2023 ASEE Annual Conference & Exposition

With the growing interest in Modeling and Simulation and lack of institutions with a degree program in the discipline in our region, designing, promoting and implementing a degree program in an institution that traditionally attracts students from the minority population will represent a viable pathway to increasing the participation of underrepresented minorities in this emerging field. After an extensive search we could not identify any Minority Serving Institution (MSI) that offers a degree program in Modeling and Simulation in our region. This paper will present the Modelling Simulation project which will help students learn the concepts of Modeling and Simulation. The modeling and simulation project work is supported by the grant from the Department of Education. To successfully train in table tennis, players must practice returning precise shots and developing a consistent play style to deal with the game's varied paces. Table tennis robots are used by players who want a guaranteed series of shots or a pattern of shots that do not change in angle or speed. These robots fire ping pong balls at the player in a controlled manner, typically by using a motor to fire the ping pong balls at certain angles in sequence. There are also robots that can make fixed and random serves, providing a comparable sensation to competing against a real player. The Ping Pong Bot* project tries to make this procedure more dynamic by using a model trained to recognize human torsos and hand signals to track the player. The goal is for the robot to engage the player in three different modes. The first mode involves the robot tracking the player and serving ping pong balls to them at their location. The second phase involves the robot purposefully shooting balls at areas where the player is not present. The third mode alternates between serving ping pong balls at the player's position and serving ping pong balls at areas where the player is not present. For this reason, a convoluted neural network is utilized, and the model created was simulated with matplotlib, where the model's accuracy was tested. The raspberry pi 4 b+ serves as the robot's brain, on which a generalized neural network was trained to detect the shape of a human player as well as hand signals for numbers one, two, and three. Servos are used to feed the firing chamber, agitate, and position the ping pong balls, and adjust the angle at which the firing mechanism is faced. The ping pong balls are fired by a brushless outrunner motor.