InnoGuard: Hybrid and Generative Intelligence for Trustworthy Autonomous Cyber-Physical Systems

The Marie Skłodowska-Curie Actions-funded InnoGuard project advances quality assurance for AI-driven Autonomous Cyber-Physical Systems, automating assessment, improving reliability, sustainability, energy efficiency, legal compliance, and trustworthiness.

InnoGuard tackles emerging quality assurance challenges in AI-enabled Autonomous Cyber-Physical Systems (ACPS), increasingly used in daily life, including public transport. Incidents such as accidents involving Tesla Autopilot highlight the need for dependable development methods. 

The project develops novel techniques to automate ACPS quality assessment and behavior evolution, integrating real-time security, privacy, and uncertainty handling. It also strengthens AI trustworthiness, improves energy efficiency using AI and Large Language Models, and validates solutions in open-source environments such as Robot Operating System-based platforms. Through advanced research, training, and dissemination, InnoGuard promotes reliable, sustainable, and legally compliant ACPS engineering.