Enhanced user interface for the control of a deep tunnel robotic platform

• Main Author: Yong, Rui Sheng
• Other Authors: Seet Gim Lee, Gerald
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Mechanical engineering::Mechatronics
• Online Access: https://hdl.handle.net/10356/141748
• Institution: Nanyang Technological University

This Final Year Report details the author's work in creating a graphical user interface (GUI) for the user to inspect and review the data collected by a deep tunnel robotic platform. The platform would be utilized by Public Utilities Board (PUB), Singapore. The deep tunnel robotic platform developed by the NTU team, can operate in dangerous and hazardous environments such as in sewerage tunnels, allowing operators to safely perform inspections of the tunnel without risks to personal safety. Front and rear cameras are installed to allow operators to drive the platform remotely, and an array of three cameras are positioned such that visual data of the whole tunnel section can be adequately captured. During the robotic platform’s deployment, a series of images and videos were collected. These raw visual data must be processed and organised before the operator is able to make sense of it. The images go through an image processing process which removes redundant frames and undergoes image stitching, allowing visual information of the tunnel to be displayed in one large flat image, similar to a map of the tunnel surface. The GUI developedby the team, allows the operator to navigate through the stitched image for suspected defects in the tunnel and validate their findings easily by referencing video data captured by the platform. It does so by linking the location of the suspected defect in the stitched image to the appropriate video frame. In addition, the GUI also demonstrates that various tunnel defects can eventually be detected automatically through the interface in future works. Currently, a functionality that allows for automatic detection of numbers painted within the tunnels has been implemented.