The smallest robotic arm you can imagine is controlled by artificial intelligence

Researchers used deep reinforcement learning to steer atoms into a lattice shape, with a view to building new materials or nanodevices.
Schematic of atoms being moved with tweezers at the nano-scale

In a very cold vacuum chamber, single atoms of silver form a star-like lattice. The precise formation is not accidental, and it wasn’t constructed directly by human hands either. Researchers used a kind of artificial intelligence called deep reinforcement learning to steer the atoms, each a fraction of a nanometer in size, into the lattice shape. The process is similar to moving marbles around a Chinese checkers board, but with very tiny tweezers grabbing and dragging each atom into place.

The main application for deep reinforcement learning is in robotics, says postdoctoral researcher I-Ju Chen. “We’re also building robotic arms with deep learning, but for moving atoms,” she explains. “Reinforcement learning is successful in things like playing chess or video games, but we’ve applied it to solve technical problems at the nanoscale.” 

So why are scientists interested in precisely moving atoms? Making very small devices based on single atoms is important for nanodevices like transistors or memory. Testing how and whether these devices work at their absolute limits is one application for this kind of atomic manipulation, says Chen. Building new materials atom-by-atom, rather than through traditional chemical techniques, may also reveal interesting properties related to superconductivity or quantum states.

The silver star lattice made by Chen and colleagues at the Finnish Center for Artificial Intelligence FCAI and Aalto University is a demonstration of what deep reinforcement learning can achieve. “The precise movement of atoms is hard even for human experts,” says Chen. “We adapted existing deep reinforcement learning for this purpose. It took the algorithm on the order of one day to learn and then about one hour to build the lattice.” The reinforcement part of this type of deep learning refers to how the AI is guided—through rewards for correct actions or outputs. “Give it a goal and it will do it. It can solve problems that humans don’t know how to solve.”

Applying this approach to the world of nanoscience materials is new. Nanotechniques can become more powerful with the injection of machine learning, says Chen, because it can accelerate the parameter selection and trial-and-error usually done by a person. “We showed that this task can be completed perfectly through reinforcement learning,” concludes Chen. The group’s research, led by professors Adam Foster and Peter Liljeroth, was recently published in Nature Communications.


Chen IJ, Aapro M, Kipnis A, Ilin A, Liljeroth P, Foster AS (2022). Precise atom manipulation through deep reinforcement learning. Nat Comms. doi: 10.1038/s41467-022-35149-w


Finnish Center for Artificial Intelligence (external link)

The Finnish Center for Artificial Intelligence FCAI is a research hub initiated by Aalto University, the University of Helsinki, and the Technical Research Centre of Finland VTT. The goal of FCAI is to develop new types of artificial intelligence that can work with humans in complex environments, and help modernize Finnish industry. FCAI is one of the national flagships of the Academy of Finland.

Kvanttibitit. Kuva: Jan Goetz.


OtaNano is Finland's national research infrastructure for micro-, nano-, and quantum technologies

  • Published:
  • Updated:
URL copied!

Read more news

Sähkötekniikan opiskelijat Venla, Iikka ja Leevi. Kuva: Filmbutik.
Studies Published:

Sähkötekniikan päivä event gathered participants from all over Finland - watch the recording!

In the virtual event, electrical engineering students were discussing about their studies and student life at Aalto University.
50 shades of red cabbage. Photo by Aalto University, Rainer Paananen
Research & Art Published:

MMD CHEMARTS projects in 2022

CHEMARTS projects by members of the Multifunctional Materials Design research group in 2022
RDM & Open Science training
Research & Art Published:
Cooperation, Research & Art Published:

ACTOR Empowers Construction Workers with Process Automation

The ACTOR project brings together six organizations to automate the collection and use of real-time data for construction management. We talked with Aalto University, Carinafour, and VTT to learn about their aspirations for the project.