WHY THIS MATTERS IN BRIEF
It doesn’t look like we humans are having much luck when it comes to solving climate change, so scientists are now getting AI onto the case.
Interested in the Exponential Future? Connect, download a free E-Book, watch a keynote, or browse my blog.
In what could be a historic moment researchers and scientists around the world are now, apparently, able to have days off, and the scourge of climate change is about to be solved because we now have an Artificial Intelligence (AI) and robots on the case. Well, that’s what the scientists behind a new company who are busy building yet another Robo-Scientist are hoping will happen after they began using a state of the art AI and a robot in a lab to help develop new materials with amazing new properties.
In their small lab in Massachusetts a robot arm dips a pipette into a dish and transfers a tiny amount of bright liquid into one of many receptacles sitting in front of another machine. When all the samples are ready, the second machine tests their optical properties, and the results are fed to a computer that controls the arm. AI software then analyses the results of these experiments, formulates a few hypotheses, and then starts the process over again. Humans are barely required, so pack up your PhD certificates and take your Furby home with you Mr and Mrs Scientist.
In all seriousness though as AI tackles more scientific challenges, from predicting the outcome of organic reactions to helping develop new products, like this NASA lander and more, the setup in question which was developed by a startup called Kebotix hints at how machine learning and robotic automation may be poised to revolutionise materials science in the next decade.
The company also believes its new approach will help find new compounds that could, among other things, absorb pollution, combat anti-biotic resistant infections, and serve as more efficient optoelectronic components. The company’s software learns from 3D models of molecules with known properties.
Software algorithms are already used to design chemical compounds and materials, but the process is slow and crude. Usually, a machine simply tests slight variations of a material, blindly searching for a viable new creation. Machine learning and robotics on the other hand, as many people hope, could make the process much faster and more effective. And Kebotix is one of several startups working on this idea.
“The goal is to use machine learning to generate candidate materials. Discovery is too slow,” says Jill Becker, CEO of Kebotix. “You have an idea for a material, you try to make it, and you test it. A few ideas are tested, with even fewer results.”
Kebotix uses several machine learning methods to design their new novel chemical compounds. The company feeds molecular models of compounds with desirable properties into a type of neural network that learns a statistical representation of those properties. This algorithm can then come up with new examples that fit the same model.
Kebotix also uses another network to weed out designs that stray too far from the original and are therefore likely to be useless. Then the company’s robotic system tests the remaining chemical structures. The results of those experiments can be fed back into the machine learning pipeline, helping it get closer to the desired chemical properties. The company dubs the overall system a “Self-driving lab.”
Christoph Kreisbeck, the company’s chief product officer, says Kebotix will start out working with molecules for electronic applications and then try to tackle new polymers and alloys.
“The AI predicts and plans what to do next, the robot automation system very rapidly tests our new molecule,” Kreisbeck says. “The machine can learn from the database and make a better decision for the next round.”
Klavs Jensen, a professor in MIT’s chemical engineering department near where Kebotix are based, but who wasn’t involved in the research, is all in when it comes to developing automated approaches to devising useful new chemicals, including methods that combine machine learning and robotics. Adding that the big catch is that such methods tend to require huge quantities of data which is generally time consuming and difficult to collect – something that also becomes more challenging as the materials get more complicated.
“You can definitely do a lot,” Jensen says. “But like anything else, it’s about the quality of the data.”
Jensen says that automation, already commonplace in the pharmaceutical industry, will become increasingly important in materials research.
“It won’t replace the expert,” he says, “but you’ll be able to do things a lot faster and accelerate the rate of [materials] innovation by multiples.”