Materials Agents for Autonomous Research

View at publisher Summarise with VIRAT
Authors: Xue Jiang, Yanjing Su
Cited by: 0
References: 175

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This chapter explores the transformative potential of AI-powered innovation in materials science, focusing on the role of language models (LLMs) in autonomous research. The text discusses how traditional materials research, often reliant on theoretical exploration and empirical experimentation, can be accelerated through data-driven, AI-enabled approaches. Key developments include the integration of LLMs into laboratory workflows, autonomous experimental platforms, and the use of surrogate models for efficient exploration of material properties. The chapter outlines the architecture of autonomous materials research agents, comprising LLM-based modules for perception, decision-making, and execution, which collectively enable dynamic, closed-loop research frameworks. Core technical modules, such as natural language interfaces for equipment control, transformer models for experimental planning, and reinforcement learning for iterative optimization, are detailed. The chapter highlights the significance of these advancements for scientific innovation and industrial development, with specific examples like the Autonomous Materials Search Engine (AMASE) platform demonstrating real-time, closed-loop interactions between experiment and theory. The chapter explores the integration of large language models (LLMs) with autonomous agents and robotic laboratories in materials science research.

Fields of Study & Subjects

Machine Learning in Materials Science Artificial Intelligence Applications Scientific Computing and Data Management

Publication details

Journal / source
Wiley eBooks
Pages
449–504
Publication date
17 March 2026
Language
EN
Document type
Book chapter