Why does the development of assistive devices have to be constrained to engineering firms and research institutions?
While there is already a market for robotic feeding arms, this one democratizes the technology by using open-source technology; getting it to those who really need it at as low of a price as possible.
Prototypes of the arm made from insulation foam and wooden dowels lay the foundation for the forms we will design down the road– the limitations and areas for improvement. These models aren’t self actuating but they give a great indication of the motion and scale.
Some earlier versions of the arm, where initial bugs were worked out before a final mechanics study.
Hardware and Software
At this stage, there are actually two types of experiments and testing, reflecting the challenges of both parts.
The end goal is to use analog potentiometers to track the position and actuate any motion with servos. This will allow the arm to record a final position set by a care-giver where the food is close enough to the mouth of the user. Once the position is set, it can be return to that position by flipping a toggle switch until a new position is recorded.
By aligning the shaft of a potentiometer with the axis of the servos, we can map the voltage to the range of the servo’s sweep.
And from a hardware standpoint, the chasis of the arm, and the placement of the servos needs to be tested and then motion tested to figure out the break points in the design.
The concept shows promise although a redesign is necessary. There is room to optimize weight and higher quality servos will streamline programming the arm so that the training feature can be implemented more efficiently.
After stepping back and looking at the failure points in version 1, we decided to rely on the make block platform rather than try to frankenstein our way into a brand new platform. This has given us the time to focus on the nuances of the firmware for the arm and looking into how to add the potentiometers to track position.
For this to be a viable kit, we need to continue fine-tuning the programming and are looking into the relationship the arm has with the utensils being used.