High-risk work environments expose personnel to dangerous accidents under rapidly changing conditions.Existing worker localization methods using RF identification, Wireless Fidelity (Wi-Fi), Bluetooth, or ultra-wideband (UWB) present tracking errors and instabilities caused by uneven anchor distribution or multipath interference.Additionally, current systems rarely integrate robust localization with automated multi-modal alerting.To address these limitations, we developed a cloud-based occupational safety system combining a custom safety helmet, UWB transceivers, and real-time IoT protocols.For precise tracking, an integrated data fusion framework blends enhanced trilateration, enhanced nonlinear least squares, and multi-circle overlap methods.A field test involving 120 participants across 150 independent trials was conducted in a 450 m 2 concrete atrium with obstructions.The results showed that the multi-algorithm fusion model achieved submeter localization accuracy, filtering out spatial jumps.The results contribute to the development of cellular UWB anchor grid scaling and modular sensor bus expansion.By demonstrating how raw micro-sensor data streams and dynamic geofencing can be integrated onto a unified edge-to-cloud layout, a wearable safety gear can be transformed from passive positioning trackers into active, intelligent hazardmitigation instruments suitable for complex industrial sites.
Ho et al. (Mon,) studied this question.