top of page

Technical Workshop |Munich

Dr. Li was invited to give a technical workshop on the “Unlocking the Power of Language Models: Artificial Intelligence for Smart Configuration in Architecture, Engineering, and Construction (AEC) Industry” at Technical University Munich on 14 Aug 2023.

Dr. Xiao LI was invited by Prof. Dr.-Ing. Frank Petzold, the chair of architecture informatics from the Technical University Munich under the HK/Germany Joint Research Scheme and delivered a seminar on August 14, 2023 with the topic of “Unlocking the Power of Language Models: Artificial Intelligence for Smart Configuration in Architecture, Engineering, and Construction (AEC) Industry.” After the seminar, they discussed potential collaboration opportunities in European research project and joint research fund between China and Germany.

Abstract
The convergence of Industry 4.0 demands and the rapid advancement of artificial intelligence technologies presents unprecedented opportunities for the architecture, engineering, and construction (AEC) industry. Natural Language Processing (NLP), the flagship technology of artificial intelligence, offers numerous tailor-made solutions for smart design and construction in the AEC domain. Among them, one crucial aspect is the characterization of user needs, which typically involves transforming unstructured user texts into structured information, facilitating mass customization within the realm of smart configuration. However, current approaches to understanding user needs usually rely on heuristic or traditional machine learning methods, resulting in significant errors and limited accuracies. This seminar will introduce a novel perspective on smart configuration for mass customization by incorporating language models into user needs understanding. Specifically, we begin by framing the user needs understanding task as a spoken language understanding task, addressing practical challenges through concrete methodologies and experimental designs. In this process, we identify two key roles for language models in smart configuration: serving as knowledge bases and functioning as backbone models. Additionally, we discuss the opportunities and challenges associated with employing language models in this domain.

Biography
Dr. Xiao LI joined the Department of Civil Engineering, The University of Hong Kong, as an assistant professor in December 2022. Before joining HKU, he was a Research Assistant Professor at HKPolyU (2021-2022), an RGC Postdoctoral Fellowship awardee at HKU (2020-2021). He was also a visiting scholar at the University of Cambridge. His research interests focus on construction industrialization and construction informatics. He has led six research projects (e.g., NSFC, GRF, HK-Germany Joint Research Scheme) as PI with funding exceeding HK$ 5 million and has authored 40+ papers in peer-reviewed academic journals with 3100+ citations (4 ESI highly cited). He is a fellow of the SYLFF Association, a National Registered Construction Engineer (Class 1), a member of CIB and the American Society of Civil Engineers (ASCE), and guest editors of several leading journals in construction engineering and management. He held several international academic awards, e.g., CIC Innovation Award 2022, SYLFF Research Grant Award, Research Abroad Award, CRIOCM Outstanding Paper Award, CIB Sebestyén Future Leaders Award, ASCE Best Paper Award, and CIOB (HK) Outstanding Paper Award. His previous research mainly contributes to the decentralized adaptive work packaging methodology for collaborative planning and control in industrialized construction. Firstly, he investigated graph-based work package generation mechanisms for complex products of industrialized construction. Then, he made a breakthrough in the stochastic optimization method of work package sizing under uncertainties. Finally, he developed a blockchain-enabled smart work packaging system for crowd intelligence in collaborative planning and control.

Project Gallery

Copyright © 2024 CI³Lab, Department of Civil Engineering, HKU. All rights reserved.

bottom of page