Humanistic Robotic Prehensile
About the project:
My project, the HRP Mk1, is a custom designed, 3D printed, humanistic, robotic prehensile. Though this is the first hand (and forearm) that I’ve designed myself, this project is a continuation of research that I’ve done previously. In high school, I downloaded and printed an open source arm design, then developed and applied a custom control method that allowed the arm to mimic my human input. This research and design project awarded me a trip to the 2016 International Science and Engineering Fair, where I earned a 4th place award in Robotics and Intelligent Machines. This was a very valuable experience that’s propagated my motivation and determination to continue my research in this field.
The next step I took, which is being showcased here, was to design a more flexible, dexterous, and humanistic hand. The eventual goal of this ongoing project will be to create a low cost, yet extremely capable, human-like arm for use in remote robotic control applications or as an effective prosthesis.
The main innovation with this current design is the utilization of special ball joints in the main finger knuckles. These unique ball joints allow for increased flexibility and movement, which improves grip strength and control ability. The hand is controlled like a marionette. Each of the five fingers are mechanized by four string tendons connected to two servo pulleys. Flex sensitive resistors are incorporated in a gauntlet I've created that collects real time data from a human user. Simple algorithms interpret this sensor data as finger positions and sends the appropriate commands to the servo motors. This combination of new age technology allows for the robot's hand movements to mimic the human's.
I’ve utilized the exotic 3D filament, NinjaFlex, which has a flexible and somewhat rubbery consistency. I printed the palm of the hand with this, which allows for the palm to bend and flex to fit with the contour of an object its grasping. My hand design employs twice as many servo motors than the previous hand I had used. This provides an entire new dimension of movement for each finger, taking one step closer to capturing the immense complexity of the human hand.
There’s already ideas for improvement and new innovation flooding through my mind. I’m going to create a new sensor glove for more precise control, and look into ways to incorporate haptic feedback. I plan to develop the next version of this hand model by next year.
Any donations or support of any kind would be greatly appreciated.