Forum

«Accelerating Chemistry with Cloud-based AI Driven Autonomous Labs»

Chemical innovations have fuelled a growth in society over the past century, from a fundamental understanding of how nature functions, to its application in the design of new medicines and materials.

Chemical innovations have fuelled a growth in society over the past century, from a fundamental understanding of how nature functions, to its application in the design of new medicines and materials.
Despite the progress in developing new reactions, the art of making molecules, however, has not changed significantly with the coming of the computer age. Research for materials and chemistry is typically carried
out manually, by trained chemists, using equipment that has not evolved significantly in the preceding decades.
In a continuously evolving scientific landscape, AI and automation are key requirements to increase the efficiency and safety of manual processes, allowing robots to autonomously execute tedious routines or even dangerous operations. IBM RoboRXN delivers on this vision with a novel cloud-based platform that combines
three technologies: artificial intelligence (AI), cloud, and automation. The core of the platform is comprised of AI models [1] [2] [3] [4], which assist the chemists by suggesting one or several synthetic routes to the desired target molecule, and can be edited in a human-in-the-loop framework to tailor the predictions to the
chemists needs. These models are trained using the information of millions of patents [5] and on-demand pipelines are already in place for tuning the machine learning algorithms with customers’ proprietary data.
Everything is backed-up by the IBM cloud infrastructure that manages the interaction of the AI model with the fully autonomous robotic system present at the IBM Zurich lab. AI, Cloud and Automation are fully integrated to deliver an end-to-end experience for the synthesis of new molecules and materials, which
serves to aid a chemist in their workflows.


References:
[1] P. Schwaller, T. Laino, T. Gaudin, P. Bolgar, C.A. Hunter, C. Bekas, A.A. Lee, ACS Central Science, 2019, 5, 1572-1583.
[2] P. Schwaller, R. Petraglia, V. Zullo, V.H. Nair, R.A. Haeuselmann, R. Pisoni, C. Bekas, A. Iuliano, T. Laino, Chemical Science, 2020, 11, 3316-3325.
[3] A. Toniato, P. Schwaller, A. Cardinale, J. Geluykens and T. Laino, Nature Machine Intelligence, 2021.
[4] A.C. Vaucher, P. Schwaller, J. Geluykens, V.H. Nair, A. Iuliano and T. Laino, Nature Communications, 2021, 12, 2573.
[5] Nextmove Software Pistachio. http://www.nextmovesoftware.com/pistachio.html, (Accessed Apr 02, 2020).

Abstract as Download


Speakers (1)

Amol Thakkar, IBM Research

Amol Thakkar, IBM Research

Manager