See stress before it shows. Save the yield it would cost.
Canopiq AI reads your NDVI and thermal flights, checks them against soil-moisture probes and puts a value on every problem. A blocked drip line, a dry pivot span or a pest foothold becomes a field task with a 24-hour target, not a pretty map in a folder.
This is the working MVP on a sample farm. Click around, run an analysis, or open it full screen.
Canopiq AI is a crop-intelligence platform for irrigated farms. Fly your fields or pull satellite passes the way you already do, and Canopiq works out which zones are stressed, why, and what each one is costing you this season.
It reads every capture before it says a word. Broken drip laterals, dry pivot spans, waterlogged low spots, pest footholds and nutrient zones are each pinned to a field, sized in hectares, checked against the soil-moisture probes underneath and priced in tonnes and dollars.
Growers, agronomists and irrigation managers stop comparing NDVI maps by eye. They work from one ranked list of what is wrong, what it will cost if left, and who is fixing it today.
Watch one thermal flight turn into a same-day repair.
The same models as the app, on one synthetic thermal capture of Field K4, 38 ha of drip-irrigated tomato at Kadawa Irrigation Estate. Click a step, or let it play.
Farms collect more data than ever. Most of it ends up as a colourful map nobody acts on.
Every season brings NDVI and thermal flights, satellite passes, probe readings, pump logs and scouting notes. Agronomists still compare them by eye to answer one question: which zones are losing yield right now, and has anyone gone to fix them?
Sample season at Kadawa Irrigation Estate (1,240 ha): 6 fields · 86 soil-moisture probes · 4 probes flagged by the anomaly model
Inspect. Understand. Compare. Prioritize. Act.
One pipeline turns a flight into a ranked list of field tasks, and every step is something you can click through in the live demo.
Inspect
Upload NDVI and thermal orthomosaics from any drone or satellite. Each capture is tagged to a farm and field.
Understand
Vision AI splits the field into zones and proposes findings, each with a location, the hectares affected, evidence and a confidence score. Agronomists confirm or dismiss.
Compare
Findings are matched against the last scouting round and the soil-moisture probes underneath, so new and confirmed problems stand out.
Prioritize
Each finding gets tonnes and dollars at risk this season from your expected yield and crop price, plus a risk score and a P1–P4 priority.
Act
Confirmed findings become field tasks with 24-hour to 14-day targets, exported to your farm ERP or a scouting report.
One flight. Every answer about the field.
Every demo below runs live in your browser on the same sample farm. Try them.
Your NDVI flights become measurable change.
Drag across September's and October's NDVI of Field K4. Canopiq found a blocked drip lateral and a new cluster of pest spots, and confirmed that the nutrient zone from September is still there.
- Every zone of the field measured, not just the obvious red patches
- Patterns separate the causes: strips, spots and soft-edged zones
- Same-field comparisons across the season
Drag the handle, or use the arrow keys.
Catch the dry zone the drone flew past.
An unsupervised model compares every soil-moisture probe with the rest of the field each day, after normalising for irrigation cycles and weather. Probes running dry or sitting saturated are flagged between flights and linked to the imagery when the two agree.
- Robust median/MAD z-score: one odd reading doesn't trigger it
- Flags both directions: under-watered and waterlogged
- Each flag names the probe, how far off it is and since when
Know which problems are worth a crew today.
A pale corner and a dead drip line are both "stress", but they aren't worth the same. Canopiq converts each finding into tonnes and dollars at risk this season, using the hectares affected, your expected yield, a loss share for each cause and your crop price. Then it ranks the whole farm.
- High-value crops rank higher: a hectare of tomato isn't a hectare of wheat
- Water problems get extra urgency, because yield is lost daily
- The score maps to P1–P4 targets your team can override
Ask your farm.
Copilot answers from this farm's findings, probe readings and scouting history, and cites the evidence. When the data can't answer a question, it says so.
- Numbers come from the record, never from the model
- Every finding ID it cites links back to the capture
- Works offline, with a language model as an upgrade
See risk where it is, not in a spreadsheet.
Every field is coloured by its worst open finding and probe anomaly. Click a field to see what is wrong and what it is worth.
- Imagery findings and probe anomalies on one map
- Drill down from estate to field, zone and probe
Five passes behind every capture.
Canopiq is more than an NDVI viewer. It measures every zone, separates the causes, checks the soil underneath and prices the result, using rules an agronomist can read.
Reads the index
NDVI or canopy temperature is recovered from the export's colour ramp, so any standard render works.
Finds the outliers
The field is split into ~1.5 m zones and each one is tested against the whole field with robust statistics.
Names the cause
Shape decides: straight strips are irrigation lines, scattered spots are pests, soft zones are nutrients, hot zones are water, cold zones are waterlogging.
Checks the soil
Probe anomaly detection confirms what the image shows, and finds dry or wet zones the image missed.
Prices and explains
The yield model turns each finding into tonnes and dollars this season, and Copilot answers cite the evidence.
Seven modules. One field record.
Canopiq Vision
Detect irrigation-line failures, water stress, waterlogging, pest footholds and nutrient zones in NDVI and thermal captures.
See it working ›Canopiq Probes
Soil-moisture anomaly detection that flags dry and saturated probes, linked to the imagery.
See it working ›Canopiq Yield
Tonnes and dollars at risk this season for every finding, with risk scores and P1–P4 priorities.
See it working ›Canopiq Copilot
Ask questions of the whole farm and get answers that cite the evidence.
See it working ›Canopiq Field Tasks
Turn confirmed findings into time-boxed tasks for crews and agronomists, exports and scouting reports.
See it working ›Canopiq Schedule
Recommend per-zone irrigation run-time changes from probe trends and canopy temperature.
Learn more ›Canopiq Forecast
Project end-of-season yield per field from vigour trends, so offtake and finance plans aren't a guess.
Learn more ›See all five Pilot modules working
Every Pilot module runs on the sample farm today. Preview modules ship to pilot partners next.
Open the live demo ›Built for everyone who answers for the harvest.
Which fields are losing money right now?
Crop value at risk ranked across every field, with the evidence behind each number.
Where should I walk first this morning?
A prioritised scouting list with the exact zones marked, instead of driving every field.
Is this zone dry because of a broken line?
Probe readings and canopy temperature side by side, with the statistics behind every flag.
Will our outgrowers deliver the tonnage?
Field-level stress and yield at risk across hundreds of contracted farms.
Is this farm being managed well?
A traceable record of problems found and fixed, for credit decisions and claims.
How do we deliver insight, not just maps?
Reviewed findings and white-label scouting reports from the same flight.
Move from colourful maps to field decisions.
Your field data stays yours.
Yield maps and probe data are commercially sensitive. They shape contracts, credit and land value. The platform is built for that.
Start with one farm, for one season.
We are running early pilots with a small group of irrigated estates, agronomy firms and outgrower networks.
Talk
A 30-minute call about your fields, crops, irrigation set-up and how problems get found today.
See it on your farm
A live demo on the sample farm, then on one of your own flights and probe exports.
Pilot
One farm for 90 days of the season, with agreed success measures and weekly check-ins. $6,500, credited in full to your first year.
How does Canopiq tell a broken drip line from a pest problem?
By shape and by source. It splits the field into small zones and tests each one against the rest. A straight strip along the row direction is an irrigation line; several small, sharp spots close together are a pest or disease foothold; a broad, soft-edged zone is nutrition or soil. A hot canopy zone means water stress and a cold one waterlogging. The soil-moisture probes then confirm or question it. Every finding stays a proposal until an agronomist confirms it.
Do we need special drones or cameras?
No. Canopiq works with standard NDVI exports (red-yellow-green renders) and rendered thermal from any multispectral or thermal drone, or from satellite imagery. Radiometric thermal can be added for calibrated canopy temperatures on Enterprise plans.
How are the yield and dollar figures calculated?
For each finding: hectares affected × your expected yield per hectare × a typical loss share for that cause × your crop price. They are planning estimates, clearly labelled, and can be calibrated against your own harvest data.
How accurate is it, and who has the final say?
On our synthetic benchmark the on-device engine found every planted problem with no false positives, including on clean fields. Real-world accuracy depends on flight quality and is measured during each pilot. A person always has the final say: nothing becomes a field task until it's confirmed.
Which probes and systems does it work with?
Any soil-moisture probe that exports readings (CSV today; live connectors on Portfolio plans). Field tasks export to CSV for your farm ERP or WhatsApp-based crew workflows.
What happens to our data?
It stays yours. Each organisation's data is isolated, encrypted, and never used to train shared models without written opt-in. Core image analysis can run entirely on-device.
Know what's stressed, and what it costs.
See Canopiq on the sample farm, then on your own flight.