BRING YOUR OWN RESULTS
Import field data
Load Polygon or MultiPolygon GeoJSON boundaries, or your team's prediction JSON. BARC reference labels stay separate from model predictions. Files stay in your browser.
Up to 2 MB · Download a sample file
A FRESH PERSPECTIVE ON YOUR LAND
Your fields. The bigger picture.
Find a field. Follow the season. Understand the harvest.
Field-year records6 fields
Harvest patterns4 silage-like · 2 grain-like
Worth a closer look2 fields to check
INTELLIGENCE, ROOTED IN YOUR LAND
A new perspective.
A clearer harvest.
Your field has a story. Let AI help you read it.
Explore harvest patterns, seasonal change and the evidence behind every answer.
＋ A VIEW FROM ABOVE Landscape photograph · not live field imagery
Photo: Bernd Dittrich / Unsplash
YOUR AI FIELD ANALYST
Real observations. Trained classification. Nemotron explanations.
PUBLIC GEOSPATIAL DATASTART WITH YOUR LAND
What can we learn?
03 Get an evidence-based answer
AI model not connected yet. You can explore the sample result below.
WHEANUT / FIELD INTELLIGENCEPREVIEW
THE ANSWER IS ONLY THE BEGINNING
Understand the why.
Your analysis will bring together a harvest classification, the evidence behind it and what still needs a closer look.
Silage-like or grain-like harvestSeasonal evidence & uncertaintyClear next steps to verify
Awaiting your team's AI connection
THE LAND IS COMPLEX. YOUR TOOLS SHOULDN'T BE.Photo: pine watt / Unsplash · landscape inspirationA LITTLE MORE CONTEXT. A LOT MORE CLARITY.
Look closer.
See beyond the boundary.
Harvest signals do not exist in isolation. Weather, vegetation and field management all shape what we see from above.
01
Find your place
Select a field on the map or draw the boundary of your own.
02
Let the evidence speak
Bring your team's geospatial analysis into one understandable view.
03
Keep people in the loop
Review uncertainty, add observations and verify before deciding.
Wheanut combines a trained harvest classifier with NVIDIA Nemotron. BARC reports use real historical Sentinel-2 data and pre-generated AI explanations. Photographs are illustrative landscapes. New fields require a live backend. Classification is not harvest-timing advice.
Find your field
6 fieldsReal BARC field boundaries, 2021–2023. Colors show held-out model predictions, not reference labels.
Yolo County, California
Silage-likeGrain-likeAwaiting analysis
Outline your fieldClick corners on the map. Add at least 3 points.
Map tiles are unavailable. Field outlines still work; check your connection to load the basemap.
SMALL STEPS. A CLEARER PICTURE.
From your field to a useful answer.
01
Find your field
Choose a research field, enter coordinates, or draw a boundary.
02
See the seasonal story
Look at how green the crop was before and after harvest.
03
Check the evidence
Review the result and flag anything that needs a second look.
Loading measured evaluation…
KEEP A COPY
Take your work with you
Download field results and your review notes, or share boundaries with your teammate.
Reviews and saved fields are stored on this device.
YOUR FIELD ASSISTANT
Trained locally. Explained by Nemotron.
A trained harvest classifier uses seasonal Sentinel-2 observations. NVIDIA Nemotron explains its results on the team GPU. Reports are pre-generated; new boundaries await a live backend.
44 real field-year reportsABOUT THIS WORKSPACE
Clear about what we know.
44 evaluated field-years from 34 Maryland fields. Predictions are held out by year and field. New fields wait for analysis. Classification does not establish harvest timing.