Skip to main content

6 Arms, 3 Thumbs: I Tested The Future of Robotic Limbs

WIRED Editor Amit Katwala traveled far and wide for a hands-on look at the future of robotic artificial limbs. Watch as he puts a wearable set of arms inspired by traditional Japanese puppetry through the paces with Dr. Masahiko Inami at the University of Tokyo. Can Amit intuitively manipulate his newly added arms to pick up objects or write with a marker? Then, we visit with Dani Clode and Tamar Makin at the University of Cambridge to get a grip on the technology behind a 3D-printed sixth digit designed to give you two working thumbs on one hand. Will these robotic augmentations be the path to superhuman performance? Dani Clode Design / www.daniclodedesign.com / instagram: @dani_clode The Plasticity Lab / www.plasticity-lab.com / @plasticity-lab.bsky.social MRC Cognition and Brain Sciences Unit, part of Cambridge University: Instagram:@mrccbu / @cambridgeuniversity Director: Anna O'Donohue Director of Photography: Thomas Beswick; Giles Cahalane Editor: Matthew Colby; Anna O'Donohue Host: Amit Katwala Expert: Dani Clode; Masahiko Inami; Tamar Makin Line Producer: Joseph Buscemi Associate Producer: Paul Gulyas Production Manager: Alexandra Dawson Herren Production Coordinator: Ericka Gourgues-Lutran Assistant Camera: Tokiomi (Toki) Tsuta Sound Mixer: Skirmantas (Spider) Ivanauskas; Michael Panayiotis Production Assistant: Grace O'Connor Post Production Supervisor: Christian Olguin Post Production Coordinator: Rachel Kim Supervising Editor: Eduardo Araujo Assistant Editor: Justin Symonds

Released on 05/09/2025

Transcript

Sometimes we could all use an extra hand or thumb,

so here I am with both

testing the capabilities of these wearable robots.

How easy are they to use? Can I hold things?

Will the movements feel natural? What are they for?

And will wearing these make me superhuman?

I'm Amit Katwala.

I'm a writer and editor for Wired,

and I've always been fascinated

by how technology can enhance human performance.

So when I heard that a team of researchers

at the University of Tokyo

were working on a set of wearable robotic arms,

I just had to go and find out more.

The design of the arms was inspired in part

by traditional Japanese puppetry.

They're intended to enhance human capabilities,

much like an instrument.

I asked one of the creators of the arms,

Professor Masahiko Inami, to tell me more.

This is a small copy of this movement.

So physical computing to avoid collision.

Also, it's very intuitive. It's easy to understand.

Now this puppet master has sensors to measure joint angle

with five potential measures for each arm.

These sensors are sensing signal.

And so based on this signal,

these arms are following this motion perfectly.

So the data from the sensors

goes into the 3D model in the laptop

and then that then controls

[Professor Inami] the bigger arms? Yes.

So the key components of the system

are the backpack, the sockets, then the motors

[Professor Inami] in the arm? Yes.

[Amit] And then the motors

[Professor Inami] in the wrists? Yes.

That kind of brings it all together?

Yes. And so each joint is connected with carbon fiber.

[Amit] So that's almost like the bones of the suit?

[Professor Inami] Yes, they're like bone.

[Amit] Professor Inami says that the robot arms

could one day be used for sport,

medicine, or rehabilitation.

So we have proposed the superhuman sport.

Maybe in future we attach or detach

this type a superhuman needs to play a new type of sport.

This is one application.

Also, we are able to find some serious application

like assisting some medical surgery.

This hand hold my wrist,

so it can also teach me how to move my arm.

So from remote side or from computer side.

It also help us conducting rehabilitation

or also acquiring some new type of skills,

such as martial arts or some other things.

Now that I understand the basic mechanics of the suit,

it's time to test it out.

So it's pretty heavy, and when I move around,

I can kind of feel it moving around behind me,

but it's actually kind of comfortable.

I look pretty cool.

I look like I've merged with a robot, which is amazing.

[upbeat music]

My silhouette looks mental. It's so cool.

Now Professor Inami is going to demonstrate

how this all works using the puppet master.

Hi. You can shake hand.

Hi, this is so cool.

I can feel it like vibrating and stuff.

It's quite strange to have these limbs

moving around you without your control.

You can kind of hear the shoulders vibrating

as they lift the arms up and down.

You can really hear it like whirring and like, you know,

you can really hear the motors working as it moves.

I think it really makes you aware

of like how much force is being put through

just to lift up these arms.

It sounds kind of mechanical,

but also like kind of, I don't know, biological.

Almost like there's some animal like purring back there

or something like that.

Quite unnatural to shake hands with yourself. There we go.

Yeah. Yeah.

So humans have got this ability called proprioception,

which is your ability to know where your limbs are in space

without having your eyes open.

It gives you the ability to touch your nose for instance

without having your eyes open.

We're gonna see if I can tell where these robotic limbs are

with my eyes shut as professor Inami moves them around.

Okay, I think my left top arm is moving up and down.

I think my bottom right arm is moving.

Okay, I think left and right arms are both moving now.

Maybe it's top left and bottom right.

I think it's here maybe. Perfect.

Okay, so I can definitely feel myself leaning this way,

and I could feel vibrations over my right shoulder,

so I think it's this arm.

I think from the way that my body is now angled,

I'm thinking that it's the bottom right arm.

So maybe like that.

Both. Yeah, that makes sense, right?

So you can see how off balance I am

because like both arms are moving, and that's like what?

That's like four kilos of extra arms,

so it makes sense that I'm kind of counterbalancing a bit.

Now that I've tried the arms

with Professor Inami controlling their movement,

I'm going to try them for myself.

The controller feels very light compared to the arm,

which is obviously a lot heavier and feels much more robust,

so I'm being quite careful with how I move it.

I feel like I should be.

I'm going to try and touch my nose, see if that works.

Hi. Yeah, it definitely feels different.

I feel I'm much more like tentative

about moving it myself actually,

like I almost got more comfortable

with other people moving it around.

But with this, I find if I make a sudden motion

with my actual hand, I could end up getting,

you know, slapped in the face with a 3D printed hand,

which is not what you'd want.

Finally I wanted to test how well the robot arms

were able to pick up items.

My job here was to try and collaborate with the robot,

which was easier for some objects than for others.

We're gonna try and grab

this little soft bird thing here, level one.

And then I'll take that.

Thank you very much. Success. [laughs]

For the next level, we're gonna try and pick up this ball.

We might need two hands for this one. Alright.

Oh, oh, oh, oh come on. We can do it.

Yes. Success. [laughs]

We're gonna try and pick up the pen,

and then the robot's gonna hand me the pen.

I'm gonna take the lid off,

and then the robot's gonna try and write something

with the pen.

So you need this gripper here.

No, not success.

That's it. Yes, yes, yes, yes, yes.

No. Just cheat a little bit.

I think if you got it like perfect, there you go,

if you got it like perfectly centered.

Okay, right. Okay, here we go.

[laughs] Artists, I wouldn't worry about robots

taking your job just yet.

Overall I'm really impressed,

and I think with a bit of practice it could get even better.

And now it's back to London. Thank you.

After trying the robotic arms in Tokyo,

I headed closer to home to Cambridge

to meet with Dani Clode

and the team that is working on a third thumb.

That's right, a third thumb. Whoa.

Dani Clod is an augmentation and prosthetics designer

at the University of Cambridge's plasticity lab.

Yeah, so this is my third thumb.

So starting on the hand piece, there's the thumb obviously.

So it's a flexible thumb.

It's completely 3D printed,

and there's kind of three main elements.

It's got a rigid hand piece with flexible straps,

and then that's connected to two motors

that I wear on my wrist, a very similar position to a watch.

I mean you can see those are kind of pulling the thumb.

And that's connected to this wearable up here.

This is the battery, which is replaceable.

And then this PCB is wirelessly connected

to what I'm wearing in my shoes and around my ankles.

[Amit] Obviously I had to try it for myself.

You've got two pressure sensors,

one underneath each big toe,

and it's just basically a pulley system.

It's a completely flexible 3D print, so it's very compliant.

Now press your toes down one at a time.

Whoa. Yikes.

[laughs] So that's your left toe.

That's so cool.

[Dani] And then your right toe.

Yeah, so I can like move it.

So, yeah.

So give a kind of really, really slow delicate press.

Yeah, you'll see that there's a proportional control.

Yeah, yeah, yeah.

So you have a lot of control over it.

Or you can go fast, do a fast quick press. Yeah.

It's quite powerful isn't it?

Yeah, [laughs] it does surprise people.

It's so cool.

The speed is controlled by how hard you press down.

So that is pretty quick,

and you can do like a bit more of a slower,

like more delicate movement.

So actually trying to like move them in tandem

and getting the pressure right

I think is gonna be the real challenge

of like learning how to use this.

On the left foot, I've actually got

quite a good level of control over like the pressure.

It's really powerful.

Look, and you can see if I squeeze like that,

you can see how much pressure

it's putting on the palm of my hand there, look.

[mechanical thumb whirring]

Sounds like a little robot

or like droid or something from Star Wars.

You can kind of see the 3D printed

like corrugation on the inside,

and then these tendons are like super thin.

What are they made of, Dani?

[Dani] It's kind of like a technical fishing line.

Okay, yes. A fishing line, yes.

I guess it's strong, and light, and like pretty robust.

Now that I know how to control it,

I wanted to test out the capabilities of this robotic thumb.

In the team's research,

they found that generally people can use it functionally

within the first minute,

but to get to those finer motor skills

you need closer to a week.

Test one, ball grip.

The thing to remember is to not use your other fingers

because that's our kind of go to is where we just grip it.

Right, press your sensor, and then just press and hold.

Yeah, so this is the little practice bit.

We'll start here. [laughs]

Difficulty, 3 out of 10.

So this is our next kind of object task.

Pick up as many balls as you can with your biological hand,

and then use the third thumb

to pick up the kind of last ball or an extra one.

Alright, come back. Maybe four was ambitious.

The goal of this test is to determine

what the third thumb can do

to expand the function of the hand overall.

Difficulty, 6 out of 10.

So I think when the hand is oriented away from you,

you don't have that proprioception, right?

You don't know where the hand is situated in space.

Whereas obviously with your own fingers,

you know exactly where they are

when you're not looking at them.

So that's one challenge, like orienting the hand

in relation to the object you're picking up

without being able to see where the thumbs are.

That's a bit of a challenge. Another test is the peg test.

It's about getting your whole hand in the right position

while using the thumb,

which is something that you normally

never have to think about.

That was good.

It's really about getting your hand in the right position,

which you never normally have to think about.

You just kind of do it.

It still feels like, I don't know,

like you know when you're like holding something with tongs

or something like that, or like, you know,

I feel very aware that this thing

is not a natural part of my body,

so I'm having to be quite deliberate with it.

[peg clattering]

[upbeat music]

I would give this a 6 out of 10 for difficulty.

Moving on, I'm going to attempt the Jenga test.

The goal with this test is to train the user

on collaborating with another finger.

Pick up two blocks, one between two fingers,

and the other one between a finger and the third thumb.

And then place them here. And then kind of build like this.

So just two is fine, and then you want do it with three.

[laughs] Again, it's that really kind of subtle movement

that you need that is kind of quite tough.

Yeah, nice, nice, nice. And then right foot in.

[block clattering]

Almost.

And also with your actual thumb as well, both of them.

There we go. Yeah, nice.

Oh yeah. [laughs]

I found the Jenga test really difficult.

Without any feedback from the third thumb

to tell me how hard I was pressing,

that sense of proprioception that I learned about in Tokyo,

it was really tricky to synchronize my movements

with my real thumb.

I would give this a difficulty of 11 out of 10.

So now that we've seen the thumb in action,

I wanted to know how using the mechanical third thumb

would affect my brain.

[machine whirring]

[Technician] In case of any problems,

squeeze the buzzer, yeah?

You want to try it for us?

[Amit] Try the buzzer?

[Technician] Yeah, just to make sure.

[buzzer beeping]

Okay, once was enough. Thank you.

[Amit] Tamar Makin is a professor

of cognitive neuroscience at Cambridge

who works with Dani to do research on human augmentation.

We use the thumb in order to introduce participants

with completely new ways to operate with their body

in order to do more than they could with just their hands.

[Amit] We took a look at exactly how my brain

was responding to the third thumb.

This is your hand area,

and the highlight tells us that it is activated.

The reason it is activated

is probably because you are moving your finger in this scan.

Is that true, were you moving your fingers?

I was moving either my toes or my fingers.

So in different trials I was moving my pinkies,

and my thumbs in some trials,

and then in some trials I was moving my third thumb

and sometimes I was moving my toe.

Now let's look what happens

when you are using your third thumb.

So this shows us your foot area being activated.

You're probably pushing your toes.

But if we zoom back to the hand area,

we don't see much activity.

And this is because your brain

didn't have to utilize your hands

and pull the muscles of your hands

in order to control the thumb.

So see this knob here, this is your hand area, right?

This area would light up when you're using your hand.

When you're using your third thumb,

you're not really using your hand,

you're really using your feet.

So it is gonna be the activity within here.

So we have this groove that is crossing along the cortex.

We call it the central sulcus.

And this little heel up in front of it is your motor cortex.

Well, my motor cortex, this is my brain.

And this one here is the somatosensory cortex.

So we control movement with this one.

And then as we move,

we're gonna get sensory information from the fingers

as they're moving from the muscles,

maybe if I'm touching something.

And this is gonna come back here

to the somatosensory cortex.

So these two parts of the brain

work in synchrony with each other.

It's this synchrony that allows you fluent motor control.

And right here, just in the middle,

this is like the best piece of real estate

you can get in that region, is your hand area.

And it has great connections to lots of other brain areas.

And it's also very well connected to this part at the top

where your feet are.

And I think that's why the toe control works so well

because we have really good ability

to communicate between these two brain areas.

If you are smart about the design,

if you can, you know, think about intuitive design,

you can make technologies that are plug and play.

[Amit] The hardest thing about using the thumb

was learning to control it,

but in the future could these robotic limbs

be connected directly to your brain via an implanted chip?

At the moment,

which is the very first earliest days of this field,

I think we want to provide a proof of concept

that, yes, we're very comfortable in our bodies

with our five fingers hand,

but any limitations in using extra fingers and arms

is a limitation ultimately of our imagination as humans.

There are multiple ways that we are currently demonstrating

to embed additional body parts into our bodies,

into our brains, into our cognitive awareness.

And there is no reason why we shouldn't redesign the world

in order to take better advantage of these technologies.

[Amit] I'm not sure I need

extra arms or another thumb just yet,

but the idea of a world by becoming superhuman

is as easy as strapping on a backpack,

I give that three thumbs up.

[clapperboard clacking]

[gentle electronic music]