Advincula Leads Project on Self-Driving Labs of the Future

A human scientist only has so much bandwidth when it comes to performing experiments. To synthesize a polymer or organic material, a chemist will need to repeat the process to try and get a higher yield. It's impossible to stay awake 24 hours a day for multiple months to speed up the process.

But what if scientists had an assistant that never slept and could process everything at a much higher rate? What if that assistant could continuously analyze vast amounts of data, identify gaps in the experiment, send instructions for recalculating or conducting a new simulation, and reset the experimental setup for the next phase?

That's the idea behind self-driving laboratories (SDLs) of the future, ones that utilize artificial intelligence and machine learning to facilitate new scientific discoveries, optimize process engineering, and rapidly accelerate research and development (R&D) for manufacturing in the fields of chemistry and materials science.

Rigoberto Advincula, UT-ORNL Governor's Chair of Advanced and Nanostructured Materials, recently published an article for The Bridge, a quarterly journal published by the National Academy of Engineering, exploring the potential of AI/ML to drive innovation in polymer materials and examine the role of SDLs in addressing challenges in polymer science and engineering.

Advincula has led a $3 million project funded by Oak Ridge National Laboratory (ORNL) and the U.S. Department of Energy, which is centered on self-driving labs and has been part of the INTERSECT initiative of ORNL. SDLs have the potential to be intelligent operating systems for conducting experiments, embodied by an automated laboratory-operated robot and controlled by computers that autonomously set or reset experimental parameters.

"This is looking at the laboratory of the future, where you have scientists with a lot of tools being able to do their discovery of radiopharmaceuticals and critical materials faster," Advincula said. "They will have the ability to make new polymers with a higher performance or better properties much faster than has ever been done before."

Faster Pharmaceutical Development

The pharmaceutical industry was one of the first to invest in AI/machine learning and flow chemistry, a technology that runs chemical reaction in a continuously flowing stream rather than in batch production.

"That means we can do the manufacturing of drugs at a faster scale, or even onshoring, because there's a lot of talk about how drugs are no longer made in the US," Advincula said. "SDLs may be able to help us to get cheaper drugs by accelerating the science through AI."

There has been speculation that SDLs will eventually operate independently of a human being present at any stage of the research and development process.

"But in reality, it's still a pipe dream. What this self-driving lab is actually proving is that highly skilled scientists are valuable," Advincula said. "But with AI and machine learning, you have a lot of tools for automation and accelerating the discovery process, meaning a scientist can do a lot of experiments in a relatively short time and with very targeted hypotheses or very refined problems, which gets you the solution faster."

That is good news to companies with R&D budgets that need to generate significant revenue from a new project by rapidly producing products.

"This becomes very cost effective as a tool for performance," Advincula said. "It's really accelerating science or optimization, where it can save energy costs and laboratory costs, because instead of doing trial and error, you get the right experiment, right result in one try."

Advincula believes introducing UT students to the potential of self-driving labs will position them for success in the career field.

"The students right now need to get skills in this area of AI/machine learning as part of their training as an undergrad," he said. "Being exposed to this now prepares them for jobs of the future."

More information:
www.nae.edu/340925/Frontiers-i … ng-Laboratories-SDLs

Provided by University of Tennessee at Knoxville