Ten capabilities, one platform — from one-click corridor scans to custom-trained detection models.
Continuous land use monitoring powered by live satellite imagery, drone surveys, and AI — so you see encroachment as it happens, not months later.
Draw a route on the map, upload a KML/GeoJSON/GPX/Shapefile, or pick from your saved routes. Set your buffer zones, click scan, and AfriScan fetches satellite imagery, runs AI detection, and delivers a complete encroachment report — hands-off, in minutes.
Five detection engines work together: YOLOv8 deep learning, RAMP segmentation (trained on African buildings), HOT fAIr humanitarian mapping, Google Open Buildings, and OpenStreetMap footprints. Combined ensemble mode deduplicates overlapping detections and catches mud huts, zinc-roof compounds, and informal structures that single-model tools miss. Optional SAM refinement converts boxes into precise footprint polygons.
Define any number of buffer distances (50m, 100m, 200m) around your route, or switch to polygon mode for concessions and exclusion zones. AfriScan automatically classifies every structure by buffer, divides your corridor into 500m risk segments, and ranks each one high / medium / low by building density.
Compare any two imagery datasets — drone vs satellite, wet vs dry season, this month vs last year — with side-by-side or swipe viewers. AI diff mode highlights new, removed, and changed structures automatically. Track encroachment growth over time with hard evidence.
Human-in-the-loop tools for ground-truth confirmation: a click-to-mark interface for inspectors, and a gamified "Spotter" challenge with leaderboards for crowdsourced building marking. Verified marks merge back into your corridor results and feed directly into custom model training.
Fine-tune YOLOv8 on your own labelled imagery to catch the exact structures you care about in your terrain. Auto-chip orthomosaics into training tiles, review and confirm labels in-browser, and export a ready-to-train YOLO dataset. Train locally or ship to Colab/RunPod GPUs, then re-import your custom model.
Persistent dataset management for every image you work with — upload drone GeoTIFFs or fetch satellite tiles from Google, Esri, or Bing at zoom 16–20 (0.15–2.4m/pixel). Every dataset stored with its bounds, CRS, and capture date, ready for scan, compare, and report.
Upload raw drone photos and AfriScan processes them into georeferenced orthomosaics directly in the dashboard via WebODM, then feeds them straight into detection. For high-risk areas satellite flags, deploy Afridrone teams for centimetre-resolution verification.
PDF encroachment reports with executive summary, risk ratings, per-segment tables, and GPS-tagged photo crops of every structure. Per-location House Reports pull all available imagery for a single coordinate. GIS-ready exports: GeoPackage, GeoJSON, Shapefile, and KMZ for Google Earth — every detection with confidence, buffer zone, and distance-to-route.
Prefetch satellite tiles for offline field use, watch live progress of every scan via real-time streaming, and resume jobs after restarts from persistent job history. The full dashboard runs on any browser — desktop, tablet, or phone — so routes, scans, and verification happen wherever the work is.
Draw your pipeline route directly on the satellite map, upload a KML/GeoJSON file, or select from existing routes. Set your buffer zones (50m, 100m) and click scan.
AfriScan automatically fetches high-resolution satellite imagery along your route, runs multiple AI detection models, identifies every building and homestead within your buffer zones, and classifies each pipeline segment by risk level.
Explore results on an interactive map with colour-coded risk segments, toggleable building layers, and satellite imagery. Click any detection to see its distance to the pipeline, GPS coordinates, and satellite photo.
Download a professional PDF report with executive summary, risk analysis, and building photos. Export to Google Earth (KMZ) or GIS formats. Compare with previous scans to track changes over time.
Live satellite and drone intelligence over your infrastructure — see encroachment before it becomes a problem, not after.
Get In Touch