Regenerative AI (RAI) is a new approach to Machine Learning Operations (MLOps) that combines advanced technology with sustainability and resilience. RAI adapts to changing needs through continuous learning, self-healing, and secure operations. Its design focuses on reducing environmental impact with efficient data and model reuse while maintaining strong security. A key feature of RAI is lifecycle renewal, which helps optimise processes, extend the use of models, and keep systems adaptable and reliable. By using methods like transfer learning and model upcycling, RAI reduces resource-heavy tasks and ensures long-term effectiveness, making it a powerful tool for sustainable innovation.

RAI creates a circular AI system where lifecycle renewal is central to its success. This process extends the life of AI models with self-healing tools and performance monitoring. Outdated models are repurposed through automated systems, with transparency ensured by audit trails. Lifecycle renewal also includes strong cybersecurity, such as secure access controls and real-time threat detection. Sustainability is also considered in every step, using energy-efficient methods and tracking environmental impact. With lifecycle renewal at its core, RAI focuses on building AI systems that are efficient, secure, and environmentally friendly for the long term.

Activities of RAI group ranges from RAI research to application specific AI implementations in several regenerative innovation domains including digital energy/sector coupling, battery systems, power electronics, hydrogen and CCUS, green IIoT, sensing systems, operations research and strategy design.