실내 건설 현장에서 작업자의 위치를 파악하는 기술 휴대폰 활용

This study presents a mobile phone-based indoor positioning system aimed at improving safety and resource management in construction sites. The system leverages WiFi and BLK RSSI signals along with barometric sensor data to estimate both horizontal and vertical positions of workers, requiring only low-cost signal emitters and widely available smartphoncs. The proposed system consists of a signal generator, a mobile application for RSSI measurement, and a BIM-interfaced server platform; it applies a regression model for altitude estimation and a prc-traincd deep learning model, fine-tuned with construction site data, for distance estimation. The mean positioning error was 2.96 m in the Open environment and 3.98 m in the Closed environment, showing a difference of approximately 1.02 m. For floor classification, a total of 40 tests conducted in school buildings and apartment complexes achieved over 85% accuracy. This represents an improvement of about 15 percentage points compared with previously reported results of WiFi-only or BLE-only based methods (around 70%). In addition, the 2D positioning error remained below 4 m on average, which corresponds to a 20 30% improvement compared with conventional RSSI-bascd trilatcration methods (average error of 5-6 m). Because of its integration with BIM, the system supports real-time monitoring, safety zone detection, and progress tracking. The proposed solution demonstrates practical applicability in dynamic construction settings, offering a scalable and economical alternative to traditional indoor positioning systems.