Developing a reusable and extensible digital platform for maintaining dynamic urban digital twins (UDTs), with a focus on integrating built infrastructure with its surroundings. 

Lead Researcher: Prof. Chi Tsun (Ben) Cheng 

Researchers: Toh Yen Pang, Azizur Rahman and George Wu 

Keywords: Digital Twin, Urban Planning 

Overview

Urban digital twins can help researchers and planners understand how buildings, infrastructure and environmental systems interact. However, creating useful digital twins requires information from very different sources—including building and terrain data, engineering models, drone-based photogrammetry and live sensor feeds—to be brought into a consistent and maintainable environment. 

This project developed a reusable framework for integrating these different forms of information. The framework addresses practical barriers that commonly limit digital-twin applications, including fragmented data, incompatible systems, scalability and the difficulty of keeping models connected to changing physical conditions. 

The research progressed through three complementary applications:

1. Geospatial integration — Building, satellite, terrain, and 3D model data were integrated in NVIDIA Omniverse, using path-tracing to simulate sunlight and shadow.

2. Real-time monitoring — Bidirectional communication via the MQTT protocol connects the digital environment with physical sensors and assets for real-time monitoring and control.

3. Agrivoltaic application — The framework was applied to an agrivoltaic setting, combining a photogrammetric orchard model, a solar-system engineering model, and autonomous robotic systems.

Together, these applications demonstrate how one digital-twin framework can connect geospatial context, detailed engineering models and live systems, supporting analysis and decision-making across urban, agricultural and infrastructure challenges. 

Outcomes, Engagement and Impacts 

• Physics-Based Digital Twin - LinkedIn  | Video

Industry recognition with Physics-Based Digital Twin LinkedIn post (May 2026):  The Physics-Based Digital Twin LinkedIn post generated 1,972 impressions and reached 1,002 members, with 5 reposts. The post attracted reactions from senior industry and sector leaders, representing significant external validation of the project's technical direction: Vice President, NVIDIA; Engineering Director, NVIDIA; Director, NVIDIA; CFO, Sunflow Energy; CEO, Food Agility CRC; Managing Director, Memko System

• Multi-scale Urban Digital Twin - LinkedIn 

• System Engineering Digital Twin  - LinkedIn |  Video​
  
  

Contact:

Prof. Chi Tsun (Ben) Cheng - ben.cheng@rmit.edu.au