batsigwata/Portfolio

Portfolio of Geospatial and Remote Sensing projects spanning 17+ years. Includes DigitalGlobe award-winning clutter dataset, Botswana rangeland monitoring on vegetation loss and wildfire risk, and deep learning work in Ireland detecting invasive Rhododendron ponticum. Showcases expertise in GIS, remote sensing, and GeoAI. Batsirai Benjamin Gwata – Senior Remote Sensing & GIS Specialist | AI & ML Enthusiast Welcome to my GitHub repository! I am Batsirai Benjamin Gwata, a Senior Remote Sensing & GIS Specialist with over 17 years of international experience. My work spans Remote Sensing, GIS, Environmental Monitoring, Land Use Planning, and Deep Learning for spatial analysis. I am currently pursuing a Doctor of Business Administration (DBA) in Artificial Intelligence & Machine Learning at Walsh College, USA. 🌍 About Me Current Role: Director of Hyperspec Africa Education: BSc (Hons) GIS & Remote Sensing | BSc Computer Science and Environmental & Geographical Science – University of Cape Town Experience Highlights: Conducted Botswana Rangeland Monitoring (2022–2024) using Landsat 9, SAVI/NDVI, rainfall, and fire data, identifying 70% vegetation health decline. Winner of DigitalGlobe Award for High Resolution Clutter Mapping for Wireless Network Propagation (presented at Geospatial World Forum 2011 and USA Defense & Security Conference 2012). Strong experience in translating spatial datasets into policy briefs, thematic maps, and actionable insights for stakeholders. Skills: Remote Sensing & GIS (Landsat, Sentinel, WorldView) Deep Learning & AI (TensorFlow, CNNs, anomaly detection, time-series change detection) Spatial Data Analysis, Feature Extraction, Land Cover Classification Drone & Ground-Truth Data Collection & Analysis Policy-Oriented Geospatial Reporting 💻 Machine Learning & Deep Learning Expertise I specialize in integrating Machine Learning and Deep Learning into geospatial workflows to provide actionable insights. Key areas include: Land Cover Classification: CNN-based classification for up to 60 land cover classes Time-series change detection using satellite imagery Anomaly Detection: Identifying environmental changes, degradation, and land-use irregularities Data Pipelines & Processing: Preprocessing satellite imagery (Landsat, Sentinel, WorldView) Fea …