HKU's AI System eCheckGo Revolutionizes Building Inspections in Seconds

Researchers at the University of Hong Kong (HKU) have developed an AI-powered system called eCheckGo to streamline building inspections. The technology uses advanced imaging and machine learning to detect structural defects quickly and accurately. Early tests show it could transform how cities monitor infrastructure at scale. The eCheckGo system works by generating interactive 3D point clouds of building facades. Inspectors can then examine defect locations and assess their severity in detail. A key feature is its ability to process dozens of images within seconds, identifying abnormalities with high reliability.

A pilot study in Kowloon demonstrated the system’s efficiency. It assessed 9,172 buildings in under four hours, proving its potential for large-scale urban management. The technology also cuts labour costs by eight times compared to existing automated methods, making continuous surveillance more feasible. At its core, eCheckGo combines a proprietary Large Defect Model (LdM) with traditional AI tools. This hybrid approach allows for multi-modal analysis, interpreting both visual data and textual inputs related to inspection standards. The LdM enhances accuracy by integrating semantic prompts that refine defect classification. The system’s mobile app lets inspectors capture images directly or pull from public datasets like Google Street View. This flexibility simplifies data collection while maintaining thorough coverage. Since its introduction, eCheckGo has attracted strong interest from government bodies and private firms looking for smarter monitoring solutions. Looking ahead, the HKU team plans to expand the system’s capabilities. Future updates may include detecting water leaks and dampness, as well as generating automated reports that meet professional standards.

eCheckGo has already shown its value in rapid, cost-effective building assessments. With ongoing development, it could soon handle a wider range of defects and produce standardised inspection reports. The system’s adoption by public and private sectors suggests a growing demand for AI-driven urban maintenance tools.