🤖 AI Summary
Low accuracy and poor experimental efficiency in materials science image analysis hinder scientific discovery. Method: This paper proposes a multi-agent AI collaboration framework featuring a novel dual-large-model structured debate mechanism integrating ChatGPT and Gemini, enabling interpretable and robust visual understanding through complementary reasoning. The approach combines a multi-agent architecture, prompt-engineering–driven debate protocols, and domain-specific fine-tuning for materials imagery. Contribution/Results: On materials phase identification, the method significantly improves classification accuracy; particle counting error is reduced by 32%; cross-task generalization is empirically validated; and the system is successfully deployed in the real-world experimental platform CRESt. This work overcomes inherent limitations of single-model approaches and establishes the first large-language-model–based collaborative paradigm for scientific image analysis grounded in structured model debate.
📝 Abstract
The Copilot for Real-world Experimental Scientist (CRESt) system empowers researchers to control autonomous laboratories through conversational AI, providing a seamless interface for managing complex experimental workflows. We have enhanced CRESt by integrating a multi-agent collaboration mechanism that utilizes the complementary strengths of the ChatGPT and Gemini models for precise image analysis in materials science. This innovative approach significantly improves the accuracy of experimental outcomes by fostering structured debates between the AI models, which enhances decision-making processes in materials phase analysis. Additionally, to evaluate the generalizability of this approach, we tested it on a quantitative task of counting particles. Here, the collaboration between the AI models also led to improved results, demonstrating the versatility and robustness of this method. By harnessing this dual-AI framework, this approach stands as a pioneering method for enhancing experimental accuracy and efficiency in materials research, with applications extending beyond CRESt to broader scientific experimentation and analysis.