pre-processing to filter and anonymise all raw video streams prior to transmission, preserving worker privacy in compliance with EU data regulations.

Viewpointsystem develops lightweight, industrially rugged XR glasses integrating binocular eye-tracking. Based on the VPS NEXT, Viewpointsystem’s latest smart glasses, it is planned to develop different see-through glasses prototypes following a human-centric approach. By providing cognitive monitoring features through binocular eye tracking, combined with the OMNI-SENSE suite to be developed during the project, this sensor fusion framework will allow to automatically trigger workplace alerts and adaptive interventions when thresholds are exceeded. The XR glasses with eye tracking will also integrate diverse miniaturized sensors supporting simultaneous localization and mapping (SLAM), and features such as measuring the alignment of parts. Then, it will be possible to verify that technicians execute manual tooling assembly steps correctly and in the required sequence, reducing procedural errors on the factory floor.

FILL contributes to the MANUFACTOR project through its “Next World Factory”, a dedicated training and learning environment for future manufacturing skills. In this controlled setting, apprentices can safely practice and train specific production tasks supported by modern assistance technologies. The focus lies on hands-on learning, intuitive human–machine interaction, and effective upskilling. This approach helps bridge the gap between education and real industrial applications while improving workforce readiness.

Provides the specialised industrial demonstration environment at its Ludvika manufacturing site to validate augmentation during high-voltage direct current transformer assembly. HE utilises the integrated multi-agent system to streamline complex, torque-critical bushing and tap changer assembly tasks. This pilot aims to reduce heavy lifting injuries, accelerate historical talent proficiency cycles, and decrease production defects.

Coordinates the Portuguese automotive line pilot case, evaluating cognitive augmentation across intense vehicle assembly and inspection cycles. VWAE utilises wearable tactile devices and virtual gauges embedded into smart glasses to inspect car body gap and flush parameters. This implementation eliminates static printed documentation and allows fluid workforce job rotation across complex assembly lines.

Sets up a dedicated, operational learning factory testbed facility in Patras, Greece, to evaluate physical worker burden. TFCC deploys advanced robotics stations, tactile wearables, and 3D vision grids to systematically monitor proximity hazards and optimise heavy assembly steps. This testbed validates individual hardware prototypes and trains engineering students, managers, and small-business operators on safe Industry 5.0 systems.

Organises the strategic dissemination, exploitation, and multi-stakeholder communication campaigns across all European target audiences. F6S manages the Stakeholder Advisory Panel, institutes the multi-site co-creation working groups, and sets up collaborative design workshops. This work ensures that operators, managers, and trade union representatives actively shape technical features to foster workplace trust.