Skip to main content

Use cases

Agricultural problems, addressed with hardware already in the field.

Each of these is a decision a farm is taking today with less information than it would like. Terra raises the resolution of what is known. It does not take the decision.

01 · Use case

Crop Health

Crop stress is visible before it is severe, but only to someone standing in front of it. On a large block, the interval between one walk and the next is the interval in which a problem grows.

How Terra approaches it

Terra Vision analyses imagery from existing cameras and drone flights for colour, texture and canopy change, and compares each zone against its own earlier record. Terra Intelligence reads those changes against soil moisture and weather, so a change is interpreted rather than merely noticed.

What comes out

  • Per-zone health indicators
  • Visual change measured over time
  • Suspected stress raised with evidence
  • Zones ranked for inspection

CAM-04 · Zone D · 12:40

3 detections

Leaf damage 0.71Suspected lesion 0.64Insect 0.58

Suspected findings · confirmation required

Illustrative. Detections carry a confidence value and the frame they came from, so a finding can be examined rather than taken on trust.

02 · Use case

Pest Intelligence

A pest count on its own says almost nothing. What matters is whether pressure is rising, where it is concentrated, and whether it is spreading toward blocks that are still clean.

How Terra approaches it

Terra Pest counts detections from camera imagery and compatible monitoring hardware, tracks them per zone over time, and correlates rising counts in neighbouring zones. Environmental conditions that favour development are read alongside the counts.

What comes out

  • Counts by zone and period
  • Hotspots and direction of spread
  • Outbreak risk raised early
  • A prioritised scouting list

Pest pressure by zone

6 weeks · detection level

Illustrative pest detection level by zone and week, on a scale of none, trace, low, rising and high.
ZoneW1W2W3W4W5W6
Zone AZone 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 BZone 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 CZone 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 DZone 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 EZone 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.
0 · none1 · trace2 · low3 · rising4 · high
Illustrative. Rising counts in adjacent zones are what separate a spread from a set of isolated sightings.

03 · Use case

Disease Risk

Leaf symptoms and favourable weather are each ambiguous alone. Together they are the reason one block needs looking at today and another does not.

How Terra approaches it

Terra Vision surfaces suspected symptoms from imagery; Terra Climate supplies humidity duration, temperature and rainfall; Terra Intelligence combines them into a scored risk that names both contributions.

What comes out

  • Suspected symptoms with confidence
  • Environmental risk in context
  • Elevated-risk zones identified early
  • Findings routed for agronomic confirmation

Site conditions

4 observed · 3 forecast

Daily range, °C

31

33

34

35

36

37

35

Hours above humidity threshold

4

5

7

9

11

12

10

Mon

Tue

Wed

Thu

Fri

Sat

Sun

observedforecast

Illustrative. Forecast accuracy is a property of the forecast provider, not of Terra; the platform states which source a figure came from.

04 · Use case

Soil Monitoring

A field is rarely uniform. Averaged across a block, a dry corner and a saturated one produce a number that describes neither.

How Terra approaches it

Terra Soil normalises readings from connected probes, holds them per zone and depth, and compares zones against each other and against their own history. Anomalies are separated into agronomic change and instrument fault.

What comes out

  • Moisture, temperature and EC by zone
  • Zone-to-zone comparison
  • Trends read against the zone's history
  • Instrument faults caught, not passed on

Moisture at depth

% volumetric

Illustrative soil moisture at three depths in two example zones.
DepthZone AZone B
10 cm
58
34
30 cm
64
41
60 cm
69
58
Illustrative. Zone B is drier through the upper profile while the deeper reading holds, a pattern that a field-level average removes.

05 · Use case

Irrigation Intelligence

Irrigating uniformly wastes water in the zones that do not need it and under-serves the ones that do, and the difference is usually not visible from the pump house.

How Terra approaches it

Soil moisture behaviour, canopy stress signals and environmental demand are read together per zone. Terra Optimize ranks the zones by severity and by how much of the evidence agrees, and shows the readings behind each recommendation.

What comes out

  • Water-stress probability by zone
  • Ranked irrigation priorities
  • Evidence attached to each recommendation
  • Advisory only: the farm decides

Recommended next actions

  1. Irrigation review

    Zone B · priority High

    Moisture below the zone's own baseline for four days, with canopy stress rising in imagery and three days above the usual maximum.

  2. Scouting priority

    Zone D · priority High

    Detection counts up across three capture cycles, with the neighbouring zone showing the same direction of travel.

  3. Instrument check

    Zone E · priority Medium

    Moisture channel static for 41 hours while neighbouring probes responded to the same rainfall. Treated as a data-quality problem, not a soil condition.

Illustrative. Recommendations are advisory: Terra proposes and the farm decides. The platform does not operate irrigation or machinery.

06 · Use case

Farm Monitoring

Coverage is the constraint on any large operation. Adding people scales linearly with area, and adding a proprietary hardware estate is a capital project before it is an improvement.

How Terra approaches it

Terra runs on the cameras, sensors, gateways and weather sources a farm already owns, normalises what they produce, and reports by exception: the platform's job is to be quiet until something warrants attention.

What comes out

  • Continuous coverage without new hardware
  • Exception-based alerting
  • One record across the whole operation
  • Instrument health tracked with the data

Soil moisture · Zone B · 30 cm

14 days · % volumetric

405060below baselinezone baseline rangeday 1day 14
Illustrative. A reading is judged against the zone’s own history rather than a fixed threshold, which is what makes a slow drift legible.

07 · Use case

Yield Intelligence

Yield expectations formed at the start of a season and revised only at harvest leave a long stretch in which planning is guesswork.

How Terra approaches it

Counted fruit and stand data from Terra Vision, growth stage from Terra Twin, and the season's conditions from Terra Climate are held together, so a forecast can be revised against what the farm has actually done and what this season is doing.

What comes out

  • Counted observations by zone
  • Growth tracked against season history
  • Condition-adjusted expectations
  • Forecast inputs, stated as estimates

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

Illustrative. Every reading, detection and risk in the platform is attached to a node of this model, which is what makes the farm queryable rather than merely recorded.

08 · Use case

Precision Agriculture

Most agricultural decisions are still taken at the resolution of the whole field, because that is the resolution the available information supports.

How Terra approaches it

Terra raises the resolution of what is known: readings and detections resolved to zones, evidence retained per zone, and risk scored per zone. Where camera coverage is close enough, individual plants can be counted and tracked.

What comes out

  • Zone-level state and risk
  • Plant-level observation where imagery allows
  • Decisions taken at the right resolution
  • A record that improves with coverage
  • Signals

    • Low soil moisture
    • High temperature
    • Canopy stress in imagery

    Elevated water-stress probability

    Raised as a probability with the contributing readings attached, and routed to irrigation priorities.

  • Signals

    • High humidity duration
    • Leaf symptoms detected
    • Favourable temperature range

    Elevated disease risk

    Reported as a suspected condition requiring agronomic confirmation, never as a diagnosis.

  • Signals

    • Rising pest counts
    • Adjacent zones affected
    • Conditions favour development

    Possible infestation spread

    Direction and pace are estimated from the zones involved, and inspection is prioritised accordingly.

Next step

Request a Farm Assessment

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.