The art of being Soft and Strong

If you imagine something to be soft, flaccid and bendy, you will not expect it to be able to give a strong resistive force. Such is the case for soft robots. Because of their soft nature, they aren’t usually able to withstand high loads. (Just imagine a rubber beam or pillar - not the best structural support.)

So far I’ve managed to design devices that can move the wrist and fingers. But these are considered “lighter” joints. For the final piece though, it has to be capable of mobilizing the elbow - not an easy task. The elbow is significantly heavier and the bicep is capable of producing high torques during flexion. Regardless, I took a shot at it.

The first proto with a fold-based design

Initially, I drew up an actuator with a fold-based design. I tested out the prototype and realized that it was wildly insufficient to produce enough torque. The primary reason was that the size of the folds was too small. Second, the number of folds was not enough. The first problem was easy, just increase the height and width of the folds so that the surface area increases for the deflection to increase. But the second one was hard. There’s only a limited length on the limb to play with and the actuator definitely can’t be stretching that far because that would make wearing the exoskeleton very uncomfortable. So, in order to increase the number of folds, I had to revert to a bellow based design. While one fold is better in performance than one bellow, the reality is that for every fold, two bellows can be 3D printed. If other fabrication techniques are used, say molding and casting, the fold-based design definitely works better because of dimensional restrictions i.e. the walls can’t be too close together, there’s a limit to how thin the walls are etc. Doesn’t quite make sense the first time around but read this over and over and it will eventually make sense.

The final proto with bellows

Ultimately, the prototype was able to do what it needs to do and I was able to publish my results in a conference. The full paper can be found here.

I have to admit that reverting to a bellow design was not what I intended. It goes against my design philosophy and to me, it seems like I took a step backwards. But that’s just me. Any other engineer would have been proud that they were able to solve a problem. I guess, I’m more like an inventor who takes pride in my work more than anything else.