IBM’s RoboRXN helps scientists research new molecules while working from home.

The time and cost incurred to develop a new material or molecule can take as long as 10 years and $ 10 million. A lot of actions in the chemical research are repetitive, or cumbersome because they rely on a trial and error practice. At IBM (NYSE: IBM) Research Europe, scientists have reinvented chemical research and molecular discovery by incorporating artificial intelligence, cloud technology, and robotics into the mix. The result: RoboRXN. (MIT Technology Review)

In 2017, scientists at IBM Research Europe hit upon the idea of predicting chemical reactions using machine learning models. The effort was dubbed “RXN for Chemistry.”

It was a runaway hit when it was launched on the IBM Cloud in August 2018 as a free service.

Even today, RXN For Chemistry is a highly dependable data-driven method for forward reaction prediction with more than 90% accuracy.

RoboRXN

In 2019, the scientists moved onto the creation of RoboRXN.  It is a platform that has “machine learning algorithms autonomously designing (AI) and executing (Automation) the production of molecules in a laboratory remotely accessible (Cloud) with as little human intervention as possible.”

The process is revolutionizing chemistry, dragging it out of its traditional procedures and reshaping it as a high-tech activity. It can play a dramatic role in the discovery of new drugs to tackle health problems like the global COVID-19 pandemic. Consider this:

1. A scientist is working at home and wants to test a new multiple
2. He logs on to RoboRXN through a web browser
3. On a screen they can draw the molecular structure they have in mind
4. The platform now uses its machine language training to predict the ingredients required
5. It also sequences their mixing
6. The platform thereupon sends the entire instructions for execution to a remote, automated lab
7. After the experiment, the platform sends a report to the scientist.

Implications

“Just imagine if an automated system like RoboRXN could help chemists cut the discovery period of a new treatment for COVID-19 or any other virus in half,” writes Teodoro Laino of IBM. “Or what if RoboRXN could help accelerate the development of a fertilizer which doesn’t require consuming 1-2% of the world’s annual energy supply for its production?”

“The possibilities are endless when it’s humans + machines.”

Related Story:  Artificial Intelligence: A Robotic Scientist That Works 21.5/7    

Image: Screenshot from IBM video