Periodic Reporting for period 1 - TALOS (roboTics and Artificial intelligence Living labs improving Operations in PV Scenarios)
Periodo di rendicontazione: 2023-10-01 al 2025-03-31
As we steadily grasp towards the EU’s carbon neutrality goal for 2050, the Photovoltaic (PV) sector remains at the core of the renewable energy market. However, even though the total number of PV installations continues to grow, some O&M activities are still costly, hazardous and/or time-consuming, dragging this growth down. With the funding from the Horizon 2022 programme, TALOS (roboTics and Artificial intelligence Living labs improving Operations in PV Scenarios) aims to improve these activities via the introduction of state-of-the-art Robotics, Artificial Intelligence (AI), Digital Twin (DT) and Internet of Things (IoT) technologies in the Solar Parks’ O&M workflow.
There are several activities to be tackled by the autonomous/semi-autonomous robotic solutions to be developed during TALOS: inspection of the park infrastructure, cleaning of the PV panels, mowing of shade-inducing vegetation and AI-systems for fault detection and for recommendation of appropriate O&M activities. To this effect, three distinct types of solar parks will be used for the proof of concepts – conventional/Land PV (LPV), Floating PV (FPV) and AgriPV (APV) – where each site will have different requirements, restrictions and ecosystems.
Every solution will be connected to a standardized platform, which will allow for seamless cooperation between them, ensuring maximum efficiency. Inspection robots will provide detailed reports on the state of the solar park, adding to the existing data taken from Supervisory Control And Data Acquisition (SCADA) systems, fueling the DT. By supplying this information to the recommendation engine, a robotic-coordination tool will be capable of optimally schedule and trigger maintenance activities, such as the automatic cleaning and mowing robots.
This initiative is expected to be able to boost PV park performance ratio by 10%, reducing the risk exposure of O&M workers by 90% and the human burden of monitoring crops (in the APV case) by 90%. It is also expected to lower the resources required to perform some O&M activities, such as by reducing the amount of water used in cleaning and farming (the latter in the case of APV) by 35%, and to reduce the amount of greenhouse gas emissions by over 450 ton per year.
TALOS will thus greatly boost the efficiency and improve the logistics of PV parks, enabling a continued investment in the PV sector.
There are several activities to be tackled by the autonomous/semi-autonomous robotic solutions to be developed during TALOS: inspection of the park infrastructure, cleaning of the PV panels, mowing of shade-inducing vegetation and AI-systems for fault detection and for recommendation of appropriate O&M activities. To this effect, three distinct types of solar parks will be used for the proof of concepts – conventional/Land PV (LPV), Floating PV (FPV) and AgriPV (APV) – where each site will have different requirements, restrictions and ecosystems.
Every solution will be connected to a standardized platform, which will allow for seamless cooperation between them, ensuring maximum efficiency. Inspection robots will provide detailed reports on the state of the solar park, adding to the existing data taken from Supervisory Control And Data Acquisition (SCADA) systems, fueling the DT. By supplying this information to the recommendation engine, a robotic-coordination tool will be capable of optimally schedule and trigger maintenance activities, such as the automatic cleaning and mowing robots.
This initiative is expected to be able to boost PV park performance ratio by 10%, reducing the risk exposure of O&M workers by 90% and the human burden of monitoring crops (in the APV case) by 90%. It is also expected to lower the resources required to perform some O&M activities, such as by reducing the amount of water used in cleaning and farming (the latter in the case of APV) by 35%, and to reduce the amount of greenhouse gas emissions by over 450 ton per year.
TALOS will thus greatly boost the efficiency and improve the logistics of PV parks, enabling a continued investment in the PV sector.
The TALOS project kicked off by laying down a solid strategic and methodological foundation. A unified structure was created to define use cases for land-based, floating, and agrivoltaic solar systems. These included detailed descriptions, goals, conditions, involved actors, and technical connections. The process was refined through workshops and meetings with pilot owners and stakeholders. A Social Innovation Roadmap (SIR) was co-developed to ensure that user needs were considered throughout the development timeline. System specifications and architecture were mapped out, defining how components interact and setting interoperability standards. Key performance indicators (KPIs) and a monitoring plan were also established to track both technical and social outcomes. A demonstration framework and Open Call challenge definitions were put in place to guide pre-pilot and pilot planning in line with stakeholder expectations.
The TALOS platform was designed as a complete solution—from data ingestion to task management and visualization. Built around early use cases and system requirements, it brings together various data sources, algorithms, and robotic solutions. Two APIs were developed: one for managing pilot site data and another for controlling robots. Modules for data ingestion and governance were also added. For fault detection and recommendations, two AI-driven approaches are being developed using Digital Twins—one based on SCADA historical data and another combining real-world and synthetic fault data. This work has involved signal analysis, machine learning reviews, PV plant modeling, and initial model training. A method for integrating data from different pilot sites is also being finalized.
Robotic solutions for monitoring and operations & maintenance (O&M) have made strong progress. Requirements were defined for robotic systems across all PV types. These include a multi-modal inspection setup with two UAVs, a surface vehicle, a UGV for agricultural monitoring, a solar panel cleaning robot, a vegetation management robot, and a floating PV cleaning platform.
The project also ran a successful Open Call to attract innovative companies. It received 166 applications, far exceeding expectations, and selected 12 beneficiaries (10 SMEs and 2 startups). These companies are now developing their solutions under a Support Programme guided by TALOS partners.
Pilot site planning and logistics have been moving forward steadily. Integration roadmaps were created using pilot specifications and Gantt charts. Regular coordination meetings helped align technology development with pilot site conditions. Open Call beneficiaries included pilot demonstrations in their mentoring plans. Weekly and monthly meetings helped refine requirements and testing logistics. A first end-user feedback questionnaire was created for initial field tests, and a draft training plan is being prepared based on partner input and feedback.
Work has also progressed on evaluating the environmental and socio-economic impact of TALOS technologies using Life Cycle Assessment (LCA) and Life Cycle Costing (LCC). Methodologies were defined, including goal setting and tailored data collection strategies. A System Overview Survey was created for all pilot types. The survey for the floating PV pilot is complete and ready for delivery, while those for land-based and agrivoltaic pilots are still being drafted.
Finally, dissemination efforts are helping boost the visibility and impact of TALOS. This includes publishing project materials, engaging stakeholders online, and participating in key events. The exploitation strategy is focused on turning technical results into market-ready solutions. The Open Call’s strong outreach contributed to its success, and the team is now working closely with partners to coordinate training materials and engage end-users—ensuring effective knowledge transfer and long-term impact.
The TALOS platform was designed as a complete solution—from data ingestion to task management and visualization. Built around early use cases and system requirements, it brings together various data sources, algorithms, and robotic solutions. Two APIs were developed: one for managing pilot site data and another for controlling robots. Modules for data ingestion and governance were also added. For fault detection and recommendations, two AI-driven approaches are being developed using Digital Twins—one based on SCADA historical data and another combining real-world and synthetic fault data. This work has involved signal analysis, machine learning reviews, PV plant modeling, and initial model training. A method for integrating data from different pilot sites is also being finalized.
Robotic solutions for monitoring and operations & maintenance (O&M) have made strong progress. Requirements were defined for robotic systems across all PV types. These include a multi-modal inspection setup with two UAVs, a surface vehicle, a UGV for agricultural monitoring, a solar panel cleaning robot, a vegetation management robot, and a floating PV cleaning platform.
The project also ran a successful Open Call to attract innovative companies. It received 166 applications, far exceeding expectations, and selected 12 beneficiaries (10 SMEs and 2 startups). These companies are now developing their solutions under a Support Programme guided by TALOS partners.
Pilot site planning and logistics have been moving forward steadily. Integration roadmaps were created using pilot specifications and Gantt charts. Regular coordination meetings helped align technology development with pilot site conditions. Open Call beneficiaries included pilot demonstrations in their mentoring plans. Weekly and monthly meetings helped refine requirements and testing logistics. A first end-user feedback questionnaire was created for initial field tests, and a draft training plan is being prepared based on partner input and feedback.
Work has also progressed on evaluating the environmental and socio-economic impact of TALOS technologies using Life Cycle Assessment (LCA) and Life Cycle Costing (LCC). Methodologies were defined, including goal setting and tailored data collection strategies. A System Overview Survey was created for all pilot types. The survey for the floating PV pilot is complete and ready for delivery, while those for land-based and agrivoltaic pilots are still being drafted.
Finally, dissemination efforts are helping boost the visibility and impact of TALOS. This includes publishing project materials, engaging stakeholders online, and participating in key events. The exploitation strategy is focused on turning technical results into market-ready solutions. The Open Call’s strong outreach contributed to its success, and the team is now working closely with partners to coordinate training materials and engage end-users—ensuring effective knowledge transfer and long-term impact.
Yet to obtain comparable (with SoTA) results. Foreseen for the last year.