US EPA meeting to discuss performance targets for non-regulatory air quality sensors

Over 800 people participated in the June 25-26, 2018 US EPA Air Sensors 2018 meeting, in which performance targets for non-regulatory air quality sensors were discussed.

In recent years, the use of low-cost sensors has grown considerably. Yet, the quality associated with these sensors is not fully known, or is highly variable between different makes/models of sensor, and depends greatly on how the sensors are operated.  Would the establishment of performance targets potential improve the quality of low-cost air quality sensors for non-regulatory applications?

The European Union has made great strides recently to evaluate and form a working group to establish performance targets for air quality sensors.

The US EPA meeting, presentations highlighted recent studies that describe the (good) performance that has been found with current particle matter sensors and ozone sensors, which has allowed for them to be used in a variety of studies and use cases.

In my presentation on “apples to apples vs apples to oranges performance testing”, I first discussed the relative merits of controlled laboratory testing of sensors, which would allow for consistent testing conditions, easy third party verification of testing results, and potentially less uncertain, lower cost, and timely results, and “apples to apples” comparisons between sensor makes/models.  Next, I discussed the importance of field testing in real-world applications that present numerous practical challanges for manufacturers, yet provides reassurance for users that sensors would likely work under real-world scenarios. These field tests would acknowleddge that different use cases in different field settings offer a challenges “apples to oranges” variety of conditions.  If sensors are able to perform well under such challenging and varied testing conditions, they’d likely be useful for non-regulatory applications.

On the 3rd day, smaller panel deliberated the relative merits of sensor evaluation, performance targets, binary vs tiered certification, and other issues.  We are working on a document that would provide summarize some of the perspectives we have on the subject.

Comments are closed.