Our data · Ground truth

Ground truth — from disease detection to AI models.

Site visits and ground truthing validate model outputs and ensure on-the-ground realities are captured in the data. This capability bridges the gap between remote sensing and field conditions — and is a key differentiator for clients who require verified agronomic insights, not just remote estimates.

Talk to the team See what's inside
Want to find out more?

Satellites see pixels. Surveyors see crops.

A stressed NDVI signal could be drought, disease, a pest infestation, or simply a late planting. Only someone standing in the field can tell you which. HSAT runs a global network of 2,000+ contributors who visit fields, geo-tag what they see, and feed that evidence back into our NDVI and crop-intelligence models — closing the loop between remote signal and ground reality.

We also deliver AI-powered disease detection and pod counting directly from these field visits — every photograph returns a plain-language disease report, a risk score out of ten, and an automated pod count, with human QA at scale via our offshore and MTurk teams.

AI-processed field imagery of a cocoa pod with disease markers
Types of data available

Field evidence, structured and model-ready.

Global field visits

2,000+ contributors

Field visits from HSAT's global team of 2,000+ contributors — on the ground in the crops and countries that matter most.

Geo-tagged data

Photos, surveys, risk flags

Geo-tagged crop photos, structured surveys and risk assessments — every observation tied to a precise location.

Disease mapping

90%+ detection accuracy

Disease mapping at 90%+ accuracy — including CSSV in cocoa, rust in coffee and cereals, and virus yellows in sugar beet.

Regional reports

National risk maps

Aggregated regional reports and national risk maps — field-level evidence rolled up into portfolio-ready intelligence.

Model integration

Feedback loop into NDVI

Ground-truth observations feed directly back into our NDVI and crop-intelligence models — correcting drift and anchoring forecasts in reality.

AI-processed analysis

Pod counting & yield inference

Pod counting and yield inference from AI-processed field imagery — scaling expert-grade agronomic insight across thousands of visits.

2,000+
Global field contributors
90%+
Disease detection accuracy
Every model
Validated on the ground
Verified, not estimated

Ready to anchor your forecasts in ground truth?

Get field-verified agronomic intelligence — disease maps, yield inference and geo-tagged surveys — feeding directly into the models you already rely on.

Book a meeting