Collapsing sewers are often the result of voids which arise from cracks In the sewer pipe. The water that flows in or out of the sewer has the ability to wash away soil and thus may result in voids. The lack of foundation around the sewer ultimately results in a higher probability of collapsing. This master thesis comes forth out of The Technology Innovation for Sewer Condition Assessment - Long-distance-information-system (TISCALI) project. It aims at utilizing, integrating, and further development of relatively low-cost, of the shelf, techniques to create an objective and quantification of defects in sewers and to determine their constructive strength and stability. This master thesis project will focus on one of these techniques called Ground Penetrating Radar (GPR).
Nowadays GPR is operated by a specialist from the surface to generally locate objects in the sub surface environment, which is time consuming and subjective work. GPR uses electromagnetlc waves which have a penetrative ability, and thus can acquire data about the refelctive properties of the surrounding material. The data is actually the reflection of the transmitted signal that reflects on contrasts or dielectric properties from different materials, after which it is received by the antenna. A robot that operates from within the sewer can be equipped with a GPR. The advantage of doing data acquisition from within the sewer Is the reduction In distance to the voids, which facilitates the use of higher frequency systems with a higher resolution. The generated radargram can be used to train a model that can automatically classify voids without relying on a specialist and makes for a potentially more time efficient and less subjective system.
This project will focus on developing a void detection algorithm with the ability of classifying and locating voids in the absence of GPR specialists.