USC Ginsburg Hall Digital Twin
Research-grade living lab digital twin — multi-source IoT streams fused with 3D building geometry for standardized evaluation and reproducible building research.
Outcome. Building a research-grade digital twin of USC Ginsburg Hall — integrating live IoT sensor streams with 3D building geometry to enable standardized data collection, reproducible evaluation, and rapid prototyping on real-world building data.
Role: Research Member · Dates: 2025– · Context: USC, Advisor: Prof. David Gerber · Stack: Python, Azure Blob Storage, IoT
Overview
Building performance and occupancy research often suffers from fragmented data pipelines and non-reproducible evaluation setups. This project constructs a living lab — a continuously instrumented digital twin of Ginsburg Hall at USC — to serve as shared research infrastructure for repeatable, real-world experiments.
The system bridges physical sensing and computational analysis: sensors capture the building’s real-time state, a cloud ingestion layer normalizes and stores the streams, and a 3D geometric model ties sensor readings to spatial context.
Highlights
- Integrated multi-source IoT streams (temperature, occupancy, lighting) with 3D building geometry for spatially-aware analysis
- Implemented cloud-based ingestion pipeline using Azure Blob Storage with real-time API access
- Designed for standardized evaluation — enabling consistent benchmarks across sensors, spaces, and time periods
- Supports rapid prototyping with real-world building data for downstream research tasks