Agricultural intelligence software
Terra connects cameras, sensors, environmental data and existing farm infrastructure into one intelligence layer for crop health, soil conditions, pest activity, risk detection and farm optimisation.
Any camera. Any sensor. One agricultural intelligence layer.
Terra is designed to integrate with compatible hardware, protocols and data sources. Compatibility is not universal and is not assumed: which of your devices can be connected, and what each of them can contribute, is established during the farm assessment before anything is committed.
Your hardware. Our intelligence.
Terra is software. It is designed to sit above the equipment already installed on the farm and make sense of what it produces. It is not there to replace that equipment, and it does not require a proprietary sensor estate before the first insight arrives.
Terra is software. It connects to the cameras, probes, stations and gateways already installed on the farm, over the protocols they already speak. There is no proprietary Terra sensor to buy, because there is no proprietary Terra sensor.
Instrumentation usually arrives one supplier at a time, and each brings a portal that knows only its own devices. Terra reads them together and holds the result against one model of the farm, which is where correlation becomes possible.
The output is not more telemetry. It is a short list of zones worth walking, each with the evidence that put it there and a statement of how confident the platform is about it.
How Terra works
Six stages, in order. The first belongs to the farm; the rest are what Terra adds on top of it.
Yours
Whatever is already in the field. Terra begins here rather than with a purchase order.
Ingestion
Devices, streams and data sources are connected over the protocols they already speak, with credentials and access scoped per source.
Ingestion
Units, sampling intervals, timestamps and identities are reconciled, and every reading is bound to the zone and crop it belongs to.
Reasoning
Computer vision, sensor fusion, analytics and prediction run over the combined record rather than over one feed at a time.
Model
The farm as a structured, continuously updated model: fields, zones, crops, instruments, conditions, history and open risk.
Output
Ranked and explained. Every output names the signals behind it, states its confidence, and says what it recommends confirming.
Product suite
Each module reads a different part of the farm. They are useful alone and considerably more useful together, because correlation is where most agricultural signals actually live.
Data & sensing
Reads the sensors and gateways a farm already runs, and turns their output into a consistent, comparable record.
Explore Terra Sense
Analysis
Computer vision for crop inspection, running on the cameras and imagery a farm already produces.
Explore Terra Vision
Analysis
Turns scattered pest sightings into population trends, hotspots and early warning across the farm.
Explore Terra Pest
Analysis
Reads soil conditions zone by zone and shows where a field stops behaving like one field.
Explore Terra Soil
Analysis
Brings weather data and on-farm environmental conditions into the same picture as the crop.
Explore Terra Climate
Reasoning
Core
The reasoning layer. Combines every signal on the farm into risks, predictions and recommendations that name their evidence.
Explore Terra Intelligence
Reasoning
A continuously updated digital model of the farm: its fields, zones, crops, instruments, conditions and risks.
Explore Terra Twin
Decision support
Turns risk into an ordered list of what to inspect, irrigate and monitor first.
Explore Terra Optimize
Computer vision
Terra Vision analyses imagery from cameras, drones and handheld capture for the visual evidence an inspector would look for. Every detection carries a confidence value and the frame it came from, and is reported as a suspected finding for confirmation rather than as a diagnosis.
CAM-04 · Zone D · 12:40
3 detections
Suspected findings · confirmation required
Sensor intelligence
Terra normalises the output of compatible probes, stations and gateways, binds each reading to the zone it belongs to, and judges it against that zone’s own history. A drifting probe is identified as an instrument problem rather than reported as a soil condition.
Soil moisture · Zone B · 30 cm
14 days · % volumetric
Pest intelligence
Terra Pest reads pest evidence from imagery and from compatible monitoring hardware, and Terra manufactures no traps. It tracks how counts move across zones and weeks, so pressure is read as a direction of travel rather than as a single day’s number.
Pest pressure by zone
6 weeks · detection level
| Zone | W1 | W2 | W3 | W4 | W5 | W6 |
|---|---|---|---|---|---|---|
| Zone A | Zone A, week 1: none. | Zone A, week 2: none. | Zone A, week 3: trace. | Zone A, week 4: none. | Zone A, week 5: none. | Zone A, week 6: trace. |
| Zone B | Zone B, week 1: trace. | Zone B, week 2: none. | Zone B, week 3: none. | Zone B, week 4: trace. | Zone B, week 5: trace. | Zone B, week 6: none. |
| Zone C | Zone C, week 1: none. | Zone C, week 2: trace. | Zone C, week 3: none. | Zone C, week 4: none. | Zone C, week 5: trace. | Zone C, week 6: trace. |
| Zone D | Zone D, week 1: trace. | Zone D, week 2: trace. | Zone D, week 3: low. | Zone D, week 4: low. | Zone D, week 5: rising. | Zone D, week 6: high. |
| Zone E | Zone E, week 1: none. | Zone E, week 2: trace. | Zone E, week 3: trace. | Zone E, week 4: low. | Zone E, week 5: low. | Zone E, week 6: rising. |
Farm digital twin
Terra Twin represents the operation digitally: its fields, zones, crops, instruments, conditions and open risks, kept current from the connected sources. It is what lets a question be asked of the whole farm and answered in one place.
Model structure
Example Farm
2 fields · 5 zones · 56.7 ha
Field North
2 zones · tomato
Zone A
12.4 ha · 3 cameras · 6 sensors · healthy
Zone B
15.1 ha · 2 cameras · 5 sensors · water stress
Field South
3 zones · pepper, cucumber
Zone C
9.7 ha · 2 cameras · 4 sensors · healthy
Zone D
11.2 ha · 4 cameras · 4 sensors · pest risk
Zone E
8.3 ha · 2 cameras · 3 sensors · monitoring
Farm health
An illustration of the Terra console: an index for the farm, a condition per zone, and the open signals ranked by severity and confidence. The values below are examples, not data from a deployment.
Terra Console · Example Farm
5 zones · 56.7 ha
Zone ATomato · 12.4 ha
91
3 cam · 6 sen
Zone BTomato · 15.1 ha
63
2 cam · 5 sen
Zone CPepper · 9.7 ha
88
2 cam · 4 sen
Zone DPepper · 11.2 ha
34
4 cam · 4 sen
Zone ECucumber · 8.3 ha
61
2 cam · 3 sen
Overall health
78/ 100
Index. Higher is better standing.
Moisture, temperature and EC across instrumented zones
Canopy and colour analysis from zone imagery
Moisture behaviour against zone baselines and demand
Detection counts, trend and spread across zones
Symptom detections and environmental conditions
Observed and forecast conditions for the site
Ranked by severity and confidence
Zone B
Water StressElevated water-stress probability
Confidence: Moderate
Prioritised for irrigation review and a physical check.
Zone D
Pest RiskPest pressure rising, possible spread
Confidence: Moderate
Top of the scouting list. Species to be confirmed in field.
Zone E
Monitoring RequiredSuspected sensor fault
Confidence: High
Flagged as an instrument problem, not a soil condition.
Supported data sources
Terra is designed to integrate with compatible hardware, protocols and data sources. Compatibility is not universal and is not assumed: which of your devices can be connected, and what each of them can contribute, is established during the farm assessment before anything is committed.
Any imagery Terra can reach can be analysed. Framing, resolution and capture frequency set what is detectable, not the badge on the housing.
Terra reads instrumentation a farm already owns. Which measurements exist is a property of that instrumentation, and the platform never reports a channel it was not given.
The transport matters less than the fact that it is open. If a device can publish or be polled through one of these, it can usually be connected.
Records the farm already keeps are part of the picture. History is what allows a reading to be judged against the same ground rather than a general expectation.
Use cases
Farm-scale deployment
What a deployment looks like follows from the ground it has to cover and the instrumentation already on it. These are shapes of deployment, not packages: there are no tiers to choose between.
One farm, a defined set of zones, and the instrumentation already installed on it. Usually the fastest way to establish what the connected hardware can actually support.
Several fields or blocks under one operation, often instrumented unevenly and by different suppliers, reported through one model of the farm.
Multiple sites, possibly across regions and climates, with per-site data footprints and reporting that has to hold at group level as well as in the field.
Dense instrumentation over a smaller area, where conditions change quickly and the value of the platform is in resolution rather than in coverage.
Deployment & pricing
Every agricultural operation has a different data footprint. Terra deployments are scoped according to the farm, existing hardware and required intelligence capabilities.
There is no price list on this site, and no monthly figure, because neither would be true until the farm has been looked at. A firm price is set after the assessment, when the scope is known.
Next step
Every agricultural operation has a different data footprint. Terra deployments are scoped according to the farm, existing hardware and required intelligence capabilities.
The assessment establishes what your existing hardware can support before anything is committed.