While the project began in July, the actual prototyping process was begun in late September. The focus of the Gen I prototype was help us determine the number, layout, and specification of the sensors we would need as well as a general form factor for the device.
The design consisted of a housing that would rest on the back of the hand, with 5 flex sensors and 1 force sensor. The flex sensors would be inside of cloth sleeves that would tie to the fingers, while the casing would be Velcroed in place. The force sensor would be placed on this Velcro strap. The OTC would be strapped on top of the housing.
The major flaws with this were all related to the form factor. The OTC sat far too high above the hand. This caused more torque than the Velcro could handle. Additionally, it was much more weight on the hand than we wanted. The flex sensors strapping method was also very tedious and not very secure. Finally, we decided a second force sensor on the thumb would be needed in order for all the required tests to be assessed.
To improve upon Gen I, Gen II implemented a Silicon superstructure that would fix the sensors in place. This superstructure was light weight, flexible, and elastic so that it would mate well with the patients hand and not impede their movement. The structure was created by 3D printing a reusable PLA mold and casting the silicone into it. The flex sensors were embedded into this structure. An additional force sensor was added to this design and both force sensors were placed in a PLA casing for added rigidity. These casings also worked as a magnetic clasp that would hold one of the thumb straps and the palmer strap together. These clasps also allowed for an adjustable size. The electronics were moved back to the forearm to reduce the weight on the hand, but the OTC was left on the hand using an independent Velcro mount (not pictured). The electronics now included a power switch and a button to begin calibration.
The pros of this design were the weight distribution, the stability of the flex sensors, and the sensor layout. The cons were the cable management, the casing being too small, the silicone feeling a bit fragile given how thin it was printed, and the over all fit of the design. After some feels-like feed back from the patient, we took 3 more weeks to redesign for Gen III.
After testing the Gen II glove, several design limitations were pointed out which were implemented in the Gen III system modification. First and foremost, the glove was too heavy for use by the patient so the system was adapted to be lighter and more maneuverable. This was accomplished by using less silicon and minimizing weight in other aspects of the design such as the circuit casing and the attachment mechanism. Additionally, to increase the usability of the system for our test case, the glove was redesigned for the right hand (dominant hand) so that the patient could use the device more effectively. Finally, the sensors were not being optimally activated as the distal attachment point for the straps was not located at the finger tips. This disallowed the sensors from being fully flexed and ultimately diminished the input of the device to the microcontroller.
The major drawback of the current design for the glove is that its weight is not suitable for use by both hands. To minimize weight in the future, the microcontroller housing could be attached to something besides the patient. This would significantly reduce the weight of the glove and make it usable for both hands. Additional improvements would involve wire management and other aesthetic changes to make the device more marketable.