Researchers have introduced a new technique that enables “self-driving laboratories” to gather data at unprecedented speeds, potentially revolutionizing materials discovery research. The advancement, detailed in Nature Chemical Engineering, significantly accelerates the process while reducing costs and environmental impact.
Self-driving laboratories are robotic systems that use machine learning and automation in chemical and materials sciences to expedite material discovery. These labs allow algorithms to predict subsequent experiments based on previous data.
“Imagine if scientists could discover breakthrough materials for clean energy, new electronics, or sustainable chemicals in days instead of years,” says Milad Abolhasani, ALCOA Professor of Chemical and Biomolecular Engineering at North Carolina State University. “This work brings that future one step closer.”
Previously, self-driving labs using continuous flow reactors depended on steady-state flow experiments. This method involved mixing precursors and allowing reactions to occur before characterizing the product. However, this required waiting for reactions to complete.
The new approach uses dynamic flow experiments where mixtures are continuously varied and monitored in real time. “Rather than running separate samples through the system…we’ve created a system that essentially never stops running,” Abolhasani explains.
Collecting more data improves the performance of these labs. “The most important part of any self-driving lab is the machine-learning algorithm…This streaming-data approach allows the self-driving lab’s machine-learning brain to make smarter, faster decisions,” Abolhasani adds.
The researchers found that their dynamic flow system generated at least 10 times more data than traditional methods over the same period. This enabled them to identify optimal material candidates immediately after training.
“This breakthrough isn’t just about speed,” Abolhasani notes. “By reducing the number of experiments needed…advancing more sustainable research practices.”
The paper was co-authored by various researchers from North Carolina State University and Tecnologico de Monterrey with support from several grants including those from the National Science Foundation.
