SEATTLE, Washington — Tapiwa Chiwewe was born in Zimbabwe. He attended the University of Pretoria, in South Africa, where his love for technology and innovation was deepened. However, while he was developing new thoughts and ideas, the skies above him were being clouded by deadly pollutants. Pollution in the country led Chiwewe to develop a pollution detector.
Pollution in South Africa
South Africa, like almost all African countries, runs on coal. 97% of South Africa’s primary energy is coal-based. A number of things contribute to Africa having dangerously high levels of air pollution in urban areas. This includes the common act of burning rubbish as well as the use of paraffin stoves and fires for cooking.
Chiwewe understood this problem as well as anybody else in the world. And he decided to do something about it.
The Pollution Detector
Chiwewe has developed a pollution detector platform that utilizes A.I technology to forecast air pollution levels. The platform will collect, predict and analyze data about the intensity and location of major air pollutants such as ozone and nitrogen dioxide.
Chiwewe works as a research manager at IBM, where his projects have spanned from data-driven healthcare to this A.I.-led climate technology. He has also done work modeling solar systems and analyzing cancer data. https://researcher.watson.ibm.com/researcher/view.php?person=za-TChiwewe
This project, though, Chiwewe feels can be one of the most impactful. “It’s all about touching people’s lives. You don’t want to just create technology that stays in the lab. Most of us (researchers) want to create technology that people find useful,” Chiwewe says.
How Chiwewe is Utilizing A.I. to Predict Pollution
The platform is exciting not only because of its current capabilities but more so because of how much it can learn and improve by itself. Instead of teaching a computer the theories behind climate, the researchers let the platform teach itself. Artificial intelligence technology makes this possible.
Chiwewe states, “You can do a lot of physics to understand how ozone is found in different places, but what we did is we just collected a lot of data and trained these machines on it and they were able to predict [local ozone levels]without any knowledge of how ozone works in the atmosphere.”
One of the biggest problems facing the platform is a lack of infrastructure for recording weather data in cities across Africa. For reference, Beijing has 35 air quality monitoring stations, where Johannesburg has only eight, and Johannesburg has more infrastructure than most other African cities with pollution problems.
Eventually, Chiwewe wants to update the platform so that it doesn’t require physical air quality monitoring stations. Indeed, he advocates for a “virtual monitoring system based on computer models… a way of sensing using software, rather than sensing using hardware.” An evolving system that doesn’t require the physical infrastructure that most African cities lack, would open the door to helping so many more cities avoid the negative effects of air pollution.
Applications for Government
The platform’s biggest positive impacts would come in the form of helping governments plan for events. With prior knowledge of when the worst days of pollution are going to occur, governments could preemptively close roads or issue public health alerts. In the long term, governments could target the biggest polluters and encourage eco-friendly policies. Moreover, having a great quantity and quality of information helps governments and private businesses best accommodate future health hazards.
Developing cities in Africa face a myriad of problems. Health concerns involving air pollution exacerbate almost all of them, as sickness and poor childhood development could damage the future workforce. And presently, any environmental disasters are likely to hurt developing countries far more than developed ones. Chiwewe’s pollution detector innovation looks to help reduce the risk of environmental catastrophes and better prepare Africa for the future.
– Evan Kuo