Research Areas

Analytics, Machine Learning and Computer Vision

  • Computer vision

    Currently experimenting with video data to automatically identify hazards and construction issues on site. Interested organisations can contact us for more information or visit the project page to find out more. Alternatively, you can check out Chian et al. (2021) to read on approaches for detecting missing barricade, Chian et al. (2022) on dynamic identification of crane load fall zone, and Lim et al. (2023) for "cluttered" construction housekeeping images through supervised learning.

  • Developing construction safety leading indicators through machine learning

    We has successfully developed a machine learning algorithm that can provide an indicator of project safety risk for an A1 contractor in Singapore. Results of the machine learning approach can be found in Poh, Ubeynarayana and Goh (2018).

  • Text mining

    We have applied machine learning methods to automatically classify accident descriptions. Using text mining techniques, organisations can better employ text documents in their safety analytics to gain useful insights to prevent accidents and ill health. See Goh and Ubeynarayana (2017) for more information.

  • Construction safety games

    We designed and developed a series of 3D digital simulation games throughout the years to teach learners on the importance of safety. SafeSim Risk (SSR) is a multiplayer game designed to train players to identify hazards and good practices on a virtual construction site. Do check out the project page, Lee et al. (2020) or Safiena & Goh (2022) for more information. SafeSim Investigation (SSI) teaches and train players to conduct an incident investigation by collecting evidence to unlock clues. Do check out the project page or Safiena & Goh (2020).

  • Using VR/AR/MR to teach

    Developed a virtual reality (VR) and mixed reality (MR) technology learning tool for incident investigation, do check out RESCUE Act 3 and Yang & Goh (2022). We also developed a mobile application utilising virtual and augmented reality to teach non-engineering students structural systems. For more information, check out the VARS page. Alternatively, you can see Hu, Goh & Lin (2021) for the educational impact of AR and Hu et al. (2023) for impact of VR360.

  • Digital game-based learning

    We developed digital games based on thematic findings from systematic literature review for adult learners and develop validated instruments to measure effectiveness. Do check out Tay et al. (2022) and Safiena & Goh (2024) for more information.

Education and Training

Safety Climate and Design for Safety

  • Simulation game for designers

    DfS is a process whereby the stakeholders of a construction project come together at the earliest opportunity during the different project stages to identify and eliminate or reduce foreseeable design risks throughout the project lifecycle. SafeSim Design is a DGBL software for designers to learn about design risk and how to mitigate them. The content of SSD is based on the IES-NUS DfS Library. For more information, you can check out Safiena et al. (2023). 

  • Design for Safety (DfS) climate

    This project aims to develop and measure Design for Safety (DfS) climate. DfS is an upstream approach to improve construction safety and health, through anticipating and designing out potential hazards. DfS climate refers to the shared perceptions members of a DfS review team have towards the policies and procedures of DfS. For more information, check out Lim & Goh (2023).

  • Safety climate

    Using a mixed-methods approach to understanding safety climate and the strategies to improve safety climate among truck drivers. For more information, check out He et al. (2022) to examining safety climate among truck drivers and Sun et al. (2024) using interpretable clustering approach.

  • Investigating the relationship between national culture and safety culture

    Influenced by the foreign labour policy of Singapore, the construction industry relies on many migrant workers from nations such as China, India, Malaysia, Bangladesh and Myanmar. Migrant workers make economic contributions, but they still have poor safety outcomes. We need to do more to improve safety conditions and protect them. We aim to investigate how dimensions of national cultures affect safety culture dimensions and propose a range of possible behavioural modification interventions that leverage on the understanding of the relationship between national cultures and safety culture.

  • FPSWizard

    We developed a mobile-based design support system to facilitate the selection and design of fall arrest and travel restraint systems for building construction activities. A prototype system has been developed and a series of drop tests conducted with BMS Training Specialist and Tractel Singapore to validate the calculations. Do check out the project page, Tashrif et al. (2022), and Lim et al. (2019).

Fall Protection