Mize walking with a bionic exoskeleton!!!

Is that actual Mize from here?

Fucking Brilliant if it is! Fucking Brilliant if it isn't too.

I'm really happy for the people who will be affected by this.
Nice.
 
So the exoskeleton and the therapist are doing the vast majority of the work as far as legs go. The big advantage for me is that, when I do assisted walking with a walker I put an enormous percentage of my weight through my hands because I can't feel my legs, but, in the Rewalk, you cannot lean over enough to do that or it will just stop so I'm putting easily 90% of my weight through my legs which can help fight osteoporosis.

Unfortunately they cost $100k and getting insurance to help is a long shot. :(
 
So the exoskeleton and the therapist are doing the vast majority of the work as far as legs go. The big advantage for me is that, when I do assisted walking with a walker I put an enormous percentage of my weight through my hands because I can't feel my legs, but, in the Rewalk, you cannot lean over enough to do that or it will just stop so I'm putting easily 90% of my weight through my legs which can help fight osteoporosis.

Unfortunately they cost $100k and getting insurance to help is a long shot. :(


What does the ReWalk use to control the actuators? Is it manually controlled through the crutches?

Is the noise we hear in the video from the motors that loud or are the smartphone's microphones exaggerating it a bit?

You seem to be struggling with balance despite the apparently very safe support from 2 crutches and 2 legs. Do you feel there's a lack of "safe balance" in your abdominal region?

Sorry if the questions seem weird, but this is my actual line of work in research :)
 
What does the ReWalk use to control the actuators? Is it manually controlled through the crutches?

Is the noise we hear in the video from the motors that loud or are the smartphone's microphones exaggerating it a bit?

You seem to be struggling with balance despite the apparently very safe support from 2 crutches and 2 legs. Do you feel there's a lack of "safe balance" in your abdominal region?

Sorry if the questions seem weird, but this is my actual line of work in research :)

I'm a T3 complete paraplegic which means I have no volitional control or sensation below my nipple line. So, yeah, balance is a big issue. No abs, no lower back, no intercostals, etc.

It took some time just to learn to sit without using my hands/arms to stabilize. Today many people think my injury is lower because I can balanced pretty well just by shifting my shoulders and head when I'm not tired. But in reality I'm working my ass off to not fall over!

The exo is activated for each step my successfully shifting one's weight far enough to the stance side (and that side hip slightly forward) that the swing foot (rear most) is unweighted. You must also do this shift within a time window while not leaning too far forward. The initial activation is through a watch-like controller that another PT is holding.

The VA covers these for vets but only is they do 10 weeks of training because there's a lot to learn and practice is key. The move from sitting to standing, for example, is pretty challenging. I managed to nail it, but only after a dozen tries. It doesn't help that I'm not great at crutches since I've never really used them as I went from totally able bodied to a state where I'm too disabled to use them :)

Finally, the motors are loud. The phone microphone might be exaggerating it a little, but they're loud.
 
I'm a T3 complete paraplegic which means I have no volitional control or sensation below my nipple line. So, yeah, balance is a big issue. No abs, no lower back, no intercostals, etc.
I have to say you had me fooled pretty well! It looked to me like you had at least some control over your abs.
I think you'll do great with an exoskeleton in the future.


The exo is activated for each step my successfully shifting one's weight far enough to the stance side (and that side hip slightly forward) that the swing foot (rear most) is unweighted. You must also do this shift within a time window while not leaning too far forward. The initial activation is through a watch-like controller that another PT is holding.
So it's kickstarted through manual activation and then it uses angular distance combinations in the knee and hip joints of the exoskeleton?
Did you think it was easy enough or is there a learning curve on the control?


Finally, the motors are loud. The phone microphone might be exaggerating it a little, but they're loud.
That's what I thought :/
They're probably using DC motors that spin really fast (in the order of ~3000 RPM) and then they have harmonic reducers to trade rotation speed for torque.
I think until we get significant leaps on electroactive polymers (think Deus Ex Human Revolution), it's either that combination of DCMotor+gears or pneumatic muscles... that are terrible to control and need compressed air canisters to work, or air pumps, all of which also making noise.



Unfortunately they cost $100k and getting insurance to help is a long shot. :(
Have you seen Ekso and REX?
It seems to me that REX offers a higher level of assistance (it's also a lot bulkier), but Ekso is supposed to be substantially cheaper, around $50k.
https://exoskeletonreport.com/
 
Pneumatics is probably too bouncy for use in human applications I would think, unless fantastic leaps and bounds have somehow been made that I'm completely unaware of. :)

Linear electric motors would be more preferable in that sense, as you can regulate their power output by human measures virtually instantaneously and very precisely, and they're at least fairly quiet; you'd probably have some PWM hum/whirr from the motors and power electronics, but way less than geared DC or an air compressor. :D They might not be ideal in this application though, as they'd draw power constantly to keep a person upright and would probably be a bit of a battery hog...
 
Pneumatics is probably too bouncy for use in human applications I would think, unless fantastic leaps and bounds have somehow been made that I'm completely unaware of.
But maybe they can be bouncy?
Skeletal Muscle behavior is bouncy, and no one really needs the 0.5º precision of the servo motors being used in these solutions. The point is to allow a gait cycle to be formed, not to get paraplegics to do ballet dancing.

I'd say we're going through that phase in human robotics right now, as we're trying to figure out exactly what we need and don't need, the complexity of the variables from biodata being used, the control schemes (doesn't make sense to use PIDs with the current parallel computing power of ULV SoCs, we should be using neural networks everywhere), etc.
The current generation of exoskeletons is well over-engineered in some aspects, of that I'm sure. And that's also what's making the current solutions so damn expensive to make.
And accordingly, with a high BoM they need to sell these things for $50-100k, because at that price they can't have enough unit volume to scale production. They're practically selling hand-made prototypes, so it's a lose-lose situation.

I really think - and hope - these devices will start advancing really fast within the next 5 years or so.
 
But maybe they can be bouncy?
Skeletal Muscle behavior is bouncy, and no one really needs the 0.5º precision of the servo motors being used in these solutions. The point is to allow a gait cycle to be formed, not to get paraplegics to do ballet dancing.
.
Is that part of the reason behind the success of prosthetic blades?
Context beyond just sports.
 
But maybe they can be bouncy?
Skeletal Muscle behavior is bouncy
Yes, I suppose you're right, but we have the ability to essentially pre-tense our muscles depending on the situation so we don't literally bounce when we move... :) Unless the control mechanism of a pneumatic exoskeleton gains knowledge of our intended movements (like, brain interface-style), it would be difficult for it to perform in a similar way I would think. Maybe if we could give it a form of trainable 'muscle memory', allowing it to recognize patterns in our movements and adjust accordingly. AI for exoskeletons essentially, I suppose you could call it. :p

I really think - and hope - these devices will start advancing really fast within the next 5 years or so.
Just today in fact I read an article in a newspaper about a guy who lost both legs and an arm in a collision with a train who is starting up his own prosthetics company because the current market-dominating manufacturers weren't offering good products. His goal was to make much cheaper and better offerings; I took it as he wanted to be a 'market disrupting' newcomer in the artificial limbs business. Granted, not exoskeletons - yet. Maybe some day, if he gets his business off the ground...
 
Back
Top