Sun Ho Ro
← All projects
Computer VisionRemote SensingEnvironmental MonitoringGeospatial Analysis

Oil Seepage Detection

AI-enabled computer vision and geospatial analysis workflows for detecting tar patties, oil seep indicators, and erosion patterns near refinery-impacted river environments.

Client
Marathon Petroleum
Location
Lovell, WY
Year
2025
Role
GeoAI Engineer

Context

Environmental monitoring near Lovell, Wyoming required better spatial understanding of tar patties, potential subsurface oil seep locations, and river behavior near affected areas. The project used imagery and geospatial analysis to support contamination monitoring and remediation planning.

Challenge

Contamination-related features can be visually subtle and spatially fragmented. The workflow needed to interpret features across satellite imagery, multispectral drone imagery, river corridors, erosion zones, and site-specific environmental context.

My Role

I led the AI-enabled imagery and geospatial analysis workflow. My work focused on detecting and delineating tar patties and seep indicators, then relating those features to river erosion and channel migration patterns.

Approach

I combined satellite imagery, multispectral drone imagery, computer vision methods, and GIS analysis. Detected features were mapped and interpreted in relation to river movement, erosion conditions, and proximity to potential seep areas.

Output

The project produced mapped detections and geospatial interpretation layers showing tar patties, seep indicators, and river behavior near areas of environmental concern.

Impact

The work improved automated identification and localization of contamination-related features. It supported environmental monitoring, risk assessment, and remediation planning with a clearer spatial understanding of site conditions.

Tech & Techniques

Satellite ImageryMultispectral Drone ImageryComputer VisionGISFeature DelineationErosion Analysis