Investigation of digital twins and metaverse

• Main Author: Chow, Harry Yun Fatt
• Other Authors: Cai Yiyu
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; BIM
• Online Access: https://hdl.handle.net/10356/181509
• Institution: Nanyang Technological University

Building Information Modelling (BIM) is the digitalisation of a building or structure's physical and functional characteristics, representing them in a virtual space [1]. It is a process that involves generating and managing digital versions of places and structures. BIM is more than just a 3D model; it encompasses building geometry, spatial relationships, geographic information, component quantities and attributes (e.g., manufacturer information), and allows for the integration of various types of information (e.g., cost estimates, material inventory, project plans) integrated into the model. Individuals, companies, and governmental organisations use BIM to plan, design, build, operate and maintain various physical infrastructures, including water, waste, electricity, gas management facilities, roads, bridges, ports, tunnels, etc. The BIM market is valued at USD 7.9 billion in 2023 [2] and is projected to reach USD 15 billion by 2028[3], indicating a compound annual growth rate (CAGR) of 13.7% [4]. Currently, 73% of construction companies use BIM in their projects [5], and countries like the UK, USA, and Singapore have mandated the use of BIM in public sector projects [6]. BIM adoption is on a steady growth trend throughout various regions and sectors. The growth of popularity for BIM represents a shift towards the digitalisation of how buildings and structures are supposed to be planned, designed, constructed, and managed. In recent years, innovations like integrating BIM technology with laser scanners have introduced the concept of scan-to-BIM (Scan2BIM) [7], allowing 3D models to be created and generated from physical spaces through the scanners. Although this has greatly benefited the industry, this approach has its challenges. The process requires the raw point cloud data from the LiDAR sensor to be manually processed and quality-assured by an engineer, which is a tedious, labour-intensive and time-consuming operation, leading to higher costs and longer project lifecycles. This leaves a gap in the market for a more automated and efficient process. Bridging this gap would significantly improve not only the efficiency but also the competitiveness and sustainability of the BIM industry. Over the years, numerous companies and start-ups have been looking to improve this process and leverage the advancement of artificial intelligence (AI) capabilities. CHiLX is a project headed by Nanyang Technological University (NTU) Associate Professor Cai Yiyu and has been developing and researching AI solutions for the past five years. Identifying this gap in the market, CHiLX has developed a solution to automatically and accurately transform point cloud data from LiDAR scanners into fully functional BIM Models with the help of AI and Machine Learning (ML) technologies. Being able to cut down labour hours by up to 80% not only addresses the issues of efficiency and cost linked to existing scan-to-BIM procedures but also establishes itself as a pioneering development in the field of BIM. The primary goal of this final year project is to bring the solution developed by CHiLX to the market and commercialise it. The project scope includes a comprehensive strategy that involves business development, developing go-to-market solutions, marketing strategies and a robust business model through extensive market, competitive, and industry research. This project aims to transform CHiLX from a concept solution into a commercially viable product. This report will record the methodology and frameworks used to commercialise this solution and turn CHiLX into a successful start-up.