RGB SENSOR
ROBOT 01Virtual camera · rendered scene
COOPERATIVE UNDER-CANOPY PHENOTYPING
Small robots. A wider view of your crop.
A coordinated fleet turns individual plant observations into a shared picture of a field trial.
Preparing the fieldGeometry, robots, and sensors
Drag to orbit · Scroll to zoom30 × 50 cm robots Simulated field
1 m
DEPTH SENSOR
NEAR FARGeometry depth · illustrative sensor
LAST OBSERVATION
AWAITING SCAN- Stem diameter
- — mm
- Ear height
- — m
- Quality score
- —
Start a mission to collect synthetic plant observations.
PLANTS OBSERVED0 / 256
PLANT COVERAGE0%
MISSION TIME00:00
FLEET DISTANCE0.0 m
Ready for a coordinated survey
FROM ROWS TO EVIDENCE
Every observation has a place.
Inspect the collected data, compare plots, and export the current mission.
Stem diameter by plot
SYNTHETIC OBSERVATIONSWHAT THE TEAM CONTRIBUTES
Repeatable observations.
Shared responsibility.
Robots divide the field into reachable tasks. When access changes, they preserve completed observations and allocate the remaining work.
01 Measure under the canopy02 Coordinate coverage03 Flag incomplete evidence
A quality score expresses this model’s uncertainty. It is not a calibrated confidence or validated sensor accuracy.
| Plant / plot | Robot | Stem diameter | Ear height | Quality | Mission time |
|---|---|---|---|---|---|
| Your first observations will appear here after the mission starts. | |||||
No observations yetLatest 12 observations · Full dataset available in CSV
A REAL AGRICULTURAL RESEARCH PROBLEM
Inspired by field evidence.
“A 2025 study reported robotic data collection from nearly 200,000 maize trial plots across 142 locations over five years.”
Summary of DeBruin et al., Communications Biology 8, 467 (2025).
The study collected multisensor data from 198,249 experimental units during 2019–2023. Derived plant traits were reported for 132,176 units. Those are different measures of coverage, and neither validates the performance of this simulation.
Read the peer-reviewed study ↗OUR CONCEPT PLATFORM30 × 50 cm
Width × length, including the wheels
The University of Illinois describes an early TerraSentia platform as 13 inches wide and 20 inches long—approximately 33 × 51 cm. Our original robot design uses a similar compact scale.
University of Illinois platform description ↗Adaptive AgroTech’s independent demonstration is not affiliated with or endorsed by EarthSense, TerraSentia, or the cited institutions.
What this model represents
Row navigation, task allocation, geometric sensing views, simulated measurement noise, and coverage accounting.
What requires field validation
Vision accuracy, traction, crop contact, energy use, autonomy reliability, agronomic outcomes, and total operating cost.