Compressed Air Jumping Shoes Are Not For The Faint-Hearted

[Ian Charnas] has taken a short break from building things that might injure himself, by building something that could injure somebody else instead. (Video, embedded below) Well, hopefully not anyway. After working with YouTuber [Tyler Csatari] on a few ideas, [Tyler] was insistent on getting some power-assisted jumping shoes, so [Ian] set to work mounting some compressed-air powered pistons to a pair of walking shoes. With a large backpack housing the 200 PSI air cylinder, control valves and timers. The whole affair looks solidly constructed, if a little ungainly, but does seem to work surprisingly well.

After some initial calculations of how much force each piston could exert before risking leg injury, he found that whilst it did work, to an extent, the pressure required was beyond the capability of the compressor they had on hand. After a shopping trip, a bigger compressor was located, but that still needed a modification to get anywhere near its maximum 200 psi rating. The thing is, that modification was to bypass the regulator and the safety valve, and this is definitely something you don’t want to be making a habit of. Compressed air systems like this can hold quite a bit of an explosion potential if pushed beyond reasonable limits, and care needs to be taken to keep things within safe bounds.

Cost-wise, [Ian] does mention a figure of around $3,000 USD making it a bit of a pricey project, but hey a YouTuber’s paying the bill, so it must just be a drop in the ocean for them?

Just to illustrate how useful compressed air is as a method of storing energy, here’s a compressed-air powered helicopter, and a 3D printed wankel rotary engine, which must’ve been tough to dial in and get working!

30 thoughts on “Compressed Air Jumping Shoes Are Not For The Faint-Hearted

  1. for an engineer he is an idiot at measuring. you do not measure leg lift you measure the height you can reach up and touch when standing. then the max height you can touch when jumping. these “jumps” are pretty horrible imho and not really doing anything more than pushing the knees into a bent position. but that is not a jump that is just assisted knee bending :P

      1. “Good thoughts, but I wonder why you have to call him idiot?”

        When criticizing people it’s important that you undermine their character to make to make yourself appear more authoritative.

      2. jpa – thank you for the civility.

        Yes from an engineer’s perspective, I completely agree with you that the center of mass makes the most sense as the measuring point. However, from a showmanship perspective, if we could get him high enough to dunk a basketball, or jump completely over a 6 foot obstacle (a foam obstacle, for safety’s sake) that would perhaps be more exciting. Food for thought.

    1. Hi The+Gambler. Ignoring the unnecessary personal insult, this is an interesting topic. In some sports (cross fit box jump, hurdles, etc) the goal is to get the feet as high off the ground as possible, while in others (high jump) the goal is to get the entire body as high as possible, while in others still (dunking in basketball) the goal is to get the hands as high as possible. It reminds me of the nice thing about standards being that there are so many of them :-)

      In any case, as long as my measurements are all done the same way (with knees bent, cross fit style), it is straightforward to determine the increased height (or lack thereof) provided by these jumping shoes. I found that with the boots, he jumped higher.

      The world record for box jump is 5 feet, 3 inches, and with Version 2 of the shoes it will be easy to get higher than that. You may not like the way that cross fit’s box jump or the hurdles utilize bent knees, but it would be incorrect to say that there is no precedent in sports for measuring this way. Would you agree?


        1. Dan, great question. The air tank alone is 6 pounds. Once you add in the other things on his back the total weight there is closer to 20 pounds. The pistons and steel bits add 5 pounds to each of the shoes. So in total he’s about 30 pounds heavier for having worn all this, but he was able to jump higher with it rather than without.

    1. jtl – indeed and thank you for your comment. I went with the highest Cv valves I could find at reasonable price, with a large diameter tubing (1/2 inch) and large diameter fittings. I believe I have done this properly as the pistons moved as quickly as the maximum that the appropriate curve on the datasheet indicated.

      The pressure in the reservoir was increased as we tested and became more confident in the safety of the system, and of course the purpose of higher pressures was to increase the force in the pistons and thus the acceleration of the mass (namely, Tyler).

      One thing I’d be curious of, from those of us here who can speak intelligently (as jtl does) and without personal insults, is whether there’s a way to measure what the limiting factor is in the piston speed. When the piston is unloaded, I expect the limiting factor is just the speed at which we can move air into the cylinders (so, limited by the resistance to air flow of the valves, tubing, and fittings) however when the piston is heavily loaded I expect the piston speeds are more limited by F=mA. I wonder if there’s a way to tell which is the dominant factor in the system I built.

      1. A high speed camera would allow you to check this. Measure the no load speed of the pistons at max working psi, then load the pistons with body weight at same pressure and check again. If slower, then you’ve hit a bottleneck. At that point, reduced tubing friction, faster solenoid valves could be considered. Shorter pistons may aid in faster acceleration as well…. I would love to see a side by side of short pistons vs longer to see which provides greater boost. Cheers!

    1. And likely the concussive force breaking both ankles and the fibulas, and possibly the tibias – maybe shattering the patella in both legs also. Guessing breaking bones in your face from the legs flying up and smashing there too.

      Not something I would sign up for… Fun for all but the wearer…

  2. Reminds me of Claude Shannon who came up with the theory that is still being used by advanced satellite modem manufacturers to characterize how good modems, modcods, FECs are (how close they are to the Shannon curve), who used his knowledge of math to knock over Las Vegas and most importantly, invented the gasoline powered internal combustion pogo stick.

  3. For mounting the boots you could have just looked at how roller skates are done with single bolts through between the center of each foot in front and one bolt directly centered on the pad at the heel of your foot. I’m curious how you measured the central point of how far to move the boot into the cross braces, I would have thought that mounting farther back toward the sole would have been better for weight distribution and force, reducing the strain on the foot. These look WAY too far forward away from the heel towards the front, which has the effect of bending the angle back towards the shin and stressing the bend points along your foot when triggered, especially the top of the ankle itself.

    1. Bobby, thank you for the comment and that’s a very interesting suggestion. Mounting things to humans is indeed difficult – we just weren’t made for easy means of accessory attachment ;-)

      The roller skates are an interesting idea! I love roller skating myself, and yes modifying those would have been much simpler than the approach I used. I like it.

      Regarding where to put the bracket, initially yes it seemed like the wise place would have been the heel, where the forces could transmit more directly to the leg. However after speaking with physical therapists I realized that this would limit the jump height, which I’ll explain – since humans jump from the balls of our feet, I can benefit from mounting the piston bracket there. The idea is that Tyler uses his leg muscles to jump as high as he can, and then just before his feet leave the ground he presses the trigger to fire the pistons. At this point the pistons push against the ground, pushing him up even further. So in this scenario the forces he can provide himself are superimposed upon the forces provided by the pistons, launching him up to a nice height. If I had instead mounted the pistons towards the heel, he wouldn’t be able to jump by himself, and then the only upward force would be provided by the pistons. Better to have his leg forces plus the piston forces. Very interesting topic and thank you for bringing this up.


      1. I spent most of my younger days speed skating – so the skates I have use nylon plates with only single attachment points inline, and they hold up fine considering the forces involved in that.

        True, the ligament and flex involved are more in the equation the farther out you move on the fulcrum. And it would allow a plus+plus to the energy where moving it towards the heel loses the first side of the jump. I get it.

        I just see in my mind the pistons firing and the micro-fractures moving through the heel up through the legs. Meh, maybe it’s just me. When I was younger and invincible I would have done the same. But given I hear those creaks just standing up now I guess those days are over. I was actually surprised that the width of each foot platform didn’t interfere with a standing jump – I bet it was pretty close.

        1. Bobby, very fair points!

          I’m similarly not a “shoot from the hip” type guy and I was super concerned for Tyler’s safety. I consulted with physical therapists, read research papers on foot forces in athletes, and tried to think about all the things that could go wrong and plan safety features for them. I think the biggest risk mitigation was in starting at like 5 PSI in the tank and then going up slowly by 5 or 10 PSI each time as we built confidence. If his feet started to hurt we were going to stop. I like goofy inventions but I’d be wracked with guilt if my invention seriously injured someone. So I’m with you on that.

  4. Might want to measure the actual pressure in the cylinder compared to piston stroke.
    Our battle bot has some pretty fast cylinder activation. We found common valves/plumbing too restrictive. Pressure slowly builds up, then plateaus. We placed a buffer and custom air-assist valves right on the cylinder. This was a big help, but eventually we put a small cylinder piston inside a larger piston cylinder with custom brass heads. The outer cylinder was our buffer and the inner coaxial cylinder held the actual piston. A free-floating ring acted as the large valve which dumped air from the outer cylinder into the inner cylinder.

    1. Great comment! With my setup I’m not sure if the piston speed is limited by flow rate, or by F=ma. If I had an electronic pressure sensor on the cylinder (like you suggested) that would let me see what kind of pressure drop I’m seeing between the tank and the cylinder. If it’s negligible then I’m limited by F=ma, whereas if the pressure drop is more like 50% of tank pressure then I’m definitely limited by flow rate and should look into a custom setup like you were suggesting. This is the best I could do with off the shelf parts, but custom parts are more fun :-)

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.