Robots are getting ever closer to outperforming humans in all aspects of life – even when it comes to learning about how the world works.
Researchers at Toyota are using artificial intelligence to speed up the discovery of the ideal chemical makeup for electric car batteries.
AI-powered robot arms engineered by the team place precise drops of chemical reagents in test tubes under the guise of human supervisors.
Over the next few months, the machine intelligence behind the system will take over the planning of experiments as well, according to Toyota.
Researchers said the ‘robot graduate student’ will decide how to modify the concentrations of the ingredients it’s testing without the need for human assistance.
‘It’s automating not only the manual part of doing the experiment but also the planning part,’ Brian Storey, the Toyota Research Institute scientist leading the project, told Bloomberg.
Dr Barnabás Póczos, a machine learning researcher at Carnegie Mellon University in Pittsburgh who is also working on the project, added: ‘I can easily imagine cases in which AI would recommend experiments to try to synthesise a chemical molecule that you wouldn’t think possible, but the AI will be able to do it.’
Automakers have been investing heavily in developing new batteries and fuel cells to increase the range of electric vehicles.
Mr Storey said Toyota’s AI is helping to identify new materials for batteries and fuel and run computer tests to narrow down the field for simulation tests by researchers.
The research is in-part pursuing a replacement for platinum as a fuel-cell catalyst.
‘We don’t have a ton of platinum on this planet and it costs a lot money,’ he told Reuters in 2017.
‘Platinum is a great catalyst, but is there another compound out there that uses little platinum or no platinum at all?’
Toyota is investing around £25 million ($35 million) in its North American research arm, the Toyota Research Institute (TRI).
The Institute is collaborating with a number of US academic institutions including the University of Michigan and the Massachusetts Institute of Technology, and British material sciences company Ilika.
A number of other projects around the world are using artificial intelligence to drive research efforts, Bloomberg reports
AI designed to identify and categorise patterns has been deployed to identify wild dolphin calls from hydrophone recordings.
Similar software has been used by astronomers to detect the dull glow of planets in telescopic images of distant galaxies.
The discovery of the Higgs boson, the so-called God particle, in 2012 utilised an algorithm that searched through billions of particle tracks produced within Switzerland’s Large Hadron Collider.
According to Mr Storey, AI could one day help scientists boil down the fundamental principles of physics to reveal the secrets of the universe.
He told Bloomberg: ‘People have wondered if you could have the computer automatically figure out the principles underlying physics.
‘I don’t think we’re going that far out now.’