Nwosu Obinnaya Chikezie Victor
Nwosu Obinna Chikezie Victor is a multidisciplinary researcher and geospatial technology enthusiast with expertise in geophysics, environmental science, artificial intelligence, and open mapping. Holding an MSc in Environmental Technology from Teesside University (UK) and currently pursuing a PhD in Environmental Sciences at the University of South Africa remotely/virtually, his work bridges the gap between academia and practical applications of geospatial data for sustainable development.
With a strong technical background in machine learning, data analysis, and GIS tools, Victor has contributed to projects leveraging AI for disaster resilience, climate adaptation, and community-driven mapping. His research includes AI-assisted seismic interpretation, flood risk modelling, and optimising construction productivity through geospatial automation—topics he has published on in peer-reviewed journals.
Beyond research, Victor is passionate about education and community empowerment. As an online tutor and former Cisco Networking instructor, he trains students in geosciences, programming, and 4IR technologies. He actively mentors youth in STEM and advocates for inclusive participation in open mapping initiatives. His leadership roles—from coordinating academic elections to organising student congresses—reflect his commitment to collaborative problem-solving.
Victor’s career spans industry and academia, including roles in civil enforcement, renewable energy (solar technology), and peer review for scientific journals like IGI Global. A lifelong learner, he holds certifications in AI, blockchain, and agile methodologies, underscoring his belief in technology’s role in solving Africa’s pressing environmental and infrastructural challenges.
At State of the Map Africa 2025, he aims to share insights on AI-OSM integration for humanitarian response while learning from the continent’s vibrant mapping community.
Intervention
Introduction/Background
The increasing frequency of climate-induced disasters, such as floods and droughts, poses serious challenges to underserved communities across Africa. These challenges are exacerbated by infrastructural gaps, limited access to spatial data, and a lack of locally informed decision-making tools. While traditional mapping often fails to capture the dynamic nature of community vulnerabilities, open mapping—especially via OpenStreetMap (OSM)—offers a participatory and scalable alternative for resilience planning. Despite its potential, there remains a gap in systematically integrating artificial intelligence (AI), machine learning (ML), and community-centric data into open mapping workflows to generate real-time, actionable insights.
Main Aim and Purpose of the Study
This study presents a methodological framework that integrates AI-assisted image classification, open geospatial tools, and community-driven data collection to improve spatial awareness, resource allocation, and emergency preparedness in vulnerable regions. The research specifically evaluates how open mapping, powered by AI and community collaboration, enhances flood response, healthcare delivery, and environmental monitoring in under-resourced areas of Nigeria, South Africa, and Kenya.
Methodology
Our approach combines multiple methodologies:
Data Acquisition: Using drone imagery and Sentinel-2 satellite data over targeted regions, including flood-prone and informal urban settlements.
AI/ML Integration: Applying convolutional neural networks (CNNs) and supervised classification models (e.g., Random Forest, U-Net) to extract features like road networks, buildings, vegetation cover, and flood extent.
Community Validation: Engaging local mappers and community leaders in YouthMappers chapters to validate automated outputs via MapSwipe and Field Papers.
Toolchain: Leveraging QGIS, JOSM, Google Earth Engine, and the OSM iD editor alongside custom Python scripts for model deployment.
Quality Assessment: Performing intrinsic and extrinsic quality assessments using metrics such as positional accuracy, semantic accuracy, and completeness based on ground truth surveys and authoritative data sources.
Key Findings
The integrated workflow improved map coverage in three pilot regions by over 70%, while reducing the time required for feature mapping by 60% compared to manual-only methods. In areas with recurrent flooding, risk zones were identified with 92% classification accuracy, enabling proactive community evacuation planning. Engagement with community mappers increased map validation rates and fostered digital inclusion, especially among women and youth. The framework also demonstrated the capacity of AI-augmented mapping to support real-time decision-making during emergencies, with significant potential for scalability across other African countries.
Scientific Contribution and Practical Benefits
This research contributes to the growing body of knowledge on AI-enhanced participatory mapping by:
Presenting a reproducible, open-source mapping workflow that integrates machine learning with local knowledge.
Demonstrating how AI can complement rather than replace human mappers in producing high-fidelity, inclusive spatial data.
Providing insights into the sociotechnical challenges of implementing such systems in low-resource settings (e.g., power outages, digital literacy gaps).
Highlighting how open mapping can bridge the gap between geospatial technology and humanitarian resilience planning.
All scripts, data, and trained models will be published in a GitHub repository under an open-source license to ensure reproducibility. Data outputs will be uploaded to OpenAerialMap and OSM for community use, while metadata and evaluation protocols will be documented in the Zenodo repository.
Implications for the OSM Community
The results show that AI-assisted mapping, when combined with inclusive community practices, offers an effective pathway for increasing the utility, adoption, and sustainability of OpenStreetMap in Africa. The approach provides a scalable solution to both routine and disaster mapping in settings with limited resources. By fostering knowledge exchange between researchers, local mappers, and humanitarian actors, this study opens new directions for collaborative geospatial innovation and ethical data governance in Africa’s mapping ecosystem.
Keywords: OpenStreetMap, AI-assisted mapping, disaster resilience, participatory mapping, satellite imagery, YouthMappers, QGIS, open geospatial science, humanitarian response, African data innovation