**4. Conclusions**

Low-cost dust sensor modules (n = 3) were constructed using an Arduino Uno R3, a FONA808 module and a Sharp dust sensor. The modules employed the *LogIt* platform our laboratory has previously developed to wirelessly transmit GPS and particle mass concentration data to a web server [40]. The dust sensor was calibrated in the lab against a commercial optical PM mass concentration monitor and subsequently used to monitor the air quality in Lubbock, TX during daily activities, such as walking, cooking, sleeping, sweeping, traveling or working. The concentration of particulate matter was collected in real time with GPS coordinates. The optimal limits of detection for the unmodified Sharp dust sensors tested ranged from 8 μg m<sup>−</sup>3–17 μg m<sup>−</sup>3. The particulate matter concentration remained relatively low, and constant during outside measurements, but data had moderate spikes inside dwellings during certain activities. High concentrations were often observed near restaurants. Multiplexing the sensors (using many simultaneously) can better constrain human exposure to PM, and subsequently better determine health effects of particle pollution.

In the future, our sensors should be calibrated and challenged against additional aerosol compositions and size distributions to improve performance and reliability. Because particle size distribution and composition may affect sensor response we advocate for direct calibration of the Sharp sensors alongside real-time EPA federal equivalent methods such as the TEOM or beta-attenuation monitor using authentic aerosol as the sample. Such direct comparison/calibration on authentic aerosol typical of a specific location will provide best performance. Additionally, the state of the science can be advanced by developing in-field calibration protocols for the low-cost sensors to routinely assess and check sensor calibrations.

**Author Contributions:** H.A. and C.J. contributed by collecting and reporting data and authoring manuscript text. J.E.T. contributed by editing manuscript, providing guidance, and project concept. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors would like to thank Scott Hiemstra for design and construction of the project box used to contain the sensors.

**Conflicts of Interest:** The authors declare no conflict of interest.
