| Description |
Millions of people are affected by infectious and chronic diseases every year. Epidemiologists and exposure scientists strive to create more accurate models of such diseases. However, researchers who deploy wireless sensors in large human subject studies for the development of such models face specific challenges because of their human-intensive and costly nature. Furthermore, although significant research has addressed wireless sensor networks (WSNs) and internet-of-things (IoT) systems, the specific needs and challenges for exposure science and environmental monitoring in human subject studies are not well addressed by the current state-of-the-art. In particular, many wireless sensor systems and IoT systems sold commercially claim to improve health and wellbeing. However, there is often little or no scientific evidence to validate these claims. There is a strong need for systems and architectures that can help evaluate WSN/IoT so-called "smart" systems with scientific methods. In this dissertation, we explore and address the challenges of wireless sensor networks for large-scale human subject studies and environmental exposure science. The design of the current WSN systems for human subject studies is evaluated and the new design and architecture of novel wireless sensing systems are introduced to enable exposure scientists and epidemiologists to perform research experiments that require: less than 0.1% data loss over both wireless and wired data connections, extremely low deployment time (3 person-seconds per sensor deployed for large-scale contact sensing studies and 30 min per home for in-home exposure studies), and zero on-site maintenance during a deployment. Next, a new Thing-Enabled Self-Science (TESS) framework is proposed to validate WSN and IoT automation systems via repeated measurement randomized controlled trials. To test the framework of TESS and demonstrate its use, we test TESS on a smart air quality system (SmartAir). Using the furnace fan to pull air through a filter can clean the indoor air. We show that SmartAir maintains clean air similar to leaving the furnace fan on all of the time, but using 58% less energy. In summary, the WRENSys platform, AirFeed, and TESSsystems meet the specific requirements of epidemiological researchers, and advance the state-of-the-art in the use of IoT and WSN systems for the purpose of improving human health. |
