Get ready for a revolution in the world of laboratory experiments! The future is here, and it's robotic!
Imagine a chemistry lab where four tall robots, each standing at 1.75 meters, seamlessly navigate through the workspace, effortlessly transporting materials between automated workstations. These robots, guided by an intelligent AI system, make decisions on the next steps based on real-time data, even during the quiet hours of the night when human chemists are fast asleep.
Professor Andy Cooper, a pioneer in this field, introduced robotics to his lab a decade ago. His groundbreaking research, published in Nature, showcases how AI-driven robotics can significantly boost productivity. "At 3 a.m., while everyone is catching some Z's, the robot has already conducted 50 experiments and is ready to make its next move based on fresh data," he explains.
These lab robots, adapted from industrial units by Kuka of Germany, move with precision between automated reactors and analytical equipment. They work alongside human researchers, thanks to sensors that ensure a safe shared space. The potential applications are vast, ranging from drug discovery to developing new materials for carbon capture.
Liverpool University, recognizing the success of this innovative approach, has announced plans to establish a £100 million AI-driven materials chemistry research hub. But Professor Lee Cronin at Glasgow University is also making waves in this field. His spinout company, Chemify, has raised an impressive $43 million in 2023 and an additional $50 million this year. Cronin's vision is ambitious: "Chemify aims to design and create any molecule on demand, covering the entire spectrum of chemistry, from drug discovery to new catalysts and electronic materials. Our next step is nothing short of a revolution in digitizing and automating chemical discovery and manufacturing."
While both UK pioneers are taking different paths, they complement each other. "My approach is to use industrial robots to integrate labs, which I believe will prove scalable and cost-effective. Lee, on the other hand, is building bespoke facilities suited for specific applications. There's room for both strategies," Cooper notes.
In June, Chemify opened its first Chemifarm, a fully automated 2,000 sq m facility in Glasgow, valued at £12 million. "We aim to work with 20 partners by next year, and then we'll expand and build Chemifarms globally," Cronin says. Beyond the robotic hardware, Chemify has developed a programming language called chi-DL, which Cronin hopes will become the industry standard for digital chemistry.
According to Cooper, labs worldwide are rapidly adopting robotics and AI. "There are at least 30 to 40 labs using these systems now, with some making significant investments, particularly in China, which leads the world in robotics production."
Sami Haddadin, a leading figure in scientific robotics, moved from the Technical University of Munich to set up a lab at Mohamed bin Zayed University of Artificial Intelligence in Abu Dhabi. He advocates for a collaborative global network of AI-driven labs, where data and computational resources can be shared to tackle complex scientific problems beyond the capabilities of individual institutions. While international collaboration is still in its early stages, Haddadin emphasizes the need for standardized data formats, hardware protocols, and interoperable software to make this vision a reality.
"A network of robotic laboratories will generate an unprecedented amount of data, even compared to particle physics and astrophysics. We need infrastructure to ensure this data is analyzed, stored, and distributed globally with proper access," he says.
Rob Brown, head of the scientific office at Sapio Sciences, believes AI-driven automation will transform research methodology. "Today, research is typically 20% virtual design and 80% experiments. This will shift to perhaps 80% virtual and 20% experimental, but automated labs will always be essential."
Everyone involved in lab automation agrees that AI will enhance, not replace, human talent. "Scientists currently spend a significant amount of time on tasks that don't contribute directly to their project's goals. Their role will become more engaging, focused on in-depth scientific knowledge and innovation, rather than data entry and repetitive tasks in the lab," Brown explains.
For Cronin, the human contribution lies in creativity. "I've seen no evidence that AIs are creative. Humans are here to stay and will remain at the heart of scientific progress, just without the need for hands-on, toxic chemical exposure."
Cooper summarizes this new dynamic as "hybrid intelligence." He adds, "Human and artificial intelligence are often portrayed as opposites, but in reality, we'll want to combine human hypotheses and conjecture with automated reasoning using large language models. Human reasoning is deeper but slower, while automated reasoning is faster but shallower. The winning combination is to bring these two together."
This exciting development promises to revolutionize laboratory experiments, making them more efficient, creative, and, most importantly, safer for human researchers.
