Matura Thesis

Entwicklung eines autonomen Roboterarm-Systems zum Pingpongspielen
Physik & Technik | Elias Bauer, 2006 | Fürigen, NW
Meine Arbeit beschäftigt sich mit der Entwicklung und dem Bau eines ...
Weiterlesen …

Abstract:

This thesis deals with the development and construction of an autonomous robot arm and camera system. Catching a ball using a net and volleying a ping-pong ball with a bat were chosen as application examples.

The self-developed autonomous system consists of three components: (1) Robot arm with (2) a camera system for dynamic image acquisition and (3) computer software for recognizing and localizing the ball in three-dimensional space, predicting its trajectory and controlling the robot arm.

The robot arm was designed and built using CAD software and a 3D printer.

The components of the system were optimized with regard to the planned dynamic applications, with a focus on speed, weight and robustness. The paper explains the selection of the appropriate components, such as high torque motors and high frequency cameras.

Additional solutions were developed and implemented for the two applications of the system: Among other things, methods are shown how the cameras use AprilTags to automatically determine their position in space, how the ball is recognized against its background, how measurement errors of the cameras can be minimized and how temporal deviations of the recordings can be overcome.

The paper also describes how the trajectory of the ball can be predicted using its recorded positional data and shows a method for calculating the optimal point at which the ball should be caught or played back by the robot arm. Furthermore, the paper deals with the calculation of the optimal racket position and sequence of movements to play the ping-pong ball back with a racket mounted at the end of the arm.

A series of tests show the successful application of the robot arm system both in catching a ball with an average success rate of over 87%, as well as in a ping-pong game with a return rate of around 90% and a rally of up to 20 strokes.

Finally, the multimedial presentation of the thesis with animations and real-time visualizations within Unreal Engine is described.

Thesis (German):