We developed a prosthetic rehabilitation device to help people who have limited mobility in their hand musclces. The device allows for three different grips that will assist users in daily tasks.
The bioelectric energy is represented with green arrows, the mechanical energy with bright pink, the power with light pink, and data lines with blue. The heart of the system is the battery, which fuels the power to the pico, sensors, and motor boards. The sEMG sensors and sensor pads work together to get bioelectric data from the arm into the system, and the machine learning model uses said data to decide which grip to switch to. The data is directed into the motor driver boards, and then, the linear actuators themselves, resulting in mechanical movement.
Click to enlarge the image.
We used sensor pads to collect data from the arm, with the signals informing the device which gesture or grip the hand would like to make. The data is then sent to the pico, and later the motor driver boards and linear actuators, resulting in the desired mechanical movement.
Mechanical Electrical Firmware SoftwareOur final Myo Amp prototype did not perform as intended due to a combination of hardware limitations and logistical challenges. The motor driver we initially selected was not powerful enough to operate two linear actuators, which resulted in component failure during testing. In the future, this will be addressed by selecting a motor driver with appropriate power capabilities. Additionally, delays in receiving critical sensors significantly extended our development timeline, reducing the time available for testing and iteration. While we were able to make partial hardware substitutions to continue progress, these constraints limited the reliability and overall performance of the final system.
Sprint 1
Inital idea, components, and sketches
Sprint 2
A working prototype
Sprint 3
Last run: final assembly and debugging
