The workshop looks ordinary: concrete walls, tool chests, black mats on the floor. Then the young man turns sideways and you see the metal frame strapped from his hip to his shoe.
He squeezes a hand control. The powered knee snaps forward with a hiss and a hard metallic clack. A moment later, he crouches and jumps while wearing the device. It is clever, unsettling and almost impossible to stop watching.
What happens in the clip
The first test is a powered kick. The mechanism drives the lower leg forward, then returns it. During another attempt, the strap around the foot slips and has to be tightened. That small moment matters more than the dramatic kick: if the frame shifts while it is moving, the machine and the knee no longer share the same path.
Next comes a jump test. The device may help produce force on the way up, but it cannot make the wearer lighter on the way down. The landing still has to be absorbed by the feet, ankles, knees, hips and the surface underneath them.
The idea is real—the safety work is the hard part
Exoskeletons are not science fiction. Medical versions can assist rehabilitation, while industrial designs may reduce physical load during specific jobs. The difference between a tested device and a garage prototype is not imagination; it is the layers of engineering around the idea.
Commercial and medical systems are evaluated for mechanical strength, electrical safety, software behaviour, repeated loading and foreseeable misuse. A homemade powered joint can skip those safeguards without the builder realising how many failure modes are hidden in a simple movement.
- Joint alignment: a human knee does not move like a perfect door hinge, so a rigid pivot can rub, pull or twist as the leg bends.
- Mechanical limits: the device needs a physical stop before the body becomes the stop.
- Straps and fasteners: a loose attachment can turn controlled assistance into a moving weight beside the limb.
- Emergency shutdown: the wearer should be able to remove power immediately without searching for a switch.
A safer way to test a build like this
The exciting part is pressing the trigger. The useful part happens earlier: bench-testing the frame, cycling it repeatedly without a person inside it, starting at low power and checking every joint and strap between runs.
A sensible sequence is boring on purpose: test the mechanism on a stand, then try slow unloaded movement, then walking, then a controlled squat. Jumping belongs near the end of the process, not the beginning. Eye protection, a spotter and a forgiving surface are basic precautions, not overreaction.
The part worth remembering
This is a genuinely inventive project, and the builder’s excitement is contagious. But a machine strong enough to make the movement impressive is also strong enough to make a small mistake matter.
Build the machine. Then give the safety system the same attention as the part that moves.
Further reading: NIOSH on wearable technologies and exoskeleton research and NHS guidance on urgent knee symptoms.
Watch the full exoskeleton test
Look closely at the foot strap and the landing—those are the moments that tell the real story.