Mobile Device Uses Machine Learning To Inexpensively and Accurately Monitor Air Quality

Particulate matter, a blend of liquid and solid particles, is a huge contributor to the worldwide air pollution. Smaller particles are considered to be specifically hazardous. According to the World Health organization (WHO) has affirmed that particles 2.5 μm or smaller can cause cancer. Further, as per the estimates of the organization, every year 7 Million population die prematurely owing to the health harms of air pollution.

Mobile Device Uses Machine Learning To Inexpensively and Accurately Monitor Air Quality

The researchers at UCLA have developed an inexpensive mobile device to estimate the quality of air. It functions by identifying the pollutants and estimating their size & concentration with the use of mobile microscope linked to a machine-learning algorithm and a smartphone automatically examines the pictures of the pollutants. The innovation is aimed to offer more people across the world to precisely detect the hazardous airborne particulate matter.

Aydogan Ozcan said, “Researchers looking out for solutions to the worldwide problem have found that fast, precise, and high-throughput quantification and sizing of particulate matter in the air is vital for monitoring air pollution.”

At present, the air quality test is mostly conducted at the air sampling stations. However, they utilize extremely advanced devices that are costly and cumbersome; they also need trained and skilled personnel. Even the portable particle counters that are commercially available cost less but are not precise and even cannot process huge air volumes rapidly.

c-Air, the UCLA platform, is accurately  similar to the high-end devices and most importantly its fabrication cost is much less. It consists of a holographic microscope, which seems to be the size of a computer chip, and an air sampler. It can examine 6.5 L of air in just 30 seconds and also produces pictures of airborne particles. It is connected to a smartphone wirelessly and functions through a distant computer with the use of machine learning algorithm that determines the particle size from the produced images.

The team suggests that owing to the machine learning ability of c-Air, it can rapidly adjust to determine certain air particles, such as diverse kinds of mold and pollen. This is a major step forward to detect the air pollutants, isn’t it?

Leave a Reply

Your email address will not be published. Required fields are marked *