Researchers at the Massachusetts Institute of Technology (MIT) have developed an inexpensive air sensor calibration that uses Machine Learning techniques to improve the accuracy of measuring air pollution. With the release of a cheap mobile air pollution detector, the general population can contribute significantly to monitoring air quality.
According to the WHO, air pollution causes over 4 million yearly fatalities. It is a severe issue for public health, yet it isn’t often monitored closely. To enlist more individuals in the cause, the researchers at MIT develop the Flatburn detector.
“We want groups of concerned people or individuals all over the world to be able to evaluate their own levels of air pollution, pinpoint its causes, and work in concert with government agencies and other relevant parties to improve the quality of the air,” said Carlo Ratti, head of MIT’s Senseable City Lab.
The Flatburn gadget is a component of the City Scanner initiative, which use mobile technology to get insight into urban life. According to Fabio Duarte, chief research scientist of Senseable City Lab, the original idea and remains the project’s goal was to democratise environmental data. ” We’d like everyone to be able to evaluate information and interact with citizens and officials,” he underlined.
The open-source detector may be 3D-printed or assembled from low-cost components. Now that it has been thoroughly tried and tested compared to existing state-of-the-art machines, the researchers are making all their findings public, including instructions on constructing the device, running the programme, and deciphering the results.
Simone Mora, a research scientist at Senseable City Lab and co-author of a newly published paper detailing the scanner’s testing process explained that they have been doing several pilots around the world.
“We have since refined a set of prototypes, with hardware, software, and rules, to make sure the data we gain are robust from an ecological science point of view,” he elaborated.
Around 2017, MIT researchers began prototyping a mobile pollution detector, first intended for use on garbage trucks in Cambridge, Massachusetts. The sensors may be charged via USB or solar panel, and they retain data on an internal memory card that can be accessed through the internet.
The latest research phase has been putting the devices through their paces in New York City and Greater Boston, comparing them against established pollution detection infrastructure to determine their efficacy. For four weeks in 2021, researchers in New York employed five detectors to capture 1.6 million data points, which they compared with input from state officials. The group used mobile sensors in Boston, comparing the performance of the Flatburn devices to that of a state-of-the-art system by Tufts University and a state agency.
Both nitrogen dioxide and acceptable particulate matter concentrations were measured over a 10-meter radius using identical detectors. Often connected with combustion processes like those seen in power plants, internal combustion engines in automobiles, and fires, “fine matter” refers to tiny particles.
The research team’s development found that Flatburn devices can produce reliable results. Lower concentrations of fine particulate matter can be detected and estimated with mobile detectors than the devices currently in use. But it only happened with a strong enough correlation, with adjustments for weather conditions and other factors.
However, the study did show that the devices will have a six-month functioning life if used in a mobile situation, such as atop vehicles. However, they also found some problems that users of Flatburn detectors can encounter. Among these are “drift,” the slow but steady change in the detector’s readings over time, and “ageing,” or the more fundamental decline in a unit’s physical state.
Researchers are confident that the devices will perform well, so they’ve released Flatburn as an open-source utility with detailed instructions. It includes advice on analysing the data alongside government representatives and locals to influence policy.