AAAR 36th Annual Conference October 16 - October 20, 2017 Raleigh Convention Center Raleigh, North Carolina, USA
Abstract View
Indicating Particulate Matter (PM) Exposure with a Smartphone App
GANG CHEN, Bruce Urch, Frances Silverman, Arthur W. H. Chan, University of Toronto
Abstract Number: 368 Working Group: Instrumentation and Methods
Abstract Exposure to particulate matter (PM) emitted from solid fuel cook stoves is a significant health risk in developing countries, leading to an estimated 2 million premature deaths every year. One of the main obstacles to adopting clean cook stoves is the poor understanding of the health impacts of PM in these areas. Current commercial PM sensors remain too costly to deploy widely and indicate PM exposure. The objective of this project is therefore to develop an affordable and relatively accurate PM-indicating smartphone app by image analysis. Here we use image analysis to determine the amount of PM collected on a filter substrate, e.g. face masks. An image of the filter substrate is first obtained using a smartphone, and the smartphone app we develop will determine the darkness of the filter substrate using standardized gray ramp scales. To obtain the relationship between the filter darkness and PM mass, we have analyzed over 700 filter samples collected in urban Toronto by the Synchronized Hybrid Ambient Real-time Particulate (SHARP) monitor. Preliminary results show a good correlation (R2>0.70) between PM concentration and the darkness on the filter substrate, demonstrating the feasibility of our method. The algorithms will be further improved using machine learning techniques to obtain better predictabilities. Further tests using a breathing machine and commonly used face masks are under way, and their results will be discussed. This project will potentially lead to a relatively accurate method to measure PM exposure at significantly lower costs compared to commercially available PM sensors, and can be used to inform the general public in developing countries about PM exposures.