Garbamonscommunity events · research brief
TALOS final event: how robots and AI could change solar maintenance
On 24 September 2026 in Lisbon, TALOS will present robotics and AI for land-based, floating, and agrivoltaic solar operations and maintenance.
The Lisbon event
TALOS will hold its final event in Lisbon, Portugal, on 24 September 2026. The event is in person only. The project page says the venue is still to be confirmed, so travellers should check the organiser's living event article before making plans.
The event will bring together the project consortium, Open Call innovators, policymakers, and the wider solar-energy community. It is intended to present results and the technical and policy lessons from a three-year Horizon Europe project that tested autonomous systems for solar operations and maintenance.
Why solar maintenance is a technology problem
Solar plants need inspection, cleaning, fault detection, and vegetation management. Some tasks are repetitive. Others expose workers to heat, electrical equipment, difficult terrain, water, or floating infrastructure. As the number of panels grows, operators need to find faults and plan maintenance without sending people everywhere at all times.
TALOS works across three settings: land-based photovoltaic systems, floating photovoltaic systems, and agrivoltaics. Its public materials describe drones, ground robots, floating vehicles, digital twins, AI-based anomaly detection, and multi-robot planning. The aim is human and robot collaboration, not a claim that people disappear from the system.
What the project is trying to prove
CORDIS lists project targets that include increasing plant performance by up to 10 per cent, reducing operation and maintenance costs by up to 5 per cent, saving up to 35 per cent of water in relevant tasks, and cutting worker risk exposure by up to 90 per cent in demonstration scenarios. These are targets for the project, not independent measurements of every solar plant or guaranteed commercial results.
That distinction matters. A robot can perform well in a test bed and still face problems with weather, communications, battery life, surface conditions, worker training, insurance, cybersecurity, or local rules. The final event is a chance to ask which results survived real operating conditions and what remains experimental.
Questions for the final demonstrations
- Which task is automated, and where does a human still approve, supervise, or intervene?
- How were faults and maintenance decisions validated against an independent baseline?
- What happens when a robot fails, loses its connection, or encounters a person or animal in the work area?
- How much water, energy, time, and replacement hardware does the system use over its full life cycle?
- Can the method work at small sites and agricultural sites, or only at large, controlled installations?
The environmental value is in the details
Automation is not automatically sustainable. Its value depends on whether it reduces hazardous work, material and water use, downtime, and unnecessary travel while keeping the solar plant reliable. It also depends on the land and water choices around the panels. A better maintenance robot does not resolve every question about habitat, food production, or floating infrastructure.
TALOS is scheduled to end on 30 September 2026. The useful legacy will be the public evidence, software, and methods that operators can test, along with an honest account of the limits. That is the standard to carry into future clean-technology claims.