Projects & Collaborations¶
My development program centers on operationalizing non-contact surface-water measurement technologies within the USGS. These projects span image velocimetry, national imagery infrastructure, entropy-based discharge computation, and emerging AI applications.
-
IVy Tools
The first operational image velocimetry software for USGS — 25,000+ lines of Python enabling discharge measurement from video.
-
NIMS
National Imagery Management System — cloud infrastructure serving 1,000+ cameras and 110M+ archived images.
-
CameraDCP
Edge computing platform for autonomous camera monitoring at 80+ USGS streamgages using Raspberry Pi hardware.
-
SurfVelTools
12,000-line Python application operationalizing entropy-based discharge computation from surface velocity observations.
-
Probability Concept
Shannon entropy method for estimating mean velocity from minimal observations — now official USGS guidance.
-
ECHO AI Skunkworks
WMA-authorized rapid-prototyping team evaluating AI-assisted tools for camera imagery and hydrologic workflows.
-
River Ice Detection
Machine learning models (RIce-Net) identifying river ice presence from NIMS/HIVIS imagery with 94% accuracy.
-
ML Image Velocimetry
Deep learning frameworks for fully autonomous velocity estimation — Auto-STIV and AIPIV collaborations.
-
Training & Teaching
National training curriculum for non-contact methods — 170+ staff trained across seven sessions (2022–2026).