Wearable sensors for personal temperature exposure assessments: A comparative study

The impacts of heat on human health has sparked research on different approaches to measure, map, and predict heat exposure at more accurate and precise spatiotemporal scales. Personal heat sensor studies rely on small sensors that can continuously measure ambient temperatures as individuals move through time and space. The comparison between different types of sensors and sensor placements have yet to be fully researched. The objective of this study is to assess the validity of personal ambient temperature sensors. To accomplish this objective, we evaluated the performance of multiple low-cost wearable sensors (HOBOs, iButton Thermochrons, iButton Hygrochrons, and Kestrel DROP D3FW Fire) for measuring ambient temperature in a (1) field exposure study by varying the placement on human subjects and in a (2) field calibration study by co-locating sensors with fixed site weather station monitors. A secondary aim involved investigating consensus between validation metrics that can be used in future sensor comparison studies. Bland-Altman analysis, correlation coefficients, and index of agreement statistics were used to quantify the difference between sensor and weather station ambient temperature measurements. Results demonstrated significant differences in measured temperatures for sensors based on sensor type and placement on participants. Future research should account for the differences in personal ambient temperature readings based on sensor type and placement. •Differences in temperatures occur based on sensor placement on the participant's body.•Differences in temperature occur between different types of sensors.•The Kestrel is the most correlated to the weather station out of all sensors.•Devices attached to the weather station facing towards solar radiation have higher rates of error.