Automated & Autonomous Experimentation

Self-driving laboratories - robotic systems guided by AI that can design, execute, and learn from experiments with minimal human intervention - represent one of the most tangible applications of AI in science. These systems combine robotic lab equipment with machine learning algorithms that decide what experiment to run next based on previous results, optimising the path toward a desired outcome. In materials science, chemistry, and biology, such systems can run experiments around the clock, testing hundreds of variations far faster than a human researcher could. The AI component is crucial: rather than simply running pre-programmed experiments, the system actively learns which parameters matter most and focuses its efforts accordingly. Early results are promising - autonomous systems have discovered new catalysts, optimised chemical reactions, and identified useful material compositions in days rather than months. The barriers to wider adoption include the high cost of robotic lab setups, the difficulty of encoding complex experimental procedures, and the need for human oversight when unexpected results occur. These systems work best for well-defined optimisation problems rather than open-ended exploration.