IN Brief:
- AINETUS has moved into the LF Energy open-source ecosystem for wider grid-sector use.
- The tool supports AI-assisted power grid operation through visualisation, recommendations, and operator-in-the-loop workflows.
- The software was developed within AI4REALNET, which applies AI to electricity, railway, and air traffic network operations.
The AI4REALNET project’s AINETUS grid operation assistant has been made available as an open-source solution through LF Energy, creating a shared software route for AI-assisted electricity network operation.
AINETUS has been developed to support decision-making in power grid operations by combining AI components with operator-facing interfaces and visualisation tools. The system helps operators assess grid conditions and AI-generated recommendations in real time while keeping human supervision within safety-critical workflows.
Moving the software into LF Energy gives it a governance and maintenance structure beyond the current research programme. Open-source hosting also gives grid operators, research institutions, and technology developers a route to inspect, adapt, and contribute to the software rather than relying solely on closed proprietary development. AI4REALNET is scheduled to conclude in March 2027, making long-term stewardship part of the deployment pathway.
The electricity network use cases behind AINETUS focus on transmission grid operation. One use case uses SCADA data and energy management system tools to identify and manage congestion. Another addresses the transfer of AI assistance from simulation into real-world operation, including alerts to the operator when data limitations prevent full autonomy.
Its technical approach draws on supervised learning, reinforcement learning, human-in-the-loop decision-making, and autonomous AI systems operating under human supervision. That blend reflects the safety-critical nature of power network operation. Grid control cannot be treated as a conventional automation exercise; recommendations have to be bounded, explainable, and visible to the operators responsible for real-time decisions.
Electricity networks are becoming harder to operate as generation, demand, and flexibility become more distributed. Solar, wind, batteries, EV charging, heat pumps, interconnectors, and large flexible loads can all alter system conditions. Transmission and distribution operators need more advanced tools to interpret changing power flows, manage constraints, and maintain system stability without relying only on manual analysis or static planning assumptions.
AI tools are entering that environment with a narrow but useful role. Their strongest application is not the replacement of control-room personnel, but the acceleration of system-state analysis, constraint identification, and operational response testing. Poor data quality, model opacity, and recommendations that work in simulation but fail in abnormal conditions remain practical barriers to deployment.
AINETUS addresses part of that challenge by keeping human oversight within the workflow and flagging cases where data limitations restrict autonomy. That approach gives AI a supporting role inside power system operation rather than treating it as a standalone control authority.
The open-source route also changes the economics of digital grid operation. Proprietary tools will remain central to many control systems, but shared frameworks can reduce duplication, support technical scrutiny, and improve interoperability between research outputs and operational software. Open-source models have already been used across grid simulation, data exchange, and digital substation work; AI-assisted operation is a natural extension where cyber security, assurance, and governance are strong enough.
Other parts of the grid digitalisation stack are moving in the same direction. National Grid’s dynamic line rating rollout uses real-time sensor data to increase transmission capacity, while UK Power Networks’ Future Fleet project models the local network impact of electric HGV charging. Both developments point to a power system increasingly dependent on data quality, simulation, forecasting, and operational intelligence.
AINETUS adds a control-room layer to that shift. As grids become more constrained and more dynamic, software used to interpret system conditions will become as important as the physical infrastructure used to move power. The open-source release gives the industry a practical testbed for AI-assisted operation, with validation, cyber assurance, operator acceptance, and integration into live control environments now setting the pace of adoption.
