Design This Day | Episode 10

How to build a bionic hand.

Aadeel Akhtar joins Teague Futurist Devin Liddell to talk about building advanced bionic hands, restoring the sense of touch, and designing technology for both humans and robots. Aadeel is the CEO of PSYONIC, a company making advanced bionic limbs that are used by people and humanoid robots.

Aadeel Akhtar:
The way that hand worked at the time too, was that when you turn it on, it'll undergo a training procedure where you have to train the AI algorithm to recognize your muscle patterns. And so we undergo the training process. Juan makes a fist for 15 seconds, a pinch for 15 seconds, hand open for 15 seconds, tripod grip for 15 seconds, relax. And then the hand opens up and it should be ready for him to control. And so I ask him to close his hand and nothing happens, and I'm just like, I don't know what else to do at this point.

Devin Liddell:
Welcome to Design This Day, a podcast about the future, the futures we want, and the people working right now to make those futures real. I'm your host, Devin Liddell. I'm a futurist at Teague. I spend a lot of my time thinking about the gap between present challenges and future solutions.
Today I'm speaking with Aadeel Akhtar, CEO of PSYONIC, a company that makes advanced bionic limbs for both humans and robots. On the show, we often ask what kind of work should we give to machines and what should remain uniquely human? This kind of tension is exactly where design matters most. Speaking with Aadeel, who helped develop a bionic hand that works for both people and robots, felt like the perfect opportunity to dig deeper into this question. But first, let's hear how Aadeel started down this path.

It's 2014, and Aadeel is a grad student at the University of Illinois. At the time, most prosthetic hands still worked in pretty basic ways. Many relied on cables and harnesses attached to the shoulder, or they were electric-powered hands that could only manage a couple of stiff grips. But Aadeel and his team had developed a prosthetic hand that could be closely controlled by the tiny electrical signals that fire when muscles flex for patients, this promised something radical, a prosthetic that would actually feel like part of their body and restore everyday motions that most people take for granted.

So when a non-profit called the Range of Motion project, invited Aadeel's team to Ecuador to demonstrate their device on a patient, it felt like a breakthrough moment. International media would be there, even the US embassy plan to attend. But there was one small tiny problem. The prototype had stopped working. Aadeel and his colleague Mary arrived in Ecuador and dove into troubleshooting mode. But on the eve of the big demonstration, the hand was still barely moving. That's when Dave Krupa, the organizer, pulled Aadeel aside and told him he was out of time.

Aadeel Akhtar:
He was like, okay, so media's coming at 9:00. We're going to pick you up at 6:00 AM and whatever you have working at that point, we're just going to have to roll with it. 6:00 AM rolls around. Mary, the other graduate student and I, we have been up for 32 hours straight. I have been looking through hundreds of lines of code, trying to figure out what might be wrong there. She's been looking at this giant mess of a breadboard, trying to figure out what could be wrong on that side.

6:08 rolls around and we hadn't heard from Dave. And I see this one line of code that looks a little bit off, and I just send Dave a text message and I'm like, "Hey, I think I might've found something. I just need a little bit more time." And he's like, "Okay." And it turns out there was one line of code where two variables are switched on the opposite side of an equal sign. And as soon as they switched those two variables, everything started working. Had Dave come to pick us up on time, we wouldn't have figured it out. So we were lucky that he was late to come get us.

Devin Liddell:
Aadeel brings the prosthetic hand to the demonstration site and meets the patient who will be trying it on. His name is Juan. He had lost his hand in a landmine explosion 35 years ago. The news cameras are ready, the American diplomats are watching.

Aadeel Akhtar:
So the first thing we do is we hook Juan up to all of the muscle sensors and we try to see if we can read his muscle signals. And every single channel was dead and we weren't reading anything. And I was like, how is this possible? This was just working on me. So Mary and I, we go into debug mode trying to figure out what's wrong with this breadboard while Juan and Dave are trying to stall with the media telling the backstory of the range of motion project and how Juan lost his hand to buy us enough time essentially to figure out what's wrong. And so we noticed that this breadboard was powered off of two nine volt batteries, and one of the batteries had puffed up to about double in size and was burning hot to the touch. And had we not noticed at that point, we might've been on the news for a different reason.

Devin Liddell:
You're close to an explosion.

Aadeel Akhtar:
Right. But fortunately, I had a fresh nine volt in my pocket. I was able to take it out and hook it up properly, and all the muscle signals appeared and we were like, okay, crisis averted. But it's one thing for us to be able to read his muscle signals. It's another thing for the mechanics of the hand to respond accordingly. And so the way the hand worked at the time too was that when you turn it on, it'll undergo a training procedure where you have to train the AI algorithm to recognize your muscle pattern. So essentially the hand would make a pinch and then the user will try to emulate that same movement with their residual limb. We undergo the training process and then the hand opens up and it should be ready for him to control. I ask him to close his hand and nothing happens.

I'm just like, I don't know what else to do at this point. And so this was completely serendipitous, but I happen to be Muslim and Juan happens to be the leader of all the Muslims in Ecuador. And to put this in perspective, Ecuador is a 99.9999999% Catholic country. So for me to, A, meet another Muslim, but B, the leader of all the Muslims is completely serendipitous. And so we found this out after we had met each other. And so he turns to me and he's just like Bismillah, say Bismillah. And for those who don't know, Bismillah is just invoking the name of God. It's something that Muslims say before they start anything. And I am just like, okay, Bismillah, Bismillah. And I hit the reset button on the microcontroller and then the hand opens up and this time I ask him to close it and the hand closes and the media comes rushing around the table asking him how it feels to be controlling this hand.

I take a huge step back, a huge sigh of relief. And so as Juan is going through this, right, he's telling these international news stations that he felt as though a part of him had come back, and that was because he made a pinch with his left hand for the first time in 35 years. So when he said that, that's when I realized if I stay in academia, if I finish medical school, this just ends up as a journal paper. If we want everyone to feel the same way that wanted, we had to commercialize the technology. And that's when cynic was born.

He felt as though a part of him had come back, and that was because he made a pinch with his left hand for the first time in 35 years.

Devin Liddell:
Wow, that's astonishing.

Aadeel Akhtar:
The way my wife puts it in retrospect is that I went down on that trip as an MD PhD student and I came back as a CEO because I was never intending to start a company. That was actually never the goal. The original plan was to work at an academic hospital where I would see patients once a week. But what Juan had made me realize if we wanted to make the impact that we really did, and if we wanted everyone to feel the same way that he had done, we had to commercialize this technology and we had to make a startup out of it. And it was a complete change for me in the sense that I didn't have any business training, I had engineering training, I had medical training, but business was something that I'd never seen before and I'd never considered it as well.

Devin Liddell:
What Aadeel realized in that moment is something many innovators eventually face. Taking an idea out of the lab and into the world is an entirely different kind of challenge. Commercializing a breakthrough isn't just about the brilliance of the invention itself. It's about navigating everything around it. Startups often stall not because the technology isn't solid, but because managing a budget calls for tough decisions, pitching investors is harder than it looks, and recruiting the right team is its own art form. And that leap getting from prototype to something real is often where the hardest design choices are made. But first I wanted to get into how the cynic hand actually works. I mentioned earlier that the hand essentially uses tiny electric signals from your body to guide its movement. Aadeel explains how this works.

Aadeel Akhtar:
For those who aren't familiar with electromyography, what you're probably familiar with instead is EKG or electrocardiography. And so those are the sensors that go over your heart, you see in the ambulance and they go, beep, beep. And then if you die, you flat-line, beep, right? So if you can imagine that instead of placing those sensors over your heart, if you place them over your muscles when you have no activity, when you're just relaxing, you'll see kind of a flat-line of activity. But then when you flex your muscles, you'll see this huge burst of electrical activity happening that your muscles are amplifying from the electrical signals coming from your nerves. And that's what we can measure on your skin, and that's what we use to control these devices.

Devin Liddell:
But this technology called surface EMG has limitations. Aadeel explained that even sweat or a change in your hydration levels can change the sensor's ability to pick up these subtle signals plus frequencies from everyday devices or even just a plain light bulb can interfere with the signals as well. It's why Aadeel and others are already developing the next generation of this technology to go directly to the source.

Aadeel Akhtar:
And so what we're seeing more now, and we're working with a lot of these companies, we're working with the Navy Hospital here in San Diego and UCSD here as well, is the next generation of these bionic devices that are going to be directly attached and connected to your nerves and your muscles inside the body. We're seeing implanted electrodes that go directly in your muscles as well as micro wires that can go directly into the nerves, and then they route through cables that go through your bones, and then you attach the hand to a titanium implant that goes in your bones and you directly communicate with your nerves. Or we can wirelessly transmit these things as well too. And so we're going to see more and more of those technologies over the next five years or so. And that's kind of where we're heading with this stuff, where there's seamless integration between humans and robots.

We're going to see more and more of those technologies over the next five years or so. And that's kind of where we're heading with this stuff, where there's seamless integration between humans and robots.

Devin Liddell:
This might be a strange question, but there are ways in which we use our hands that are purely functional. I need to twist a jar open. There's a screw that have dropped. I need to reach down and pick it back up. There are other things we use our hands for that are sort of more in emotional territory. I mean like how we would pet a dog for example, or how you might touch someone's shoulder to comfort them. Is there anything in your work that has helped you maneuver how people use the PSYONIC hand to maneuver that gap or to do both of those things? Is that anything that ever comes up?

Aadeel Akhtar:
One of the things that I didn't mention yet was that with those implanted electrodes, this is the first bionic can to give users touch feedback. So right now they feel a vibration. When you touch our fingers through vibration motor, it's kind of like your cell phone vibrating. But with the implanted electrodes, we can stimulate the nerves directly and make that sensation feel like it's actually coming from your missing hand.

And to be able to give them that sense back, I think going to be really special and something that we're going to be seeing more and more of over the next 10 to 15 years to be able to have some of the feeling of touch has a very emotional impact. And even with the vibration motor, our very first user in the US sergeant Gary to Anderson, he told us that he could actually feel his daughter's hand when he's holding it. And that was something that he couldn't do with any other prosthetic, even though it still felt like a vibration to be able to know that he's actually holding his daughter's hand. Just that emotional impact in and of itself goes a long way.

Our very first user in the US, Sergeant Garrett Anderson, he told us that he could actually feel his daughter's hand when he's holding it... Just that emotional impact in and of itself goes a long way.

Devin Liddell:
Moments like that really capture the heart of Aadeel's work. This bionic hand was reconnecting users to experiences that feel profoundly human. But in 2022, Aadeel attended a robotics conference that unlocked a whole new direction. He met a robotics engineer who took one look at the hand and recognized a different potential. He told Aadeel that roboticists would jump at the chance to put that prosthetic hand on a robot. Today, Aadeel's team is partnering with 50 plus tech companies, including Google and Meta to build AI robots that will have anthropomorphic hands, and those hands will be almost as capable as our own in terms of human users of the hand and robotic users of the hand. Is there any distinction with the unit itself, or is the unit completely identical? It is the same unit that's used by both.

Aadeel Akhtar:
Yeah. The only difference is the wrist connector because you have to adapt it to whatever robot you're applying it to or the human users. So for example, the one that I'm holding right now, that the hand is on what we call an electronic quick disconnect, which is on the human side, and it allows the user to quickly pop the hand on and off. But then on the robotic side, we might have a four bolt circle wrist, but everything else, all the hardware inside the hand is exactly the same. All the motors, all the sensors are exactly the same between the humans and the robots.

Devin Liddell:
Amazing. And aesthetically it's the same, it looks the same.

Aadeel Akhtar:
Yep.

Devin Liddell:
It'd be indistinguishable. Okay.

Aadeel Akhtar:
Exactly. And that was by design too, because the thing is if you're building a robot to do human tasks, we've optimized our hand for humans to do human tasks. So it makes sense that the robots end up using the same thing.

Devin Liddell:
Earlier this season, we spoke with Matt Ondler at Axiom Space where they're building the world's next space station. Matt said, astronauts and robots will likely work side-by-side to keep the station running. For robots to handle that kind of environment, some will need human-like features, hands that can grip, twist, and manipulate tools just like we do. So it's no surprise NASA turned to PSYONIC. They wanted to test one of Aadeel's bionic hands on their humanoid robot named Valkyrie.

Aadeel Akhtar:
We had visited NASA a year and a half ago or two years ago when they first put our hand onto Valkyrie, their humanoid astronaut robot. And we brought down our user experience specialist, Anika Berlin with us. They had Valkyrie doing a bunch of different tasks that you find on the International Space Station, so pushing buttons, handling knobs and switches that you might find on the ISS. And they were struggling with a zipper that was on the wall. And the way that they were going about the zipper was that they were trying to get the hand to basically make a pinch on the zipper and then zip it that way. But remember when I said that about 80% of the tasks that we do are either a power grip or a key grip. So typically when we're manipulating a zipper, you're putting your thumb on the side of your finger and then gripping the zipper in between there.

You're not actually doing a pinch, you're actually doing a lateral grasp or a key grip. And so Anika, she goes up to the zipper and she does a key grip on it, and she easily does it. And the NASA engineers were like, we didn't even think about using the hand in that way. And so we have these preconceived notions on how we even use our own hands that we don't realize that that's not actually how we're doing it in reality. And I think those are some of the most surprising things that we found is these little things that are these seemingly simple tasks that we almost take for granted, but we haven't actually done a deeper analysis on how actually are we doing those things.

Devin Liddell:
Interesting. And will PSYONIC always be a maker of hands for both humans and robots? Or do you think there's a scenario where you just shift into providing hands for humanoid robots?

Aadeel Akhtar:
So my goal has always been to be on the human side. And in my head, we can't decouple the human from the robot side because again, if you're building a robot to do the human tasks, you need the humans to have done those human tasks as well too. And there's a lot of data that we can learn from our human users as well too. I mean, there have been many robotics companies that have watched our human tutorial videos on how to crack an egg or how to tie shoelaces and then use that to teach their robots how to do the same thing with the exact same hand. And because it's the exact same hand, it's almost like a one-to-one translation then between what our human users are doing and what the robots are doing. And so what we build for the robots benefits the humans. And what we build for the humans benefits the robots is a completely synergistic relationship.

What we build for the robots benefits the humans. And what we build for the humans benefits the robots. It's a completely synergistic relationship.

Devin Liddell:
I love that. Yeah, that's a great way to describe that. Are there examples of how robots and humanoid robots are treated in pop culture? So are there other scenarios where you watch, I imagine there are scenarios where you're watching a show and you're like, that's not the way it is.

Aadeel Akhtar:
Yeah, in our offices we've got the Unitree dog, the Aliengo, as well as the G1, their humanoid robot as well too. And a lot of times we will get comments on like, oh, this is from Black Mirror. And I'm just like, oh man, this is not Black Mirror. This is not what we're doing here. For us, especially with the dog, because we've got the arm on it and we've adapted the arm to have our hand on it at the end of it instead. And we've got a ton of YouTube videos of it, just showing our robot dog doing different things like laboratory work, like putting glass slides on test tubes, things like drilling into a piece of wood. And the idea is that it's basically a robot Armani mobile platform. It's very similar to what Boston Dynamics is doing with Spot. You want it to be able to be deployed so it can open doors for people, do inspections, all this kind of stuff. So yeah, we have to contend with that a little bit too.

But at the same time, I think there's also just this huge fascination with these robots as well too, just from a curiosity perspective. And when we have people come and visit the lab and they see the G1 with our hands on it and they get to interact with its just like, wow, they didn't know that this technology was real. They thought it was science fiction, but it actually exists in real life. And last week we saw the world Humanoid Robotics competition happening in Beijing, and then next week is the first Humanoid Olympics happening in Greece as well. So I'll be attending that. And I think it's going to be really exciting for the public to see where this technology is right now and then where it's going to be in the next 10 years.

Devin Liddell:
Are there any dystopian outcomes that you either are actively working to help prevent or just you actually just worry about? And maybe even another way to put it is, are there applications of the PSYONIC hand that you would turn away?

Aadeel Akhtar:
Yeah. And in particular, we're against weaponization. So using our hands with weapons in particular. Because again, the reason why we went into this field was to build people up, not break them down. And so we come from a very humanistic aspect coming from the prosthetic side in particular, and that part of us will never go away.

Devin Liddell:
I love that. All right, so we're going to go into a quick lightning round. What's your take on the most over-hyped technology at the moment?

Yeah, so I would probably say that is probably artificial general intelligence. I think there's this conception that robots are going to become sentient and that is just not how the AI works in particular. So I would say that AGI is probably the most over-hyped thing right now.

Devin Liddell:
I love it. Just because of your unique background of having both sort of an academia background, but also you're now presently a CEO and having a business background. What do you think is behind sometimes researchers, and this has been documented in a couple scenarios where researchers working on AGI or working toward AGI I should say, have sort of maybe jumped the gun a little bit and said like, no, I'm actually seeing sentience here. What do you think is behind that?

The reason why we went into this field was to build people up, not break them down. And so we come from a very humanistic aspect coming from the prosthetic side in particular, and that part of us will never go away.

Aadeel Akhtar:
I think part of it is just things like sci-fi, right? But we have these preconceived notions of what some of these things are theoretically capable of. Going all the way back to Star Wars, I mean, you had these sentient Androids like C3PO and R2D2 and everything. And I think we almost want to have a natural tendency to validate some of those ideas and things that we see in the real world without thinking through things objectively sometimes, and results in misconceived perceptions of a robot when there is none in the first place.

Devin Liddell:
Exactly. I love that you brought up the Star Wars droids too, by the way, because they're actually, the irony for me always with especially C3PO, is that actually he's intensely human. He's nervous, he's chatty, he talks too much. He would not be programmed in that way. He's a good example of humans in a lot of ways.

Aadeel Akhtar:
For sure.

Devin Liddell:
All right, so the other side of that coin, what's your take on the most under-hyped technology?

Aadeel Akhtar:
Yeah, I think it'd be nice to have more hype on battery technology in particular. And I think especially as we get to more of these ubiquitous robots, these more mobile and wearable devices, edge computing and that level of things, we need more done on the battery side. These devices need to last longer. They need to be able to handle longer lasting things. I mean, I feel like these are a lot of the failures of wearable devices right now too, is that they might only last for a couple hours as opposed to longer times. So I mean things like watches and whatnot, and if we want to move more towards these ubiquitous computing, this wearable computing, and for us like bionic limbs, then I wish there were more hype around battery technology.

Devin Liddell:
Interesting. Okay. All right. Last one. Is there a book or a show that you're currently watching or reading that is influencing how you're thinking about the work you're doing?

Aadeel Akhtar:
Yeah, so that's a good one. Shoe Dog by Phil Knight, and I read that a couple years ago, but it is one of my absolute favorite books of all time. And just so many parallels between just that entrepreneurial journey, even though it's in a completely different market, the shoe market in particular, and just Nike and the story there, but was, I felt that in my bones as I was reading it.

Devin Liddell:
I can see that. It's funny that you bring that up by the way because I will say, and we've done quite a bit of work with Nike at Teague, when you were describing your position on what things you wouldn't do with the hand, and you used a phrase, we want to build people up.

Aadeel Akhtar:
Yeah, instead of breaking them down.

Devin Liddell:
Instead of breaking them down. And when you said, we want to build people up, I actually heard a little bit of an echo with the way Nike talks about athletes. They will say things like if you have a body, you are an athlete. But it's also to me, reminiscent of sort of a very principled approach to where you're going and why you're innovating. And I think in that regard, there's some similarities between what you're doing and what Nike does. So I can see why that book resonate. I think there's probably some natural affinities there between the two of you.

That's it for today. Thank you for listening to Design This Day, a podcast by Teague. Subscribe on your favorite podcast app so you don't miss the next episode. And if you have a complex problem that needs solving, we'd love to hear from you. Visit us at teague.com Or send us an email at hello@teague.com.