Randomized trial demonstrates enhancing positioning accuracy in smart buildings using a BIM-driven approach, suggesting significant benefits for digital twin technology.
Key Points
This study aims to enhance indoor localization accuracy in smart buildings by integrating Building Information Modeling with Wi-Fi and BLE data.
Developed a hybrid localization framework combining Wi-Fi and BLE Received Signal Strength data with BIM.
Implemented two positioning models: Fingerprinting using a Multilayer Perceptron and Trilateration based on a log-distance path-loss model.
Conducted experiments on 35 reference points within a BIM model at the University of Tehran.
Fingerprinting achieved high spatial accuracy with RMSE ≈ 0.40 m.
Trilateration exhibited larger positioning deviations with RMSE ≈ 2.38 m.
A hybrid strategy combined both methods, exceeding 95% accuracy within one meter.