featured-image-standard-abstract-lines-dots-nodes-connecting

Automated system generates actuators that no human can do by hand

Published: 16:58, July 13, 2019

Researchers designed an automated system using 3-D prints that generates robotic parts for novel tasks. The parts, called actuators, are optimized according to a huge number of specifications. The automated system does automatically what a human could never do by hand.

When designing an actuator requires too many variables for a human being to test by hand, the automated system can step in.

Actuators are devices that control robotic systems mechanically in response to electric signals.

The researchers, from the Massachusetts Institute of Technology (MIT), Cambridge, USA, and University of Grenoble Alpes, Inria, France, wrote about their study and achievements in the journal Science Advances (citation below).

The authors demonstrated the system by creating actuators that show different black-and-white images at various angles.

For example, one actuator, when laid flat, portrayed a portrait of the Dutch painter Vincent van Gogh. however, when its activated, tilted at an angle, it portrays ‘The Scream,’ by the Norwegian painter Edvard Munch.

The research team also 3-D printed floating lilies with petals. The arrays of hinges and actuators respond to magnetic fields (run through conductive fluids) by folding up.

Actuators made from three different materials

According to an MIT press release:

“The actuators are made from a patchwork of three different materials, each with a different light or dark color and a property — such as flexibility and magnetization — that controls the actuator’s angle in response to a control signal.”

A computer program initially breaks down the actuator design into millions of voxels, i.e., three-dimensional pixels. Each voxel can be filled with any of the three materials.

The software then runs millions of different simulations in which voxels are filled with different combinations of materials. It eventually finds the right placement of each material in each voxel so that two different images can be generated. At one angle you see one image, while at another angle you get a different image.

A 3-D printer then creates the actuator by placing the correct material into the right voxel. It does this layer after layer.

Goal – automatically find the best design

First author, Subramanian Sundaram PhD, said:

“Our ultimate goal is to automatically find an optimal design for any problem, and then use the output of our optimized design to fabricate it. We go from selecting the printing materials, to finding the optimal design, to fabricating the final product in almost a completely automated way.”

Dr. Sundaram was a Ph.D. student at MIT. He is a Postdoctoral Associate at Boston University and the Wyss Institute at Harvard University.

In an Abstract preceding the main article in the journal, the authors wrote:

“The topology optimizer assigns materials for individual voxels (volume elements) while simultaneously optimizing for physical deflection and high-resolution appearance.”

“Unifying a topology optimization-based design strategy with a multimaterial fabrication process enables the creation of complex actuators and provides a promising route toward automated, goal-driven fabrication.”

Citation

“Topology optimization and 3D printing of multimaterial magnetic actuators and displays,” Subramanian Sundaram, Melina Skouras, David S. Kim, Louise van den Heuvel and Wojciech Matusik. Science Advances, 12 Jul 2019: Vol. 5, no. 7, eaaw1160. DOI: 10.1126/sciadv.aaw1160.

Christian Nordqvist Avatar

Other News

UK late-payment bill would cap terms and strengthen suppliers’ rights

Sep 30, 2026

Sumitomo completes battery-recycling plants designed to recover four metals

Sep 30, 2026

Smarter controls could make room for 330 GW on existing power grids

Sep 30, 2026

Biosimilars cut into Humira sales and offer savings on costly medicines

Sep 30, 2026

Global wealth hit a record, but much of the gain was on paper, MGI says

Sep 29, 2026

Progress closes $400 million Domo deal to add AI data platform

Sep 29, 2026

SOCAR and Comstock set a $1.65 billion framework for Haynesville gas investment

Sep 28, 2026

HCLSoftware plans Robotiq.ai deal to connect AI agents with older business systems

Sep 28, 2026

ONS research says payroll records could sharpen the UK labor market picture

Sep 28, 2026

Select Water agrees $700 million deal for Pilot Water’s oilfield network

Sep 27, 2026

US firms are pulling back investment in China, Federal Reserve analysis finds

Sep 26, 2026

NetApp plans PEAK:AIO acquisition to scale storage for larger AI clusters

Sep 26, 2026

Bank AI use was linked to a smaller share of small-business lending, Fed study finds

Sep 26, 2026

Iridium shareholders approve Rocket Lab takeover: what still has to happen

Sep 25, 2026

Akamai’s 11.6 billion dollar Anthropic deal ties cloud revenue to a 5.5 billion dollar buildout

Sep 25, 2026

Bentley completes 350 million pound Crewe investment as it unveils its first electric vehicle

Sep 25, 2026

Falling birth rates did not reduce total output in historical data, NBER study finds

Sep 25, 2026

Cheaper renewable power does not solve the capital problem for poorer countries

Sep 25, 2026

Facial payments may feel novel, but money worries can curb repeat use

Sep 24, 2026

Precision farming cuts water use while raising crop yields, study finds

Sep 24, 2026