AI4REALNET (AI for Real Networks)
Project description
AI for Real-World Network Operations
Artificial Intelligence (AI) technology has the potential to enhance the flexibility and resilience of critical network infrastructures to address global challenges like climate change, energy transition, increasing demand from mobility infrastructures, and digital transformation. However, AI faces several challenges: ensuring reliability, transparency, and ethical adherence to prevent errors and adversarial attacks; managing the complexity and uncertainty from aging assets, climate change, and rising demand in energy and mobility networks; enabling effective human-AI collaboration through reciprocal learning and integration of human knowledge; and overcoming scalability issues in AI methods like reinforcement learning (RL) when applied to large-scale infrastructures.
AI4REALNET project launched in September 2024 for a duration of 42 months and aims to create a comprehensive multidisciplinary approach by combining emerging AI algorithms, open-source AI-friendly digital environments, and socio-technical design of AI-based decision systems with human-machine interaction. This aims to enhance the real-time and predictive operation of network infrastructures. The project focuses on three critical infrastructures — electricity network, railway, and air traffic management — vital to Europe and identified as priority sectors in national AI strategies.
The AI4REALNET project is funded by the European Union within the framework of the Horizon Europe research and innovation programme (Grant agreement n°101119527).
Objectives
- To develop the next generation of decision-making methods, which aim at trustworthiness in AI-assisted human control with the resilience, safety, and security of critical infrastructures as core requirements
- To boost the development and validation of novel AI algorithms through existing open-source digital environments capable of emulating realistic scenarios of physical systems operation and human decision-making, enabling a direct assessment of AI-based decision quality
Implemented skills
- Data science and AI
- Human Machine interaction
- Systems Engineering
- Optimisation
Targeted markets
- Electrical networks and energy
- Transport and mobility
- Air traffic management