Scientists are revolutionizing catalyst research by combining AI-driven automation with human expertise, proving that the smartest labs aren’t just robotic—they’re collaborative.
Perspective: Role of the human-in-the-loop in emerging self-driving laboratories for heterogeneous catalysis. Image Credit: Gorodenkoff / Shutterstock
A self-driving laboratory integrates AI with lab automation and robotics. The AI plans experiments, which are executed in increasingly automated (robotized) modules. In practice, this process occurs in active learning loops, where the data from the last loop is used to refine a machine learning model. The AI then uses this model to plan the subsequent experiments in the next loop. However, rather than fully replacing human researchers, SDLs are designed to augment their role, ensuring that AI remains adaptable to human modifications. This way, only those syntheses, characterizations, and tests that are most informative are conducted on the basis of all prior collected data. Simultaneously, the automation enhances throughput, reproducibility, and safety – promising a significant acceleration compared to traditional human-led development processes.
Early implementations of this innovative concept for discovering improved catalyst materials often focus on replacing human tasks with synthesis robots. However, merely automating existing procedures does not necessarily improve efficiency. Dr. Scheurer and Prof. Reuter instead emphasize that the most time-consuming step of such catalysis research is typically the explicit testing of the materials. Given the increasing importance of sustainability, the degradation behavior of the materials in the reactor must be monitored over a long time. Therefore, throughput improvements are more likely to be achieved by developing new experimental designs, including the use of proxy experiments that can provide equivalent insights faster, rather than merely automating existing procedures.
Especially when throughput remains limited, AI's role in experiment planning is crucial. The fewer loops that need to be executed, the better. Also, humans will continue to play a vital role for the foreseeable future. While current AIs can determine optimal experiments within a given overall framework, they cannot yet question this framework or even redefine the scientific questions themselves. The design space is not rigid; instead, it evolves as new insights emerge, requiring AI to be flexible and adaptive. For the time being, these creative tasks remain the domain of humans, necessitating a human control function within the loops.
This research, published in the journal Nature Catalysis, highlights that breakthroughs in self-driving laboratories will not come from full automation alone but from rethinking experimental workflows and maintaining human involvement.
The authors thus advocate the "human-in-the-loop" principle and analyze its implications for AI development in SDLs. Rather than simply aiming for full automation, they emphasize the importance of integrating humans as ‘data-informed scientists,’ who guide AI-driven processes and refine experimental goals based on newly acquired data. The AIs must be capable of responding flexibly, robustly, and assessably to human modifications of the loop structures – a methodological challenge currently addressed by ongoing research in the Theory Department—Fritz Haber Institute of the Max Planck Society.
Source:
Journal reference: