Viability of off-site inspections to determine bridge defect ratings

This paper examines the viability of remote, image-based assessments to rate bridge defects in the UK. General Inspections are typically undertaken on all bridges in a stock every two years and is one of the primary methods of stock assessment by bridge owners. However, past studies have shown that they are subjective and potentially unreliable. This inspection process could be split up into six fundamental stages: (1) planning & preparation, (2) image capture, (3) defect identification, (4) defect grading, (5) interpretation of change over time, and (6) maintenance decision-making. In current UK industry practice, going through these stages is often a labour-intensive manual process. By framing the inspection workflow in this way, work can be done to investigate if these stages can be ‘operationalised’, without reducing the reliability of the overall process. This study explores the factors that affect the accurate obtaining of defect ratings, where image capture and defect interpretation are explicitly separated. A simple comparison of defect ratings obtained from an on-site and off-site inspection is carried out on four defects. The results were found to be similar, with plans being made to expand this into a large-scale trial. Barriers to a remote, image-based inspection are discussed and a potential future architecture for an inspection workflow is suggested.