A Sensor Network for Particulate Distribution Estimation

This paper describes the development of a sensor network designed to estimate the spatial and temporal distribution of particulate in the air. The network employs sensor nodes which are based on an optical solution and are capable of estimating the particulate size distribution. The sensor nodes employ a commercial fiberglass filter through which the air is forced to pass by means of a small pump. A video camera coupled to an inexpensive RaspberryPI Zero W is used to acquire and process the filter image. A small LoRa wireless module is coupled to the the RaspberryPI in order to transmit the acquired data over a range exceeding 10 km. The nodes can measure reliably particles down to sizes of 10 μm, usually refereed to as PM10 and a solution down to 2.5 μm, (PM2.5) is being tested. The fiberglass filter is in form of a strip and a small motor is used to move the strip and to start a new measurement when the filter gets covered in dust. The overall node cost is of less than 100$.