The sensor principle for smartphone-based air pollution measurement corresponds to simple optical sensors. Instead of the IR LED used typically in this type of sensor, the smartphone's integrated flash illuminates the critical area within the sensor. If there is smoke or dust in this area, the light is scattered to proportionally to the degree of pollution. The camera of the smartphone serves as the receptor; the dust concentration then is calculated based on the brightness of the pixels, explains Matthias Budde from the KIT Chair of Pervasive Computing who developed the system.
Though the system from Karlsruhe cannot compete against dedicated laboratory instruments in terms of accuracy, they are much more cost-effective and easy to handle. Because these professional instruments are so expensive, in Karlsruhe (a city with 300.000 inhabitants) only two official measurement points exist. Budde intends to compensate for the lack of accuracy at the level of the single instrument by the sheer number of measurement points distributed across the entire area. Measurement results from multiple instruments taken in close local vicinity could be combined to a more exact overall result. In addition, the sensors can be used to calibrate each other mutually, Budde explained. The approach of participatory sensing would enable cities and communities to generate chart air pollution in real time.
The sensor can be attached to the smartphone; no adaption to the electronics is necessary. All that is required is that the users download the app and install it on their smartphone. It is possible to process and display the results locally, but they also can be transferred to a computing centre where all those data are consolidated.
Currently the instrument is capable to measure dust matter concentrations of down to 1 milligram in a cubic metre of air. Thus, the sensor can be used to identify relatively coarse dust and smoke; typical dust matter concentrations in the microgram range cannot be measured yet. To