Exoskeleton boot that learns how you walk

When Nancy Sinatra sang 'these boots are made for walking' in 1965, she probably didn't foresee the unveiling by US scientists this week of robotic footwear that can make you move 9% faster and as though you were 13kg lighter. But that's what Stanford's Patrick Slade has come up with. His new "exoboots", he says, can give assistance to those who move less well with age or disease, and they can even make walking harder if needed to act as a training and strength-building device. The key is that these boots "learn" how you walk so they can intervene in the most effective way - like the foot equivalent of power-steering - to make your movements as efficient as possible...

Patrick - When you hear exoskeleton, you think Iron man. But realistically what it is, is it looks like a shoe with a piece that runs up the calf. And this piece has a motor that sits right behind your calf. What the device does is it puts a spring in your step where it helps provide a push-off assistance as you're walking by having the motor wind a cable which is attached to the heel of the shoe and inject some energy into your walk.

Chris - And how does it know when to do that?

Patrick - We use sensors on the device to figure out how you're moving. So for example, we have a pressure sensor in the shoe to detect when your foot's on the ground and if you've just hit the ground or if you're about to toe off, we have a sensor to understand your ankle motion. All of that information is collected by a little computer, which is worn on the exoskeleton. As it gets sensor data, it makes decisions on how to assist the person.

Chris - Can it make better decisions with time? Because everyone's gait is unique. I mean there was even a paper a few years back saying we can use CCTV footage of thieves and work out from their walk who they are <laugh>. Given the uniqueness of gait, can it adapt itself to the walker and the wearer?

Patrick - Yes, absolutely. So you are spot on. Everyone walks very uniquely, and this is one big challenge of providing effective assistance is how do you personalize it to each person? Because it turns out if you just have one assistance pattern and you slap it on everyone, it doesn't provide great benefits. A big part of our project was determining how you can look at their motion, so specifically your ankle motion and the force that you're applying. And from that learn how the exoskeleton is improving the efficiency or making it worse. So our device, over an hour of walking outside, tries out a bunch of different assistance patterns, figures out for you what is providing the best benefit and then figures out exactly what's gonna be the most beneficial form of assistance for you.

Chris - How does it know when it's getting it right?

Patrick - That's a great question. So in the lab we collected a ton of data where we had people walk with different types of exoskeleton assistance while measuring this energy expenditure. And then we train the model which can mount ankle motion and assistance data to end the expenditure. So once we've built this model, you don't have to come into the lab, you just put on the exo and we look at your motion as you move around and we can infer how beneficial certain assistance conditions are and from that we can personalize the assistance.

Chris - And how good is it? I mean, it's one thing to invent something like this and, as you say, pictures of Iron man and sort of go through your mind <laugh>, but if you take your average elderly person who doesn't get out much because they've lost confidence and they've lost strength, are we gonna see something game changing for them where they're gonna be able to make it to the park, make it to the shops?

Patrick - Yeah, I hope so. And I think so. So, in this study we performed experiments with young healthy adults and what we saw was when they wore our device that had personalized assistance to them it actually reduced their energy. And this was equivalent to removing a 13 kilogram backpack, which is a significant reduction in burden. And also it increased their walking speed by about 9%, which is clinically significant. And so now we're starting to think about and starting to perform studies with older adults to make sure that we can provide these same level of benefits. Early indications show that it's absolutely possible that we could provide increase in their daily mobility which will have an impact on their independence and what they like to do. And I like to think of it like an e-bike for example. My grandpa loves to bike. As he got older, he was able to do it less and less. We got him an e-bike and all of a sudden hills aren't a big problem and he can go for as long as he wants. And so hopefully this device will allow people to walk more easily and then they'll walk more and they'll do the things that they want to do.

Chris - Aside from just being supportive, could it also be interventional, as in therapeutics? So if you've got someone who's had a stroke, for example, they've got movement but it's weak or it's not optimal, could something like this be programmed to try to help them to restore better movement? So it's part of their rehab program?

Patrick - Absolutely. One big area of research is assisting different groups with muscle weaknesses. Stroke is a good example. These muscle weakness conditions can definitely benefit from augmenting movement. It can even be extended to things like injury recovery. So for example, if you injure your achilles tendon, you could put one of these boots on and reduce the load and then perhaps help with the recovery process and allow you to move more freely then. We can use the device to make walking harder to provide some sort of training. But there's a whole host of questions now that as we're getting close to being able to deploy these devices, that we want to make sure that we can really understand how it's gonna impact people's health.

Chris - The computer you can power with a couple of simple AA batteries, but how are you powering the motors? Because if you're putting that much effort, lifting big backpacks off of a person's shoulders effectively <laugh>, that's a lot of energy. Where is the person having to trail around on an extension lead, like a primitive remote controlled car, or can they actually take their power source with them?

Patrick - <laugh>? Fortunately, yes, it's all portable. So we use a portable battery right now. We have them wear the batteries in a little waist pack. 0.6 kilograms of battery gets you about 40 minutes of walk time and so you can swap them out if you run low. People typically walk about an hour and a half per day. So carrying two sets of batteries could potentially get you there.