AAAR 36th Annual Conference October 16 - October 20, 2017 Raleigh Convention Center Raleigh, North Carolina, USA
Abstract View
Estimating Personal Exposure to Particulate Matter Using a Low-Cost Wireless Sensor Network and Indoor Positioning System
CHRISTOPHER ZUIDEMA, Thomas Peters, Geb Thomas, Kirsten Koehler, Johns Hopkins School of Public Health
Abstract Number: 526 Working Group: Health Related Aerosols
Abstract Personal sampling is the gold standard for occupational exposure assessment. However, personal sampling is expensive, burdensome to employees, and typically suffers from a low number of samples. In most cases, one to six samples are used to judge if employees are overexposed or if facilities are in compliance with occupational exposure limits. To overcome these shortfalls of personal sampling, we have devised a technique to estimate personal exposure by combining two sources of data: geospatial hazard data from a wireless sensor network (WSN) and employee location information from an indoor positioning system (IPS). To evaluate our approach, we have deployed a 50-node WSN covering 806,400 ft2 of a +2million ft2 manufacturing facility. The WSN nodes are distributed in in a spatially-optimized pattern designed to capture hazard variability. In this heavy industrial manufacturing facility workers cut, grind, machine, blast with abrasives and weld. Consequently, each WSN node is constructed with low-cost sensors for carbon monoxide, ozone, noise and particulate matter (PM), to measure hazards inherent in these processes. This presentation is focused on PM, which is measured by the WSN with inexpensive ($10-$15) SHARP dust sensor (GP2Y1010AU0F). The WSN records measurements from each node to an online database every five minutes. Hazard maps can then be generated for time periods of interest. The facility is equipped with a commercially-available IPS originally developed to manage the location and movement of raw materials and equipment. Experiments indicate that the location of a mobile phone connected to the facility’s wireless internet can be tracked with a precision of approximately 30 ft, suitable resolution to position workers and estimate personal exposure with our technique. Estimating personal exposure is less expensive and can provide more exposure data on more employees with higher temporal resolution compared to traditional personal sampling.