demoConstruct

posted on 28 Feb 2025 by leon

democratizing Digital Twins with Progressive 3D Reconstruction

Brief

demoConstruct is an open-source platform that enables accessible and collaborative scene authoring, powered by progressive 3D reconstruction.

Designed to accommodate both professionals and untrained users, the experience is delivered through interfaces that run on common consumer devices.

demoConstruct’s progressive 3D reconstruction approach enables near real-time 3D reconstruction which supports simultaneous parallel operations across diverse user roles:

  1. scene capturing and digitization using a mobile camera,
  2. scene editing on a desktop PC, and
  3. scene exploration and inspection on a VR headset.

This platform powers many different use cases of digital twins and one of our current focus is in the healthcare sector, where we are working with Tan Tock Seng Hospital (TTSH) in Singapore.

Members

SIT

TTSH

Outputs

Usable Artifact

Open-source platform for collaborative scene authoring with progressive 3D reconstruction:

https://github.com/singaporetech/immersification-demoConstruct

Publications

Leon Foo, Nirmal S. Nair, Liuziyi Liu, Jeannie S. Lee, Songjia Shen, Indriyati Atmosukarto, Alvin Chan, Jing Shi, Yong Joo Loh, Yih Yng Ng, Michael Chia, and Chek Tien Tan. 2025. Collaborative Scene Authoring with Near Real-Time 3D Reconstruction. In CHI Conference on Human Factors in Computing Systems (CHI ‘25), April 26-May 1, 2025, Yokohama, Japan. ACM, New York, NY, USA. (preprint)

Leon Foo, Chek Tien Tan, Liuziyi Liu, Nirmal Sukumaran Nair, Songjia Shen, and Jeannie Lee. 2024. demoConstruct: Democratizing Scene Construction for Digital Twins through Progressive Reconstruction. In ACM SIGGRAPH 2024 Labs (Denver, CO, USA) (SIGGRAPH ‘24). Association for Computing Machinery, New York, NY, USA, Article 2. https://doi.org/10.1145/3641236.3664424

Leon Foo, Songjia Shen, Ambrose Wee, and Chek Tien Tan. 2023. Progressive 3D Reconstruction for Collaborative Construction of Digital Twins. Proceedings of the AAAI Symposium Series 1, 1 (Oct. 2023), 7–10. https://doi.org/10.1609/aaaiss.v1i1.27466