Nottingham links AI materials discovery with a two-arm robotic laboratory

University of Nottingham researchers have connected two robotic arms, liquid dispensing and high-temperature furnace processing in a coordinated laboratory workflow. The system can prepare a precursor solution, transfer material for heating, manage cooling and retrieve the processed sample, according to the university’s 26 October account.[1]

The work sits alongside an AI-guided search for materials used in thermal barrier coatings—the protective layers that reduce heat reaching components in demanding environments such as aircraft engines. An active-learning approach helps identify which candidate compositions are worth investigating next.[1][2]

The difficult part is joining laboratory operations

A robot that performs one repeatable action is useful, but materials preparation involves several different conditions. Liquid handling needs calibrated quantities; sample transfer needs reliable positioning; furnace work introduces heat and cooling times. The Nottingham team’s contribution is to coordinate those stages with purpose-built tools and software.[1]

The university reports calibration experiments for dispensing different liquids, followed by demonstrations of the combined preparation and thermal-processing sequence. Connecting these operations reduces the manual hand-offs between stages and creates a more consistent record of how a sample was made.[1]

AI helps choose where to spend an experiment

Active learning uses the information already available to select a useful next experiment. In materials research, that can mean balancing a composition expected to perform well against one that would resolve uncertainty. The Nottingham account says the approach identified and prepared several candidates with a pyrochlore crystal structure for investigation as thermal barrier materials.[1][2]

Low thermal conductivity is relevant because a coating should slow heat transfer to the underlying component. It is one part of a demanding application: a promising material must also survive processing and service conditions. The announcement describes candidate discovery and laboratory preparation rather than qualification for an aircraft engine.[1]

A complete self-driving laboratory would connect candidate selection to synthesis, measurement and selection of the following experiment. The reported robotic work establishes an important preparation-and-processing segment; the researchers describe further integration with intelligent experimental planning and optimisation as an opportunity.[1][3]

That segment is substantial on its own. A repeatable route from a liquid recipe to a processed sample lets researchers compare results with a clearer account of how the material was produced. It also gives an AI planning system a more dependable physical process to call when the wider feedback loop is connected.[1]

Report an error or suggest a correction