World’s smallest autonomous robots are 'smaller than a grain of salt,' cost one penny apiece — researchers expect new micron-scale fully-programmable robots to be used in medicine, microscale manufacturing, and other areas
Researchers tout potential medical and microscale manufacturing applications.
Fully programmable, autonomous robots “smaller than a grain of salt” have been developed in a collaborative effort between research teams from the University of Pennsylvania and the University of Michigan. Claimed to be the world’s smallest of their kind, these nanoscale robots are expected to find applications in medicine, microscale manufacturing, and other areas. Remarkably, the researchers contend that these robots will cost only one penny each to manufacture.
Penn Engineering was responsible for the physical structure and features of these microscopic robots, while Michigan scientists gave the robots their ‘brains.’ The Michigan lab involved in this collaboration also holds the record for creating the world’s smallest computer.
Barely visible robots move using ion propulsion
Let’s look at Penn Engineering’s work first. Scientists there designed the physical aspects of the tiny robots, each measuring about 200 x 300 x 50 micrometers. This is similar to the scales of many microorganisms, and as such, they face different physical challenges than the robots with which we are most familiar.
With the challenges of drag and viscosity that such tiny robots will face, it was decided to make them ‘swimmers.’ However, to keep them small, the scientists designed an entirely new propulsion system so that moving parts (such as limbs) wouldn’t be needed for locomotion.
Interestingly, these robots generate an electric field that “nudges ions in the surrounding solution,” allowing them to “swim.” Adjusting this ion field can help robots move in complex and/or coordinated patterns. Having no moving parts also makes them highly durable; for example, they can be picked up with a micropipette without being damaged.
Another challenge of their tiny size was the power supply. The design that was adopted uses solar power, with the solar cells making up the majority mass of the robot body. Even though solar was maximized, it yields only 75 nanowatts of power, underscoring the importance of efficiency.
Communicating with light and a waggle dance
To be fully programmable, autonomous robots, these tiny machines need on-board processing, and this is where the expertise of the University of Michigan was applied.
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
According to the source blog, the robot's physical components get their smarts courtesy of a highly efficient processor, memory, and sensors. To program the robots, pulses of light are used – thus the solar panels do double duty. Meanwhile, to retrieve data from the robots and their sensors, they have been programmed to perform a honeybee-like waggle dance using their propulsion systems.
These robots can be deployed in their hundreds to get their tasks completed. Their autonomous operational life is measured in months, thanks to solar power and their level of efficiency. The first test sample carried a temperature sensor, which is good for all kinds of analytical tasks. However, it would not be challenging to switch the sensor.
Despite all this advanced technology at play, the researchers say that these fully programmable, autonomous robots “can be fabricated cheaply at scale,” costing only a penny each to produce. What we are seeing now is “just the first chapter” of micro-robots, promise the researchers.
Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

Mark Tyson is a news editor at Tom's Hardware. He enjoys covering the full breadth of PC tech; from business and semiconductor design to products approaching the edge of reason.
-
hotaru251 and thus begins the micro robot assassination part of history.Reply
mix em in w/ a person drink/food and they'll never know. -
Geef Maybe heat based power since someone would eat/drink them. Just imagine a little steam engine with a drill driving around your body making holes. Your body quickly patches them up but when there are 10k of them all doing the same thing...Reply
Or assassinate someone by giving them a ton of them right before an MRI. Thousands come tearing out of a person's body all at once. Eww. -
bit_user Reply
No, stomach acid would destroy them and they won't get enough light in your digestive tract to power them. Furthermore, they (or their breakdown products) would be easily detectable, which is usually something you don't want in a poison.hotaru251 said:and thus begins the micro robot assassination part of history.
mix em in w/ a person drink/food and they'll never know. -
bit_user Reply
Not plausible.Geef said:Maybe heat based power since someone would eat/drink them. Just imagine a little steam engine with a drill
If all you guys want to do is just poison someone, there are lots of biological and chemical agents that would be much better at the task.Geef said:driving around your body making holes. Your body quickly patches them up but when there are 10k of them all doing the same thing...
The only role I see robotics potentially having is if you fed someone miniature enteric-coated capsules that contain a poison and could be opened via RF signal (or, maybe they'll open unless they keep receiving the RF signal!). Then, we enter James Bond villain territory, where you can inform someone they've just been fed a poison with no antidote you'll open the capsules if they don't do as you say. Still, these would stay in the digestive tract, and would therefore be subject to removal by one means or another. They'd also be subject to the limits of battery power.
You need robots for that... why?? Just put iron filings in someones food. Cheap and potentially even smaller than these robots.Geef said:Or assassinate someone by giving them a ton of them right before an MRI. Thousands come tearing out of a person's body all at once. Eww. -
Tanakoi Reply
Stomach acid doesn't break down silicon. Any other breakdown products are those which are likely to be found in the body in minute quantities anyway. Obviously such nanobots would require a power source besides solar, but chemical sources abound in the body.bit_user said:No, stomach acid would destroy them and they won't get enough light in your digestive tract to power them. Furthermore, they (or their breakdown products) would be easily detectable
Eh? The human body can't absorb elemental iron; it has to be in the ferrous state ... and stomach acid does a relatively poor job of that. Even pure ferrous iron would take 10+ grams to be a lethal dose (several thousand times the mass of one of these robots), and elemental iron would likely take 100g or more. I'm pretty sure I'd notice a tenth of a kilo of iron filings floating around my vodka martini.bit_user said:You need robots for that... why?? Just put iron filings in someones food. Cheap and potentially even smaller than these robots.
BTW, the SF author Vernor Vinge focuses on some of the astonishing (and still plausible) aspects of the multitudinous applications of such nanoprocessors in his books The Peace War and A Deepness in the Sky. -
bit_user Reply
Why do you think they're entirely silicon?Tanakoi said:Stomach acid doesn't break down silicon.
True. In bots made exclusively for medical purposes, I'd expect it should be possible to do something like that.Tanakoi said:Obviously such nanobots would require a power source besides solar, but chemical sources abound in the body.
Who said anything about absorbing? just having it in your gut would be enough to have a really bad day in a MRI scanner.Tanakoi said:Eh? The human body can't absorb elemental iron; it has to be in the ferrous state ... and stomach acid does a relatively poor job of that.
We were talking about food, not beverages.Tanakoi said:I'm pretty sure I'd notice a tenth of a kilo of iron fillings floating around my vodka martini. -
Tanakoi Reply
Did you not read my post? I specifically mentioned other breakdown products. In any case, if one was designing a nanobot for ingestion, coating it entirely with a thin silica layer is trivial, and wouldn't affect operation.bit_user said:Why do you think they're entirely silicon?
You're not going to poison someone unless it's absorbed. And if it's not absorbed, why would you be going for an MRI?bit_user said:
Who said anything about absorbing? just having it in your gut would be enough to have a really bad day in a MRI scanner.
The OP specifically mentioned food and drink. Regardless, I think I'd notice a tenth of a kilo of iron filings whether it was in dim sum, Dover sole, or a Denver omelette.bit_user said:We were talking about food, not beverages. -
bit_user Reply
If you read the post I was responding to, Geef talked about dosing someone who was about to have a MRI scan (for presumably unrelated or otherwise contrived reasons).Tanakoi said:You're not going to poison someone unless it's absorbed. And if it's not absorbed, why would you be going for an MRI?
Again, you're thinking about a toxic dose, not Geef's scenario. Also, the iron could be coated so it doesn't have that characteristic iron taste.Tanakoi said:The OP specifically mentioned food and drink. Regardless, I think I'd notice a tenth of a kilo of iron filings whether it was in dim sum, Dover sole, or a Denver omelette. -
Geef Reply
We all know my post is the best just because you can imagine a little fat robot dude driving a steam engine drill tooting the horn! :pGeef said:Just imagine a little steam engine with a drill driving around your body making holes. -
Tanakoi Reply
Got it. The fault indeed lies with my failure to encompass the thread, not yours.bit_user said:Again, you're thinking about a toxic dose, not Geef's scenario. Also, the iron could be coated so it doesn't have that characteristic iron taste.