In order to maintain a safe working environment, there is a constant need for the inspection of structures for failure. One common problem faced during inspection is restricted access to the inspection site. Furthermore, inspectors use sensors like cameras and non-destructive testing kits to detect structural failures such as cracks and corrosion. Still, due to the size and complex geometry of most structures, inspection is expensive, time-consuming and potentially unsafe.

This thesis, therefore, describes the design of a robot that can be used for non-destructive Inspections of various range of structures to improve and automate structural inspections. In order to achieve this aim, different existing robotic designs have been reviewed, with focus on wall-climbing robots in similar environments. The final design combines a set of actuator motors, magnetic tracked wheels to aid vertical movement on magnetic structures, as well as a wireless camera for visual inspection and manoeuvring the robot system for the identification of surface cracks and corrosion using YOLOv4 machine learning algorithm. The structure of the tracks allows the robot to climb over uneven surfaces like bulwark, obstacles etc. which allows inspections in unfriendly and inaccessible environments therefore reducing costs and inspection time considerably.