Using Robotics to Enhance Laboratory Experiments

Imagine running 3,600 synthesis experiments in a single day and learning something fundamental from each one. That's the power of Self-Driving Labs (SDLs). When I bring up SDLs with colleagues in biotech, the concept clicks instantly. The field already leans on robotic assays, high-throughput screening, and informatics pipelines. So it feels like a natural extension to integrate AI/ML for closed-loop optimization. But when I speak with peers in chemistry and materials science, I often encounter more hesitation. The idea of handing over experimental decisions to machines can feel at odds with a culture shaped by first-principles reasoning, serendipity, and manual iteration. That's why I was so intrigued when I met Prof. Milad Abolhasani at the NIST AI for Materials Science (AIMS) workshop last month. His lab’s work offers a compelling example of the SDL vision for materials discovery, bridging the gap between possibility and practice. From catalysis optimization to quantum materials, Abolhasani’s team is building SDL platforms that not only automate experiments, but also accelerate understanding. A recent standout is Rainbow: a multi-robot, AI-powered system that autonomously explores the synthesis of metal halide perovskite nanocrystals. By combining miniaturized batch reactors, real-time spectral feedback, and multi-objective Bayesian optimization, Rainbow executed thousands of synthesis trials in a single day: mapping Pareto fronts, uncovering structure–property relationships, and scaling up the best results. Maybe now is the time to imagine what SDLs could mean for materials innovation in your organization. 📄 Autonomous multi-robot synthesis and optimization of metal halide perovskite nanocrystals, Nature Communications, August 22, 2025 🔗 https://lnkd.in/euz72pXe

Researchers at the University of Liverpool have developed AI-driven mobile robots capable of conducting chemical synthesis research with remarkable speed and efficiency, as demonstrated in a study published in 𝐍𝐚𝐭𝐮𝐫𝐞. These 1.75-meter-tall robots can autonomously perform experiments, analyze results, and decide on the next steps in exploratory chemistry, handling tasks as effectively as human researchers but much faster. Led by Professor Andrew Cooper, the project focused on using robots to address challenges in three areas of chemical synthesis: drug discovery, supramolecular chemistry, and photochemical synthesis. The robots’ AI allows them to make swift, complex decisions, such as whether to proceed with a reaction, which might take a human hours to assess. Although these robots currently lack the intuitive "Eureka!" moments of human chemists, they execute repetitive tasks with high accuracy, and there is potential for further enhancement using AI models to access scientific literature. This technology could transform industries by expediting drug synthesis and materials science research, possibly scaling up to large robot teams for extensive industrial applications. Read more: https://lnkd.in/ehYQPbcP

🚀 High throughput experiment changing science The pre-defence by Elham Foadian in Mahshid Ahmadi group summarized her work on the robotic synthesis and characterization of hybrid perovskites - and was an opportunity to reflect on how automated synthesis changes how we do science. I recall being a part of inorganic and solid state chemistry groups in Russia in 90ies, and extensive collaborations with solid state chemists when in US in 2000-2010. At this time, exploring (synthesis, structural characterization, property measurements) of a single binary or ternary phase diagram was a good chunk of a thesis - literally year of work. Plenty of time to learn and internalize the science behind materials from multiple papers, thinking, and so on - and decide what to do next In Elham's presentation, the core numbers of explored points in concentration spaces measure 10s of thousands. Of course, these sample relatively low dimensional linear cross-sections of composition spaces as limited by robotic synthesis, but sheer numbers are astounding. And it's not only composition - it's stabilities, phase evolution, and multiple other manufacturing related parameters. Now it is possible to map these spaces in days, and in principle hours. Which in turn requires new ways of learning how summarize and interpret information obtained from such measurements, and plan experiments between dimensions of chemical spaces rather than single points. Materials science cannot change overnight, but the way it changed over the last 3 years is already more than the previous 30. And the process just starts. Clearly we need better strategies for experiment planning with these capabilities - filling the gap between human execution and often intuitive human planning. So far very few work in this area - but clear opportunity!