Wed. Feb 28th, 2024
Using AI and Satellite Images to Address Spatial Apartheid in South Africa

Raesetje Sefala, along with her team at the nonprofit Distributed AI Research Institute (DAIR), is leveraging the power of computer vision tools and satellite images to analyze the impact of racial segregation on housing in South Africa. Growing up in a township with inadequate access to resources, Sefala felt a deep sense of frustration and inequality. Now, at 28 years old, she is determined to make a difference.

To conduct their research, Sefala and her team spent three years collecting millions of satellite images and geospatial data from the government to map out townships across South Africa. This data allowed them to train machine-learning models and build an AI system capable of categorizing areas as wealthy, non-wealthy, non-residential, or vacant land.

Their findings were disheartening but not surprising. They discovered that over 70% of South African land is vacant, and there is a significant disparity in land allocation between townships and suburbs. Armed with this data, Sefala and her team are now sharing it with researchers, public service institutions, and civic organizations. Their goal is to help identify land that can be repurposed for public services and affordable housing.

The impact of their work extends beyond data analysis. By providing crucial information to organizations fighting for justice in urban planning, Sefala hopes to address South Africa’s housing crisis. Cities like Cape Town, known for their racial segregation, have a large percentage of households living in informal settlements due to the lack of affordable housing. However, the government’s myth of a lack of vacant land perpetuates the problem.

By challenging this narrative with concrete data, Sefala’s work has the potential to drive systemic change. It could influence the government to label townships as formal residential neighborhoods, ensuring fair resource allocation and expanding access to essential services like healthcare and education. Additionally, her research has already aided policy think tanks like the Human Sciences Research Council in advising the government on budget allocations for vital programs such as HIV treatment.

Through the combination of AI and satellite images, Sefala is making significant strides in dismantling spatial apartheid in South Africa. Although the road ahead is long, she remains committed to fueling social justice and empowering marginalized communities through data-driven solutions.

Frequently Asked Questions (FAQ): Racial Segregation and Housing in South Africa

Q: What is the focus of Raesetje Sefala’s work?
A: Raesetje Sefala and her team at DAIR are using computer vision tools and satellite images to analyze the impact of racial segregation on housing in South Africa.

Q: How did Sefala’s team collect data for their research?
A: Over a period of three years, Sefala and her team collected millions of satellite images and geospatial data from the government to map out townships across South Africa.

Q: What did Sefala’s team discover from their analysis?
A: They found that over 70% of South African land is vacant, and there is a significant disparity in land allocation between townships and suburbs.

Q: What is the purpose of sharing the data with researchers, public service institutions, and civic organizations?
A: The goal is to help identify land that can be repurposed for public services and affordable housing, with the intention of addressing South Africa’s housing crisis.

Q: What is the impact of Sefala’s work beyond data analysis?
A: By providing crucial information to organizations fighting for justice in urban planning, Sefala hopes to drive systemic change and address the housing crisis in South Africa.

Q: How could Sefala’s research influence the government?
A: The research could influence the government to recognize townships as formal residential neighborhoods, leading to fair resource allocation and better access to essential services like healthcare and education.

Q: How has Sefala’s research aided policy think tanks like the Human Sciences Research Council?
A: Her research has supported policy think tanks in advising the government on budget allocations for important programs such as HIV treatment.

Key Terms/Jargon:
– Computer vision tools: Tools and technologies that enable computers to understand and interpret visual data, such as images and videos.
– Geospatial data: Data that is related to a specific location on Earth’s surface. It includes information about the physical features and characteristics of an area.
– Machine learning models: Algorithms and statistical models that enable computers to learn and make predictions based on data, without being explicitly programmed.
– AI system: An artificial intelligence system, in this context referring to the combination of machine learning models and computer vision tools used by Sefala and her team.

Suggested Related Links:
Government of South Africa
United Nations Sustainable Development Goals
Human Sciences Research Council