Three domains. One sensor stack.

The physics is the same in all three. A building has a climate, a crop has a water balance, and switchgear has a failure curve. All three leave a thermal signature before they leave a visible one.

Livestock Crops & Greenhouse Farm Infrastructure

Livestock

A herd does not have one temperature.

Weather-station THI gives you one number for the whole barn. Thermal gives you one per animal, and one per square metre of floor.

Illustrative system view of Livestock climate and condition monitoring across a poultry house and dairy barn; not a thermal measurement.
Illustrative — not a measurement. Concept view of: Livestock climate and condition monitoring across a poultry house and dairy barn
2.3–4.1 °C
Comb and wattle temperature rise under acute heat stress in broilers, η² ≥ 0.70
Moura et al. 2025, Animals 15(15):2233
180–360 g
Daily milk yield lost per 1-unit THI increase above threshold, Turkish herds
Turkish academic study, DergiPark
6.43 °C
Maximum internal temperature spread measured in a commercial poultry house
Deng et al. 2024
Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

Heat stress, per animal

Forehead temperature tracks THI at r = 0.808 in dairy cattle. Rectal temperature manages only 0.557.

Surface temperature starts rising at THI 66.8–71.4. Core temperature waits until 74.1.

Peng et al. 2019, Int J Biometeorol 63:327–336. N = 488 Holstein cows.
Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

House climate uniformity

Inlet asymmetry, dead zones, exhaust-end build-up, brooder coverage. One thermal map shows what forty point sensors approximate.

Mature diagnostic practice rather than a single study. Extension programmes have used thermal house scans for insulation and inlet auditing for years.
Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

Cooling efficacy

In a 50,000-hen cage-free house, back-surface temperature dropped 6–6.7 °C immediately after sprinkler activation. It was still 5 °C cooler at three minutes.

University of Georgia field trial, 2 × FLIR T440.
Emerging Real published results from small samples, single sites or lab conditions. Sample size always shown.

Lameness and digital dermatitis

Deep learning on thermal foot images reached over 81% accuracy on the day of first clinical signs. Two days earlier, over 70%.

Small study. One farm. 38 animals. Promising, not proven.

Feighelstein et al. 2024, Scientific Reports 14:29849.

What we will not claim. Thermal is often sold for mastitis and oestrus detection. The published literature on both is split.

Two studies conclude infrared has low diagnostic ability, or fails at early detection. One sheep study found infected animals measurably cooler, not warmer.

We run both as research. We do not sell them. Why

Livestock use cases

  • Illustrative system view of Floor and air-inlet surface temperature along the inlet-to-exhaust axis; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Floor and air-inlet surface temperature along the inlet-to-exhaust axis
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    Poultry house climate uniformity and ventilation audit

    6.43 °C

    maximum internal temperature spread measured in a commercial house

    Deng et al. 2024, Animals 14(18):2638

  • Illustrative system view of Forehead and orbital surface temperature, after acclimatisation, out of direct sun; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Forehead and orbital surface temperature, after acclimatisation, out of direct sun
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    Dairy heat-stress monitoring by surface temperature

    r = 0.808

    forehead temperature against THI. Rectal temperature manages 0.557.

    Peng et al. 2019, Int J Biometeorol 63:327–336

  • Illustrative system view of Comb and wattle surface temperature against the house baseline; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Comb and wattle surface temperature against the house baseline
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    Poultry heat stress by comb and wattle temperature

    2.3–4.1 °C

    ΔT under acute heat stress, η² ≥ 0.70 — the strongest of twelve regions measured

    Moura et al. 2025, Animals 15(15):2233

Crops & Greenhouse

A closed stoma is warmer than an open one.

Canopy temperature is a direct proxy for stomatal conductance. It tells you a plant has stopped transpiring before it tells your eyes anything.

Illustrative system view of Greenhouse and vineyard canopy observation during a valid measurement window; not a thermal measurement.
Illustrative — not a measurement. Concept view of: Greenhouse and vineyard canopy observation during a valid measurement window
74 %
of Turkey's water use goes to agriculture. 94% of irrigated area still uses surface irrigation.
Climate Scorecard / Daily Sabah
0.83
Stem water potential predicted from on-the-go thermal imaging plus machine learning, Rioja vineyards
Gutiérrez et al. 2018, PLOS ONE 13(2):e0192037
60–70 %
of greenhouse production cost is heating
Industry sources
Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

Irrigation scheduling by canopy temperature

The crop water stress index is a formula, not a black box: CWSI = (T_canopy − T_wet) / (T_dry − T_wet).

In super-high-density olive, aerial CWSI separated full irrigation from deficit irrigation replacing 45% of crop evapotranspiration. Yield per unit leaf area was not significantly reduced.

In cotton, maximum water-use efficiency occurs at less water than maximum yield. That gap is the money.

Agricultural Water Management, olive and cotton trials.
Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

Greenhouse climate uniformity and envelope integrity

Torn cladding, failed seals, thermal bridging, uneven heat-pipe distribution. Invisible on a walk-through, obvious in one thermal pass.

We do not claim the camera saves energy. We claim it finds where the energy goes, and verifies a screen delivers what it was rated to deliver.

Emerging Real published results from small samples, single sites or lab conditions. Sample size always shown.

Condensation and Botrytis risk

A transpiring leaf sits 1–2 °C below air temperature at night. It can reach dew point and begin condensing while the climate computer’s air sensor still reads safe.

We produce a canopy-level condensation risk score. There is no published figure for how much disease this prevents, and we will not invent one.

Mechanism per AHDB and Penn State Extension guidance.

Where this fails, stated plainly. A ±1 °C air-temperature error produces a 28–83% error in the calculated CWSI. That is the difference between a decision and a guess.

Canopy thermal is reliable near solar noon under clear sky. Cloud makes the index unstable, and wind biases it.

In one almond-orchard study, UAS-based CWSI could not separate control from deficit-irrigated trees. Stem-water-potential measurements separated them clearly. The instrument failed, not the biology.

Crop use cases

  • Illustrative system view of Canopy temperature against a wet and a dry reference, in the correct measurement window; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Canopy temperature against a wet and a dry reference, in the correct measurement window
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    Irrigation scheduling by canopy temperature (CWSI)

    0.83 R²

    stem water potential predicted from thermal imaging plus machine learning

    Gutiérrez et al. 2018, PLOS ONE 13(2):e0192037

  • Illustrative system view of Envelope surface temperature, heat-pipe distribution and screen performance across the house; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Envelope surface temperature, heat-pipe distribution and screen performance across the house
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    Greenhouse envelope and heating uniformity survey

    60–70%

    of greenhouse production cost is heating

    Industry sources

Start with a measurement, not a subscription.

A fixed-fee assessment. One day on your site. A written report with the thermal evidence, the cost of what we found, and a verdict — deploy or stop.

Book an assessment What’s in the report

Farm Infrastructure & Energy

A fault heats up long before it burns.

Loose connections, overloaded circuits and failing bearings dissipate energy as heat before they smoke. This is the oldest proven thermal application on a farm, and the least monitored.

Illustrative system view of Electrical panel, pump motor and solar infrastructure under normal operating conditions; not a thermal measurement.
Illustrative — not a measurement. Concept view of: Electrical panel, pump motor and solar infrastructure under normal operating conditions
192,862 animals
died across 143 Belgian barn fires, 2006–2011. Animals died in 62 of the 143.
Dutch Association of Insurers / Flemish Parliament records
1M+ animals
died in Dutch barn fires over the same period
Same source set
97 %
hotspot mapping accuracy with UAV thermal on a solar field, validated against ground measurement
Remote Sensing 15(7):1914, 2023
Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

Electrical panels and switchgear

Resistive heating at a loose or corroded connection is a textbook thermal signature. Fixed cameras with automatic alerting are already deployed in adjacent industries.

Requires the equipment under normal operating load. A cold panel shows nothing.

Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

Motors, pumps and bearings

Friction from lubrication loss, misalignment or overload raises housing temperature before failure.

The value is not the single image. It is the trend — the bearing that has crept 2 °C per month across forty assets nobody inspects individually.

Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

PV arrays and agrivoltaics

Cracked cells, delamination, string faults and reverse-bias hotspots against a uniform panel background. A mature inspection discipline, directly applicable to greenhouse roofs and pumping-station arrays.

Emerging Real published results from small samples, single sites or lab conditions. Sample size always shown.

Stored crop and feed surfaces

Microbial self-heating in stored grain, hay or silage. Honest limit: thermal sees surfaces.

In a controlled barley-silo trial, hotspots were detectable only within 0.3 m of the wall or grain surface. Detection failed entirely at wind speeds of 1.0 m/s and above.

Surface scanning supplements probe networks. It does not replace them.

ASABE stored-grain silo hotspot trial, elibrary.asabe.org abstract 22243.

This one has a third-party payer. Of the three domains, infrastructure is the only one where somebody other than the farm may fund it. Insurers price barn fire risk.

Ontario data attributes about 40% of barn fires with a determined cause to electrical deterioration. If you are negotiating a policy, a continuous thermal record is an argument.

Infrastructure use cases

  • Illustrative system view of Surface temperature of connections, busbars and breakers under load; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Surface temperature of connections, busbars and breakers under load
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    Electrical panel and switchgear hotspot monitoring

    40%

    of barn fires with a determined cause are attributed to electrical deterioration

    Ontario fire-cause data

  • Illustrative system view of Cell and string surface temperature against the uniform panel background; not a thermal measurement.
    Illustrative — not a measurement. Concept view of: Cell and string surface temperature against the uniform panel background
    Proven Replicated peer-reviewed research, or mature practice in an adjacent industry. The source line says which.

    PV array hotspot inspection for farm roofs and agrivoltaics

    97%

    hotspot mapping accuracy, UAV thermal, validated against ground measurement

    Remote Sensing 15(7):1914, 2023

Start with a measurement, not a subscription.

A fixed-fee assessment. One day on your site. A written report with the thermal evidence, the cost of what we found, and a verdict — deploy or stop.

Book an assessment What’s in the report