STAR

Safe and Trusted Human Centric Artificial Intelligence in Future Manufacturing Lines

Artificial intelligence (AI) systems are increasingly improving production automation in the manufacturing sector. But for these systems to be reliable and applicable when replacing human activities in dynamic operations, they must be safe and adjustable to react to different situations, security threats, unpredictable events or specific environments. The STAR project, funded by the European Commission, will respond to this challenge by designing new technologies for the implementation of AI systems in production environments. These must be reliable, human-centred and based on secure standards. The project will research and integrate cutting-edge AI technologies such as active learning systems, simulated reality systems, digital twins and cyber defence mechanisms to enable the safe implementation of sophisticated AI systems in production lines.

AI systems in industrial plants must be safe, reliable and secure, even when operating in dynamic, unstructured and unpredictable environments. STAR will develop technologies that enable AI systems to operate in complex environments and withstand cyber attacks. STAR will research and integrate cutting-edge AI technologies with broad applicability in manufacturing environments, including: • Active learning systems that increase safety and accelerate knowledge acquisition. • Simulated reality systems that accelerate learning in scenarios involving collaboration with human robots. • Explainable artificial intelligence (XAI) systems that increase the transparency of industrial systems and boost confidence in them. • Human-centric digital twins that enable worker monitoring for safer and more reliable production processes. • Advanced RL techniques for optimal mobile robot movement and safety zone detection in industrial plants. • Cyber defence mechanisms for sophisticated attacks against deep neural networks operating on industrial data. These technologies will be validated on production lines in the areas of quality management, robot-human collaboration and agile AI-based manufacturing. STAR will remove security and safety barriers to the implementation of sophisticated AI systems on production lines. The results will be fully integrated into existing EU-level initiatives (EFFRA, AI4EU).

Translated with DeepL.com (free version)

R2M participates in STAR with the aim of improving its expertise in AI technologies in order to develop new advanced solutions for manufacturing, with a particular focus on solutions for industrial customers. In addition, R2M will support the market introduction of these technologies and develop a roadmap of feasible exploitation opportunities for the STAR solution.