Merck and the Acceleration Consortium from the University of Toronto, Canada, have collaboratively made their AI-driven experimentation planner, Bayesian Back End (BayBE), available open-source on GitHub under an unrestricted Apache 2.0 license. This initiative merges Merck’s use cases with the Consortium’s expertise in self-driving labs, aiming to advance scientific discovery.
BayBE enables smarter experiments across Merck’s businesses. For example, it provides recommendations for optimized experiments and serves as the “brain” for automated equipment, which the company says increases efficiency and accelerates results. Traditional experimental design relies heavily on experimenter intuition, which leads to inconsistencies between labs, especially for complex campaigns. BayBE uses AI to overcome these challenges, reduce time and cost, and improve sustainability.
- Merck KGaA, Darmstadt, Germany
Also of Interest
An expanding compilation of articles concerning the intersection of artificial intelligence and chemistry
