Air pollution, specifically fine particulate matter (PM), represents the single largest global environmental risk factor for premature mortality and morbidity, with outdoor air pollution causing an estimated 4.2M premature deaths annually (GBD, 2017). The effective mitigation of this environmental health risk relies critically on accurate, reliable, and widespread monitoring of key pollutants on a local scale. Distributed air quality (AQ) sensor networks can provide data necessary for:
Cost-effective monitoring solutions are necessary to enable decision-makers to monitor potential air quality issues which can lead to actionable change.
QuantAQ technology being piloted in India
QuantAQ combines more than 10 years of air quality sensor system R&D - designing, building, learning, testing, iterating, re-learning, and rebuilding integrated sensor systems to provide actionable air quality data. See below for some of the contributions our team has made to the peer-reviewed literature on air quality sensors. Most of our work has focused on developing algorithms for reducing uncertainty in sensor measurements and for identifying types and sources of pollution.
For more QuantAQ papers visit: https://www.quant-aq.com/science
Assessing the accuracy of low-cost optical particle sensors using a physics-based approach
Current regulatory approaches to measuring air pollution are sparse & expensive - a typical US EPA air quality monitoring station has a capital equipment cost ~ $150K USD and high annual operating costs. Each monitoring station characterizes the air for that one location, accounting for a relatively small geographic footprint. Additionally, air quality data from the station is often only reported on a daily or hourly basis (not in real-time) rather than on a minute-by-minute basis as the actual pollution levels are changing. In contrast, each of our integrated sensor systems is capable of providing the same information provided by regulatory monitoring stations but for a fraction of the cost. With QuantAQ, cities can deploy and support 50 sensor systems for the same cost as a single regulatory monitoring station.
As atmospheric chemists, we understand that making accurate and reliable air pollution measurements is hard.
Current low-cost solutions produce poor results - over the past several years, many low-cost commercial products have appeared on the market - few capable of providing actionable data. Typical low-cost gas sensors suffer from environmental degradation and interferences. A large part of the work that QuantAQ has done is in developing algorithms and calibration approaches that minimize these effects. Machine learning gets a lot of hype and certainly provides a simplified way at obtaining useful information; however, recognizing and reducing the bias associated with truncated training data is extremely important and it is necessary to inject domain expertise into this process to obtain useful results.
Air quality sensors are inherently imperfect. Taking an honest, transparent approach to data handling is one step towards combating these imperfections.
One of our core competencies at QuantAQ is in developing modeling techniques that allow us to reduce the inherent bias associated with limited/truncated training data sets by leveraging our comprehensive calibration process and environmental calibration chamber. This approach provides more realistic training data sets, allowing us to overcome the limitations and challenges faced when using lower-cost OEM sensors to generate actionable AQ data.
David H Hagan, PhD, CEO / Co-Founder
David is an atmospheric chemist with an interest in blending computer science with environmental chemistry to improve our understanding of air pollution. He holds Bachelor's degrees in Chemical Physics (Hendrix College) and Chemical Engineering (WashU) and a PhD in Atmospheric Chemistry & Physics from MIT.
Eben S Cross, PhD, CTO / Co-Founder
Eben is an atmospheric chemist passionate about building and sharing tools that people can use to improve their air intuition and avoid air pollution. He holds a Bachelor's degree in Environmental Chemistry from Connecticut College and a PhD in Physical Chemistry from Boston College.
For further information:
