Digital Twin based Building Environment Monitoring using BIM, IoT, AI
demonstration
DTB-BMS sensor data
anomaly detection using deep learing model
This R&D project(2021) has the purpose of developing the simple prototype related to Digital Twin based Building Environment Monitoring using BIM, IoT, AI.
This project includes like below.
Digital Twin concpet implimentation
Open source based digital twin development method
Connection dataset between BIM and IoT
Anomaly detection based on environmantal data pattern using deep learing
Visualize monitoring dataaset based on Web
Develop RESTful API
Develop anomaly detection model using deep learning
In reference, this source code shows how to use Autodesk Forge, Node red, Arduino BLE sense, Mongo DB, node.js.
Below is a description of the main directories in this repository:
- DTB-BMS: Contains the core application files for the Digital Twin based Building Management System. It integrates BIM models and real-time data for facility management.
- anomaly_detection: Includes deep learning models and scripts designed to identify irregular patterns or anomalies in environmental sensor data.
- cesium-starterkit: A template and basic configuration for using CesiumJS to visualize 3D geospatial data and BIM models within a web environment.
- cesium-workshop: Practical workshop materials and examples for implementing 3D Digital Twin features using the Cesium library.
- mapbox-app: A web-based application leveraging Mapbox for 2D/3D map visualization and building localization.
- mongodb_IoT: Contains database schemas and integration scripts for persisting time-series IoT sensor data into MongoDB.
- resident_recognition_model_using_openpose: An AI module utilizing OpenPose for human pose estimation to monitor occupancy or analyze resident behavior.
- sensingNano: Firmware and source code for the Arduino Nano 33 BLE Sense, which acts as the primary hardware node for collecting environmental data.
- sensing_nano_nodered: Node-RED flows and configurations for processing, routing, and visualizing data from the sensing nodes in real-time.
Taewook Kang (laputa99999@gmail.com)
- Taewook Kang, Yunjeong Mo, 2024, Comprehensive digital twin framework for building environment monitoring with emphasis on real-time data connectivity and predictability
- Autodesk Digital Twin platform architecture analysis
MIT License