Prof Yu-hsing WANG

The Hong Kong University of
Science and Technology

Ir Prof WANG Yu-Hsing received his B.S. and M.S. degrees in Civil Engineering from National Taiwan University and Ph.D. in Civil Engineering from Georgia Institute of Technology, where he received the George F. Sowers Distinguished Graduate Student Award for Ph.D. Students. Prof Wang is a Professional Geotechnical Engineer in Taiwan and a Fellow of HKIE. Currently, he is  the Associate Dean of Engineering, Professor at the Department of Civil and Environmental Engineering and founder/director of Data-Enabled Scalable Research (DESR) Laboratory, the Hong Kong University of Science and Technology (HKUST). The DESR Lab is a Makerspace, specialized in the applications of Vertical AI, integrated with Geotechnical Internet of Things (Geo-IoT), Big Data Analytics, and Deep Learning, etc., on sustainable urban development and city resilience.

In 2005, he received the ASTM International Hogentogler Award. In 2008 and 2017, he received the School of Engineering Teaching Award, HKUST. In 2013, he received the Distinguished Alumni Award from the Department of Civil Engineering, National Taiwan University. In addition, he has been invited for Keynote and theme lectures in the international conferences and served as associated editors and editorial board members in different journals.


Smart System for Asset Management - Artificial Intelligence-based Building Information Modelling (AIBIM)

The Drainage Services Department (DSD) envisions providing world-class wastewater and stormwater drainage services to enable the sustainable development of Hong Kong. To support this vision, DSD keeps abreast of industry developments and explores ways to improve their day-to-day work through the application of innovative technologies. In this connection, DSD has collaborated with the HKUST to explore state-of-the-art integrated technologies, particularly artificial intelligence (AI), into implementing works projects and routine maintenance works.
DSD is currently responsible for the management of ~360 public sewerage and stormwater drainage facilities. Building and civil (B&C) assets like structural concrete of the sewerage and drainage facilities are significant assets to DSD. These concrete structures are often under an aggressive environment and are vulnerable to deterioration due to chemical and biological attacks. Maintenance programs, including regular inspections, are thus carried out to ensure the structures are well maintained to a serviceable state. Manual visual inspection is currently undertaken as the standard practice used in the principal inspections by the maintenance team. However, the practicality of very extensive and frequent checks is limited by time and labor constraints and occupational safety and health concerns over prolonged work inside sewerage facilities. Therefore, AI technologies integrated with BIM (AIBIM) were explored as a feasible tool to enhance the inspection and maintenance programs to facilitate the management of B&C assets. Such an innovation also help initiate a new era of intelligent stormwater and wastewater management.

In this presentation, details of implemented AIBIM pipeline will be shared, starting with data collection with a tailored-made, lightweight mobile mapping system. Then, procedures for data annotation of the images collected will be described. Details of the two proposed deep learning models for concrete defects identification, using the supervised and unsupervised learnings, will also be included, focusing on the model architecture, training, and evaluation. The AI-processed data thus automate updating the information of BIM to facilitate the subsequent assessment management.