Components
In-depth view of our project
3D Hand Design using SolidWorks
Our hand is currently designed to fit all of the mechanics needed to play Rock, Paper, Scissors in the palm of the hand. There and strings that run through each finger, are held in place by screws, and are attached to servo motors which control each finger. Although on a larger scale, the hand is modeled to look like a robotic version of the human hand.
.png){kind=small}
Servo Motor/motion control using Arduino
The basic mechanic principle behind controlling the finger movements in the hand is tension. Each finger is connected to a servo motor via a string that runs through the finger. To change the position of the fingers, the servo motors spin which creates tension in the string, forcing the finger to either open or close. Currently, for the sake of the game, the servo motors are coded to spin according to the positions of either rock, paper, or scissors.
Machine Learning to train the hand
For the Rock Paper Scissors gameplay component of this project, we have implemented an algorithm based on a study found in a paper on human conditioning. The paper titled Social Cycling And Conditional Responses In The Rock-Paper-Scissors Game found that humans do not play exactly “randomly”, but rather more than half the time follow certain patterns-based on if they win, lose, or tie. The code implements this study’s findings and connects with the Bionic Hand along with the hand tracking feature to have the Hand imitate the computers choice and acknowledge whether the Machine or Human has won the game. Altogether, the Machine acts as a functioning intelligent Rock Paper Scissor Player.
Join our Neural Network
We have undergone the IRB review at The Cooper Union and were permitted to collect anonymous images of the cooper union community member's hand gestures in order to create a large data set to increase the efficiency of our neural network that trains the hand to recognize the gestures of rock, paper, and scissors. .