Session: 15-01-01: ASME International Undergraduate Research and Design Exposition

Paper Number: 94633

94633 - The Ai Rock Paper Scissors Machine 

The research we are proposing is a project creating an AI Rock Paper Scissors machine. The AI Rock Paper Scissors machine combines both mechanical design with computer software to create a functioning hand that a user can walk up to and play rock paper scissors against. The hand, has a display and computer attached to it that runs software to read the user’s move, dictates to the hand what move it should play, and outputs the stats on the display. There are two artificial intelligence components to this design. The first, the computer’s strategies, deals with the computer learning from previous Rock Paper Scissors games for it to develop its own strategy for playing. The second, the hand tracking, deals with the computer using a camera to take instant pictures of the user’s hand to determine if the user is playing rock, paper, or scissors. The goal of the AI Rock Paper Scissors machine was to create a fun, user friendly game through a bionic hand, while testing and learning how artificial intelligence can improve the functionality of the design. This research not only demonstrates the design and creation of a mechanically built game, but also contributes to the advancement of science as it explores the intersection between mechanical design and artificial intelligence, and how their combination can greatly improve desired results.

 The mechanical aspect of the project deals with the design and functionality of the bionic hand. The hand was designed using Solidworks, modeled after a human hand with hollowed out holes to allow for strings to flow through. These strings are attached to servomotors within the “palm” of the hand that cause the strings to tighten and loosen, which by utilizing this tensile force the strings force onto the hand, opens and closes each finger. The methodology behind this design involved analytical calculations and iterative testing. Analytical calculations were used in solving how much power to send to each servomotor and the relative distances in the design to achieve the desired bend angle of each finger. Iterative testing was used in this component of the project to test between different strings and design specifications of the hand, to achieve maximum results.

The software and artificial intelligence aspect of this project deals with the strategy of play in Rock Paper Scissors and the collection and user inputs. The computer plays Rock Paper Scissors based on an AI algorithm, in which the computer choses its next move based on data it has collected from previous games. To collect the user’s move, the computer uses a separate algorithm to do hand recognition and identify what the user has selected. After both the user and computer do their moves, the computer evaluates who has won the game and saves the data. When integrated with the mechanical bionic hand, the software drives the Rock Paper Scissors game through the hand, creating a Rock Paper Scissor machine, in which a user plays against the bionic hand.

Altogether, this research explores how the integration of artificial intelligence can help improve and shape mechanical designs. By using artificial intelligence in this project, the rock paper scissors game improved through hand recognition and the ability to collect player data and learn from it. Just as these features were used to enhance the gameplay experience for Rock Paper Scissors, there is great potential for these features across other mechanical designs and projects.

Presenting Author: Evan Kluger Student

Presenting Author Biography: Evan Kluger is a sophomore mechanical engineer at Cooper Union.

Authors:

Evan Kluger Student
Benjamin Meiner Cooper Union
Michelle Katz Cooper Union
Kanghyuk Lee Cooper Union