Current Limitations

1. Thumb actuation in progress. I have 6 more open air channels for the thumb and other miscellaneous functions. However, a better design of a proper saddle-joint for the thumb base/CMC may be required for a V2.

2. Design is meant to mimic hand bones, ligaments, tendons, and muscles. The complexity is not always practical, rather proof of concept.

3. Current actuation is absolute open/close of cheap solenoid valves. Proportional control has not been tested yet.

4. Tire compressor is old and loud.

5. I have yet to write a more robust control method beyond just buttons, perhaps some tuning that overcomes inconsistencies in the custom ligaments/antagonists if proportional control is possible with these solenoid valves.

McKibben Actuator Limitations

1. The nature of soft actuators is lower force output per unit when compared to an electromagnetic motor, as well as a much smaller range of motion. It is possible perhaps to use more advanced configurations to increase range of motion as well as McKibbens in parallel to multiply force output, but I have yet to test this.

2. Pneumatic or hydraulic soft actuator systems typically require a bulky reservoir (in the case of this project, a 0.5L air tank) which may add a lot of mass.

However, there was a solution developed by Ozgun Kilic Afsar, involving an electro-hydro-dynamic system using a charged fluid and a very high voltage to transfer fluids between agonist and antagonist McKibbens, without the need for an external reservoir. I’d very much like to explore this deeper when I have the time.