A breakthrough discovery in developing autonomous nanobots

29. 6. 2021 | Study in the Czech Republic

data-nodes-23688-gridworld-v4-1800x1200-resize-w1280-h0Scientists have been trying to construct intelligent microscopic machines for quite some time now and it seems they’re finally a step closer to their goal. The researchers from Charles University and Leipzig University have recently made a breakthrough by developing a system that combines machine learning and precise control over microscopic particles.

Their research was published in the prestigious magazine Science Robotics in which the team presented their study called Reinforcement Learning with Artificial Microswimmers. They managed to develop special microscopic particles and then precisely control their movement. It’s the first step towards nanobots which, for example, will be able to distribute healing drugs or repair damaged tissues.

“Our method uses symmetrical particles that are possible to heat locally via a laser,” Dr Viktor Holubec from the Faculty of Mathematics and Physics explains. “Particles then move away from heated spots because of to the flowing liquid and the effect known as thermophoresis.”

This new experimental method was combined with the reinforcement learning running in a computer which controls particles with a laser beam. “Brains of our particles aren’t in particles themselves yet,” Dr Holubec comments. “Our next phase is to miniaturise these components and implement them into the particles.”

The newly developed system should also help the scientists to understand how the microcosmos actually works because it’s still mostly an uncharted territory. “It can help us uncover key aspects of the dynamics of such particles. For instance, we discovered similarities with organic bacteria,” he adds.

Are you a science nerd? Do you feel inspired? Study at the Faculty of Mathematics and Physics!