ADOPTION OF COMPUTER VISION TO IMPROVE SAFETY PERFORMANCE IN SINGAPORE CONSTRUCTION INDUSTRY

Abstract

As one of the hardest-hit sectors due to the outbreak of COVID-19, Singapore’s construction sectors contribution to the nominal Gross Domestic Product dropped from 3.7% in 2019 to 2.7% in 2020 (New Zealand Ministry of Foreign Affairs and Trade, 2021). Although the sector is recovering from pandemic in 2021 driven by construction demand, it is plagued with WSH challenges (MOM, 2022c). Computer Vision (CV) applications have the capability to address this shortcoming and achieve the 2028 national Workplace and Safety Health (WSH) target. However, the level of CV technology adoption in Singapore’s Built Environment (BE) industry is low despite its potential. This dissertation presents an investigation into the Technology-Organisation-Environment (TOE) factors that limit the adoption of the CV system in Singapore Construction industry. A research method was employed combining the literature review, qualitative data collection and analysis. Case study, observations and interviews were conducted to get developer and user insights. QDA Miner Lite was used to group and analyse the identified factors, find hidden correlations, and organise recommendations. The main identified challenges were grouped into three categories and placed in order of the number of times it was bought up: technology, organisation, and environmental factors. This study will help stakeholders understand the major barriers hindering the adoption of CV systems in the Built Environment (BE) sector. The findings will devise recommendations to facilitate stakeholders in implementing CV technologies that enhance construction safety.