Design and Development of Electro-Mechanical System of Intelligent Litter Picking Robot

Low, Yi Hui (2016) Design and Development of Electro-Mechanical System of Intelligent Litter Picking Robot. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text LOW YI HUI Full Text.pdf Restricted to Registered users only Download (3MB)

Abstract

The environmental cleanliness has been a major concern to the society nowadays in order to secure the healthy lifestyle of human beings. EcoBots, an interactive robot to pick up the litter at the sea shore has been an inspiration towards the litter picking concept and this idea is then being translated into a much clearer and easy-to-aim objective, which is to pick up an object with different shapes and colours and then, categorise them according to their appearance characteristics. The major purpose of this project is to build a prototype of robotic arm which can carry out pick and place of an object via identifying its characteristics with the intelligence system developed. Deep down into the focus of this project, it can be further divided into two main deliverables: Electro-Mechanical system of the robotic arm and also the image processing techniques implemented to carry out the specific tasks. Therefore, this project aims not only on the litter picking approach, but also several industrial manufacturing industries which implement robotic arm to identify the shape of the manufacturing product at the conveyor. It is hence, covered a wide range of fields in robotic arm perspective. The methodology used in designing and developing the prototype of robotic arm includes the mechanical structure, embedded system, and also program writing. From the mechanical structure perspective, the first stage is to draft out the design of robotic arm and its picking mechanism type is identified. The next step will be drawing the design drafted into SolidWorks Software with accurate dimensions and designs. The SolidWorks drawing will then being analysed and modified in order to fit with respect to the screws and nuts used. Fabrication of the mechanical prototype of the entire robotic arm is done using 3D Printers and knowledge of G-code has been in implementation. It then being connected with the electronic components such as servo motors, Raspberry PI controller, Raspberry PI Camera Module to control the actuator movement of the arm. The camera module fixed will eventually acts as the eye of the robotic arm in order to capture and track the desired object which provides an output of pixel coordinates in x-axis and y-axis format. The output is then being processed in a series of calculation by the image processing algorithms such as gray-scaling, thresholding, edge detection and so on. An OpenCV software is being used coalesces with the Python programming language written will eventually assign and control the servo movements based on the object tracked. The core of the project in term of image processing further develops on deliverables such as colors detection, shapes detection, distance measurement between objects. In a nutshell, this project tends to provide understanding, knowledge and tactful skills regarding to the developing of an intelligence robotic arm which can perform a specific task by its own interpretation. Further development on a picking robot should be machine learning on Artificial Intelligence perspective that could recognise on any objects and determine the material type of the objects.