The Human Sciences Research Council (HSRC), in partnership with the funding agency National Earth Observations and Space Secretariat (NEOSS), hosted a community and stakeholder engagement workshop for the project titled “Flood Risk Monitoring and Early Warning System for Informal Settlements with GeoAI in South Africa” on 23 July 2026 at HSRC offices in Pretoria.
The Co-Principal Investigators (Co-PIs) are Dr Emmanuel Fundisi, Dr Mokhantso Makoae, Dr Tholang Mokhele and Prof Moeketsi Hlalele. The workshop brought together representatives from South African National Space Agency (SANSA), University of Johannesburg (UJ), Council for Scientific and Industrial Research (CSIR), National Department of Human Settlements (DHS), National Disaster Management Centre (NDMC), Gauteng Department of Human Settlements (GDHS), and City of Johannesburg (COJ), creating a collaborative platform for communities, researchers, academia and policymakers to engage with an innovative Geospatial Artificial Intelligence (GeoAI)-powered flood susceptibility model developed for South Africa’s informal settlements.
The workshop forms part of a study that was aimed at developing a GeoAI-driven Flood Risk Monitoring and Early Warning System tailored for South Africa’s informal settlements. Through accessible technology, capacity building and training, the study envisaged empowering communities and local authorities, generating high-resolution flood risk maps and providing near real-time flood alerts at community level. The workshop began with an overview of the project providing participants with background, objectives and motivation behind the development of the GeoAI model. This was followed by a presentation highlighting the current state and challenges associated with informal settlements in South Africa, providing valuable context for the need for innovative and community responsive flood risk solutions.
A key highlight of the workshop was the practical demonstration of the GeoAI model, showcasing its application within selected informal settlements in Mamelodi (Ward 28 and 86). The interactive session enabled the participants to provide feedback on the model’s usability, relevance and potential application within local communities. Participants also contributed insightful ideas and practical recommendations on how the model could be improved to better address local needs, enhance its functionality, and increase its effectiveness at community level.
In addition, the workshop presented a draft policy brief with recommendations for integrating GeoAI into disaster risk reduction and flood management. Some key takeaways from this workshop included participants highlighting the importance of combining technology and local knowledge to produce trusted and usable early warning systems, collaboration between communities, researchers, academia and policymakers to improve flood preparedness and inform future policy, and the project has potential to evolve into a broader community resilience platform, supporting multiple hazards such as crime hotspots, human security, community vulnerability mapping and service delivery planning through integrated geospatial intelligence.
