An international research team has developed an AI-driven robotic platform that autonomously designs, fabricates, and optimizes perovskite solar cells, completing the full experimental workflow in a closed-loop system. Using the platform, the researchers fabricated and tested more than 50,000 devices, achieving efficiencies of up to 27%.
An international research team has developed an AI-driven robotic platform capable of autonomously designing, fabricating, and optimizing perovskite solar cells.
“At the core of the study is the idea that robotic experimentation should do more than automate repetitive operations,” the researchers said in a statement. “Formulas and parameters are encoded into machine-readable recipes, translated into robot-executable commands, and then returned as structured feedback after fabrication and characterization. In this way, the system establishes a closed-loop workflow that links recommendation, execution, validation, and model improvement.”
Using the system, the researchers have fabricated and tested 50,764 devices. It is powered by a recipe language model (RLM) that encodes information from around 60,000 perovskite solar cell-related publications released over recent decades, as well as data generated by the platform during device fabrication. These inputs are processed through a seven-layer AI architecture comprising recipe learning, recipe generation, dataset construction (RecipeQA), fine-tuning, reasoning, evaluation, and optimization.
Automated fabrication is initiated after the reasoning stage, where new experimental recipes are proposed. Eleven robotic boxes then carry out synthesis, device fabrication, and characterization tasks while simultaneously generating a digital twin of the process. The setup includes 101 functional units, more than 1,500 components, and over 4,300 controllable parameters.
Boxes 1–3 handle chemical storage, solid sampling, and liquid dispensing. Boxes 4–11 are used for spin-coating, antisolvent application, thermal annealing, laser processing, device transfer, vacuum exchange, and thin-film deposition. These latter units are also equipped with cameras, sensors, and actuators for in situ characterization, feeding data back into the model’s evolution loop.
Overall, the researchers describe the workflow of the robotic system as progressing through four stages: an initial phase of broad, largely unguided exploration of perovskite formulations , a second stage introducing additives and self-assembled monolayers (SAMs) to enhance crystallization and interfacial properties , a third stage incorporating surface passivation to reduce defects and improve performance , and a final stage combining SAM-based hole transport layers with targeted additive and passivation strategies.
“In stage I, without interface or additive engineering, the power conversion efficiency ranges from 0% to 17.4%. The incorporation of SAMs and additives in stage II narrows the distribution and increases efficiency to around 23%,” the results showed. “In stage III, interfacial post-treatment passivation leads to a further improvement, reaching 25.6%. The final configuration in stage IV delivers an efficiency of 27.0% (certified at 26.5%).”
The researchers stated that the main innovation of their study lies in combining three advantages within a single closed-loop AI–robotics system. They described it as enabling the controlled robotic fabrication of complete perovskite solar cell devices, alongside robotic characterization that transforms high-throughput experimental results into structured evidence related to underlying mechanisms. They further noted the inclusion of a domain-specific RLM that is continuously trained to improve recipe recommendations, mechanistic understanding, and subsequent robotic execution.
The system was described in “Agentic Robotic Boxes for Perovskite Solar Cell Fabrication with Recipe Language Model,” published in Engineering. Scientists from the Hong Kong Polytechnic University, Swiss Federal Institute of Technology in Lausanne, China’s Wenzhou Institute of Technology, University of Nottingham Ningbo China, Shenzhen University of Advanced Technology, North China Electric Power University, Zhejiang University, Peking University, and the University of Oxford in the United Kingdom have contributed to the research.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.