Skip to main content

Microchip Helps Tiny Robots Move Autonomously and Untethered

A robot invisible to the naked eye has a tiny microprocessor “brain” onboard. | Cornell University, Noël Heaney

Building self-sufficient robots is no easy task. One reason is because autonomous robots need to have onboard control, using an internal processing system to undertake actions. Onboard control is even more difficult to implement on a micro scale, where systems must be downsized to near-invisibility while retaining their functionality.

Featured in Science Robotics, a new type of microchip can successfully overcome this size-dependent hurdle, enabling onboard control in untethered robots only a tad bigger than the width of a human hair.

"We developed a technique for combining silicon electronics 'brains' and microactuator 'legs' to build robots that are smaller than you can see with your naked eye," said Michael Reynolds, postdoctoral associate at Cornell University and first author of the study.

The new research could help advance microrobotics research with uses in environmental fieldwork, medical interventions, and other fields.

Making Movement on a Micro-Scale

In the past few decades, roboticists have been experimenting with control systems for tiny robotic systems. But most of these inventions were guided by outside digital control systems that responded to signals such as magnetic fields or chemical cues.

"Having onboard digital control electronics essentially gives a microrobot a 'brain' — it knows internally how to make itself move," said Reynolds. He and his team compare past microrobots to puppets because those machines relied on external manipulation.

To tackle the onboard control challenge in microrobotics, the group designed digital electronic chips tiny enough to be incorporated into robots smaller than 250 micrometers. The chips' size meant they were limited in their ability to power actuators, or components in robotic systems that turn energy to motion. But this was not a deterrent for the scientists.

"We knew right away that it was possible in principle to build the sorts of robots we show in this work," said Reynolds.

From Principle to Production

In 2020, the scientists had published on a new type of microactuator, called surface electrochemical actuators (SEAs. Most recently highlighted in Science Robotics last year, the SEAs are responsive to lower voltages and currents.

"The challenge was to find a way to bring two complicated fabrication processes — microelectronic circuits and SEAs — together into one process that would allow you, at the end, to release and operate a microrobot with a circuit onboard," said Reynolds.

The group began working with a coalition of semiconductor factories called X-FAB Silicon Foundries to combine microchip and microactuator into an autonomous microrobot.

Using techniques akin to those used for making cell phones and computers, the scientists created three types of robots. These robots all walked with different gaits faster than 10 micrometers per second. External solar cells powered the onboard chips as they controlled microrobots' motion through pre-programmed electrical signals. One type, called a dogbot, could even internally process and respond to optical commands. The microrobots' ability to autonomously move and also interact with surrounding stimuli showcases their potential for future use in the real world.

"We are excited about applications in medicine and environmental remediation for microrobots with onboard electronics — things like detecting and selectively breaking down pollutants or transporting cells through the body or performing microscale surgery," said Reynolds. He noted that the next steps for the team include optimizing the microchips' responsivity and revising the microrobots' actuators to support other types of movement.

[Credit for associated dogbot image: Reynolds et al., Sci Robotics 10.1126/scirobotics.abq2296]


Abigail Eisenstadt

Related Focus Areas