Nottingham AI Robotic Lab Automates Heat-Resistant Coating Discovery

Researchers at the University of Nottingham’s Center of Excellence in Coating and Surface Engineering have developed a self-driving laboratory that combines artificial intelligence, robotics and high-temperature processing to accelerate the discovery of next-generation heat-resistant coatings. The system automates key materials-development tasks, including precise liquid dispensing, chemical mixing, sample transfer, furnace processing, controlled cooling and sample retrieval, reducing the manual and repetitive steps traditionally required to test materials for extreme environments. The platform uses two robotic arms, a custom liquid-dispensing tool, laboratory equipment and purpose-built software, guided by active learning. This AI method analyzes results from earlier experiments to recommend which material compositions should be tested next, enabling a more efficient and data-driven search for promising candidates. Using the system, the team identified and prepared materials with a pyrochlore crystal structure as potential thermal barrier coatings, which are designed to limit heat transfer to underlying components in demanding applications such as aerospace engines, energy systems and high-temperature engineering. The work addresses a major technical challenge in autonomous laboratories: integrating chemical preparation, robotic handling and furnace-based materials processing into a single reliable workflow. Calibration tests showed accurate dispensing across different liquids, while demonstrations confirmed that the system could independently prepare precursor solutions, move samples through high-temperature processing and return processed materials for analysis. By connecting these stages, the researchers established a foundation for more reproducible and high-throughput materials development, reducing human intervention and improving experimental consistency. Professor Tanvir Hussain said the study highlights how laboratory robots can be made to operate reliably across multiple experimental phases, particularly where liquid handling meets high-temperature processing. Dr. Mohammadreza Lalegani Dezaki added that the project’s goal was to unify operations often performed separately into one coordinated robotic workflow, showing that complex materials preparation can be automated from start to finish. The findings were published in Engineering Applications of Artificial Intelligence and in a working paper on ChemRxiv.