Ada: Accelerating materials research with flexible automation and AI

13 May2020

Image source: the Berlinguette group via

An artificially intelligent, self-contained, and self-driving laboratory named Ada can explore formulations for a type of thin-film material common to advanced solar cells and consumer electronics. Advanced solar cells are used to power sustainable energy production and enable more efficient energy storage, including longer-lasting batteries. As the need to find clean energy solutions becomes increasingly urgent, it will be imperative to leverage tools like Ada to work toward finding new materials to power industrial and consumer technologies.

A team co-led by SBQMI investigator Curtis Berlinguette (Professor in the Department of Chemistry and the Department of Chemical & Biological Engineering at UBC), Jason Hein (Associate Professor in the Department of Chemistry at UBC), and Alán Aspuru-Guzik (Professor in the Department of Chemistry and Department of Computer Science at the University of Toronto) has demonstrated that it is possible for self-driving laboratories to develop and test thin films in an automated fashion, with resulting data used to inform the design of subsequent experiments. Named for British mathematician and early computer scientist Ada Lovelace, and borne out of the Paris Climate Agreement and Mission Innovation, Ada conducts experiments autonomously and “learns” how to optimize thin-film materials in a continuous loop, refining its process through machine learning.

Project Ada was first funded by Natural Resources Canada in 2018. By 2019, the Ada team had built the first fully functional Ada platform; they have since authored a milestone proof-of-concept paper showing that Ada can be effective in the search for new material formulations with desirable properties.

“Ada was able to determine that a particular annealing temperature produced a material that performed better than we had predicted; the result is a discovery that we would not likely have made using conventional methods,” explained Berlinguette. The paper, published in the journal Science Advances on May 13, offers findings that speak to the possibilities of using autonomous laboratories to refine and test organic and inorganic materials of relevance to materials science and clean energy technologies.

Thin films are chemical deposits applied to materials such as glass or silicon that improve the properties of those materials, and are useful in applications ranging from solar technology to nanomedicine. Thin films also offer the opportunity to explore materials with unique and unconventional properties, such as those that show promise for quantum applications.

Translating new functional materials from lab to market has typically followed a rigorous process that can take decades; Project Ada has shown for the first time that self-driving laboratories can be leveraged in an effort to accelerate this timeline for thin-film materials. In the future, the Ada team is aiming for an ambitious but plausible tenfold acceleration, bringing a theoretical material from experimental to commercially viable in just a few years. 

Reference: MacLeod, B. P.; Parlane, F. G. L.; Morrissey, T. D.; Häse, F.; Roch, L.; Dettelbach, K. E.; Moreira, R.; Yunker, L. P. E.; Rooney, M. B.; Deeth, J. R.; Lai, V.; Ng, G. J.; Situ, H.; Zhang, R. H.; Aspuru-Guzik, A.*; Hein, J. E.*; Berlinguette, C. P.* “Self-Driving Laboratory for Accelerated Discovery of Thin-Film Materials.” Science Advances 2020. DOI: 10.1126/sciadv.aaz8867

  • Collaborations
  • Research