Gait Simulator, Part I

This is a retrospective post so there aren't as many pictures and I won't go into as much detail (if you're really going to try and construct a gait simulator, let me know and I'll gladly fill you in on the details).

For my Senior Design Project, I worked with Stryker to create an automated gait simulator that would be used in product resiliency tests. The world of medical devices is heavily regulated by the FDA which necessitates fatigue tests which companies like Stryker need to conduct to comply with safety standards. The Stryker advisor we had worked in the lower extremity trauma department which produces intramedullary nails, plates, screws, and external fixators. All these devices need to be tested using machines that can biomechanically relevant motions to simulate fatigue and wear over time. Current machines cost anywhere from $2,000 to $15,000 depending on the model and are large and very heavy. Our challenge was to produce a low-cost, portable alternative to the machines currently being used. 

TL;DR...we made a really cheap machine that simulated walking to make sure implants that would go into your lower leg don't break prematurely.

For a detailed report (or if you particularly enjoy reading caffeine-fueled, last minute papers), contact me at the info in my About Me page. Otherwise, enjoy the pictures, videos, and brief captions below. For Part II, click here.

Above is the half finished simulator. The frame consisted of solid aluminum plates and steel brackets. The pistons could exert a combined force of 592 pounds, which is more than sufficient to test worst case scenarios in obese patients.

Simple 4 way electronic valve with built in flow regulators. One input would drive two outputs determined by the position of the internal electronic valve. The left side is the input, right side is the output which led to...

...two in-line pressure gauge/regulators to observe and regulate the pressure of each piston independently.

Picture sucks since it's a still from a video. Regular air compressor to drive the whole contraption

Completed simulator with platform.

Not pictured is the 555 timer circuit we employed to control the electronic valve. The square output signal was controlled by a potentiometer. A relay was used to control the electronic valve since it drew more current than the 555 could handle.

Below is a video of the initial verification that the design.

And a video with the completed project in an Instron to verify that it could output clinically relevant data.


The results turned out better than expected for a completely passive system. The superimposed graph below is data from a University of Washington study on cadaveric gait. The data under it is the force output curve of our simulator. The resemblance was uncanny, even with the slight variation in toe-off force. Sure there was a discrepancy in the forces of the two curves, but as long as the simulator could output kinematically correct motion, force would be a relatively simple adjustment.

Overall, it was an incredibly fun project since I got to learn how to use machine shop equipment during construction, how to set up an Instron, and other assorted skills. To read up on Part II, click here.