Robot loose in the chem lab

This will either inspire or terrify you, depending on your view of such things.

British researchers have built an intelligent robot scientist that works in the lab on chemistry experiments. Not helping with experiments, but doing them by itself – making its own decisions about what to do when.

Credit: Benjamin Burger University of Liverpool

It has infinite patience, can think in 10 dimensions, keeps going for 21.5 hours a day, pausing only to recharge its battery, and can work in complete darkness, an advantage for carrying out light-sensitive photochemical reactions.

And it weighs in at 400 kilograms, so it’s probably always right in staff discussions.

“Our strategy here was to automate the researcher, rather than the instruments. This creates a level of flexibility that will change both the way we work and the problems we can tackle,” says the University of Liverpool’s Andrew Cooper, the lead author of a paper in the journal Nature.

In particular, he notes, the new technology could tackle problems of a scale and complexity that are currently beyond our grasp because they involve searching vast, unexplored chemical spaces.

Robots have been used in chemistry research before but are typically hardwired to a specific experiment. This 1.75-metre tall but unnamed robot roams the lab as required.

It can work with equipment designed for human operation because of its human-like dimensions and physical reach, Cooper says. It uses a combination of laser scanning coupled with touch feedback for positioning, rather than a vision system.

In a recent flurry of activity, it conducted 688 experiments over eight days, working for 172 out of 192 hours.

It independently carried out all tasks, from weighing out solids and dispensing liquids to running catalytic reactions and quantifying the reaction products. To do this, it made 319 moves and 6500 manipulations, travelling 2.17 kilometres.

According to Cooper and the man who built and programmed it, PhD student Benjamin Burger, its brain uses a search algorithm to navigate a 10-dimensional space of more than 98 million candidate experiments, deciding the best one to do next based on the outcomes of the previous ones.

It has already made its own breakthrough – discovering a new catalyst that is six times more active, with no additional guidance from the research team.

The biggest challenge, Burger says, was to make the system robust.

“To work autonomously over multiple days, making thousands of delicate manipulations, the failure rate for each task needs to be very low,” he says. “But once this is done, the robot makes far fewer mistakes than a human operator.”

There is no comment to date from the HR office.