Multi Sensor Life Detector using Data Fusion Technology
Murugesan Manivel, Aravind Manikanda Raja P, S Parthiban, S. Vignesh, Vijay Balaji S · 2025
Sensors can pick up non-human signals like machine heat or wind vibrations that wastes time and rescue resources while perhaps omitting real victims. This article presents a multisensor human life detection system with the goal of increasing accuracy and reliability using data fusion techniques. Traditional multi-sensor systems are often restricted in dynamic environments, and therefore this design integrates passive infrared, ultrasonic, gas, and vibration sensors to gather complementary information. The system, which is built on an ESP32 platform, facilitates real-time data acquisition and processing. Each sensor is programmed to perform something different: PIR detects motion through fluctuations in infrared radiation; ultrasonic sensor calculates distance for awareness; gas sensor detects human-source gases such as CO₂; vibration sensor captures motions. Weighted average data fusion method combines the inputs, assigning weights based on reliability and usefulness of each sensor in different contexts. Noise removal and normalization are used to keep clean, comparable data for fusion. The system was tested indoors and outdoors with performance measured in terms of accuracy, precision, recall, and false positive rate. The results show a clear gain in detection performance compared to single-sensor systems. Gas sensor calibration and data fusion weights optimization are part of the research. This makes the system suitable for deployment in applications such as search and rescue, home security, and remote health monitoring, where there is a requirement for reliable human presence detection.