Man-machine balanced control for automation of infrastructure crack sealing.

摘要

Aging of our infrastructure is most visibly exhibited in the form of surface cracking. During construction or service of the infrastructure, cracking is a warning that should not be ignored. Sealing surface cracking is one effective strategy utilized by maintenance authorities around the country. As a result, sealing cracks in pavements, buildings, and other elements of the infrastructure is a nationwide activity. Despite or because of its ubiquity it is often overlooked, however automating crack sealing can significantly improve safety, productivity and quality. Drawbacks exist with current crack detection, mapping, and sealing methods which are either solely autonomous or solely manual. A man-machine balance can instead be used, taking advantage of the respective strength of man and machine. Such a balanced control loop was developed for an Automated Road Maintenance Machine (ARMM). Major research problems and some of the design issues involved in automated crack sealing systems development are discussed when describing the ARMM's man-machine balanced control loop. Then, this dissertation presents an innovative machine vision algorithm to accurately map surface cracks to be sealed and another to automatically traverse the represented crack network. The efficiency of the man-machine control loop including the algorithms is evaluated, and the productivity of the ARMM is also measured and presented in this dissertation. Based on the results, the man-machine control loop and the machine vision algorithm developed for the ARMM are proposed as a good model for automation of infrastructure crack sealing.;This dissertation also examines the potential broader usefulness and applicability of the man-machine control loop and the machine vision algorithm by evaluating images from other infrastructure areas. It is concluded that the man-machine control loop and the machine vision algorithm presented in this dissertation should be applicable to crack and joint sealing in buildings, pavements, bridges, tunnels, and utility pipes, etc. It is also anticipated that this dissertation research can provide context and structure for the development of new algorithms and tools for automation of infrastructure maintenance or inspection.