Managing Informal Settlement Encroachment: AI-Driven Approaches to Water Infrastructure Resilience in Makause City of Ekurhuleni
Rapid urban sprawl in South Africa has accelerated the growth of informal settlements, increasing risks to critical water infrastructure. In Ekurhuleni, the Makause informal settlement has expanded into buffer zones around pipelines and reservoirs, highlighting the urgent need for integrated approaches to infrastructure protection and urban governance. Although research on service delivery and resilience has increased, existing studies remain fragmented across disciplines and do not provide municipalities with holistic frameworks for action. This article reviews literature published between 2010 and 2019, including municipal planning documents, Rand Water reports, and engineering analyses, supplemented by field data from 105 resident questionnaires, 25 professional surveys, and direct observations. Using a mixed-methods approach, the study examines governance, water infrastructure, encroachment, and the role of artificial intelligence in predicting risks. The findings identify three critical challenges: (1) weak governance and enforcement continue to undermine municipal resilience; (2) predictive tools, particularly artificial neural networks (ANNs) with feature selection techniques, can effectively forecast encroachment risks; and (3) sustainable solutions require strong community participation alongside technical interventions. The study proposes a strategic framework that combines foresight analysis with community engagement, enabling municipalities to anticipate risks while fostering local ownership of infrastructure protection. For policymakers and utilities, the framework offers practical guidance to strengthen planning, improve resilience, and safeguard critical water systems in rapidly urbanising environments. This synthesis advances academic debates and provides actionable insights to support more coordinated, proactive, and sustainable responses to urban encroachment challenges.
