Researchers in South Africa champion a device for AI-powered air-quality monitoring
Production of the air-quality monitor, powered by artificial intelligence (AI), is a groundbreaking initiative because it demonstrates the ability of the Global South to provide leadership in pandemic preparedness and response.
The research team, under the auspices of iThemba Labs of South Africa’s National Research Foundation and the University of the Witwatersrand in Johannesburg, as well as the leadership of Bruce Mellado, a full professor at the university, has recently successfully tested the air-quality monitoring system, called AI_r.
Poor air quality is an important risk factor for respiratory and cardiovascular diseases. WHO estimates that around 7 million premature deaths, mainly from non-communicable diseases, are attributed to the joint effects of ambient and household air pollution. Air quality has generally deteriorated in most low- and middle-income countries due partly to large-scale urbanization and economic activities that have largely relied on the inefficient combustion of fossil fuels.
One of the reasons why air quality is not monitored systematically is the cost of devices. The AI_r is a breakthrough in the sense that it is set to reduce manufacturing costs by a factor of 2.5. Essentially, the local production of the technology enhances access and efficiency. Being able to produce the technology locally will not only reduce the costs, but it will enable partners to procure and deploy the technology in large numbers, thus strengthening their early-warning-system technology.
Mellado has noted that the AI_r system was designed to function on batteries, consuming a low-level of power. The system is currently deployed in different parts of Johannesburg, including the university, near a local highway and at schools in vulnerable communities. The European laboratory CERN, where devices have also been deployed, serves as a test bench.
The device measures various indexes, and it provides data in real time. The data is collected and analyzed, using machine-learning technologies to provide forecasts. The system has the potential to predict future air-quality metrics, which inspires hope for Africa’s potential to design its own early-warning systems.