Tiny robots                                                           01/08/2026
--------------------------------------------------------------------------------

I wrote parts of this blog post sometime last year but never published it. Just
rediscovered it a few days ago and decided to finish it.

Sometime in 2017, I came across this video.



This was at the height of my (first) obsession with origami, and it was also
right around the time I discovered what Arduino was, so naturally, I started
digging into it a bit more.

I spent some time trying to understand how the origami design worked, how they
got it to self fold, but most of all, how they made it move around and perform
tasks. Of course, I knew from the video (and the article that went with the
video) that it had a little magnet on its back and that was their way of
controlling its motion, but I figured maybe there was something about the
origami structure that allowed them to control it, like bending the joints to
propel it forward or pick up the little red cubes.

But after a while, I came to the conclusion that the origami structure just
allowed it to waddle from side to side. Nothing was actually bending, which
seemed to defeat the purpose of an origami robot. Why make it foldable, why have
so many joints, when none of them actually help the robot do things?

Plus, not knowing the first thing about magnetism, part of my brain was
convinced there was someone dragging a powerful magnet along the underside of
the table. Even though I knew this wasn't what they were doing, I decided they
were cheating by using a magnet, and valiantly set out to improve their robot by
giving it brains and controlling it by flexing the joints rather than relying on
the waddling motion their robot exhibited. I never was able to get anywhere with
it and did move on to other, more attainable quests, but not before I spent
quite a lot of time thinking about it and trying to fold various origami designs
that could function as an improved robot body.

I've had a lot of passing fancies like this over the years, but somehow this one
has managed to stick around. Nearly nine years later, I'm still pretty set on
making a tiny smart robot. And robots of this scale seem to be garnering more
attention recently.

Many of the recent innovations solve one of my issues with these original
origmai robots: no matter how they're controlled, many newer robots, like the
aerial robots at MIT's Soft and Micro Robotics Laboratory or the amphibious
millirobots at Stanford's Zhao Lab actually take advantage of their
flexibility for locomotion. But they still don't have brains of their own.
They're either tethered with wires or controlled magnetically -- which is fine
for some applications like surgery, where tasks can be performed in a relatively
controlled environment, but definitely doesn't work for tasks that involve
dynamic or uncertain environments.

 

Left: MIT's soft-actuated aerial robots (Soft and Micro Robotics Lab).
Right: Stanford's amphibious origami millirobots (Zhao Lab).
--------------------------------------------------------------------------------

I've always thought it would be cool to work with fingertip-sized robots. The
origami robot video from nine years ago convinced me of that. The idea that it
might be possible to squeeze everything -- computational power, sensing
capabilities, and more -- into a fully-functional robot no larger than a few
centimeters at most. That's exciting.

And also really difficult, but that's the best part.