City-scale subsurface utilities mapping by multi-channel ground penetrating radar (MCGPR) with mobile mapping system (MMS)

Abstract

The complex and tangled networks of spider-web utilities present a persistent engineering challenge in many cities, emphasizing the need for improved maintenance to minimize societal disruptions and risks. Their locations, sizes, and conditions are unknown to many cities’ municipal authorities. Given their large amount and varieties, stock check and its constant update are necessary but hard to achieve because surveys are costly, inefficient at city-scale, and interfered with by road traffic. In recent years, such a stock check by MCGPR can be achieved with high efficiency, although ‘how?’ is still uncertain. This paper develops two hybrid solutions to answer this ‘how’: (1) a multi-channel ground penetrating radar (MCGPR) workflow from its city navigation to processing to GIS, and (2) matching with quality levels in the international standard (BS PAS 128–2022). The first solution is divided into above-ground and underground. Above-ground is seamless city trajectories based on firstly building a query image-basemap database by mobile mapping system (MMS) and then matching with subsequent on-board panoramic cameras for generating trajectories during MCGPR survey. Sky-view must always be restricted in a city where seamless positioning by GNSS-RTK is challenging. The underground part is a hybrid GPR-GIS workflow integrating the desktop planning, field survey, and data processing. The second solution contains two workflows: data classification of MCGPR, three data qualities based on BS PAS 128–2022, and attribute categorization of four scenarios, which are good & fair match, uncharted utilities, and survey Unreliable (SU)/Survey Not Successful (SNS)/Patch Utilities based on an integrated utility database by the Government. After the MCGPR survey of about 23,000 sq.m. rural area and 20,000 sq.m. urban areas in Hong Kong, we concluded that 20% and 34% existing records of point features (manholes and valves) and 50% and 10% existing records of polyline features (pipes) are ‘good or fair match’, respectively. This paper demonstrates how a stock check of busy city underground utilities can be achieved in a scientific, effective, and logical approach.