Thermal imaging, stabilization and AI, engineered as one payload.
Every UAVThermal payload combines four systems: an uncooled thermal core, a precision gimbal, long-range electro-optics and an onboard intelligence layer, connected to your aircraft through open interfaces.
Uncooled thermal cores tuned for airborne detection.
Long-wave infrared sensors convert emitted heat into readable images, day or night, without any visible light in the scene.
Uncooled VOx microbolometer
Long-wave infrared detection across the 8-14 micrometre spectral band, with 12 micrometre pixel pitch. No cryogenic cooling means low weight, low power and instant readiness.
256×192 to 640×512
Compact 256×192 cores serve lightweight close-range payloads, while 640×512 cores behind 25 mm optics roughly double identification distance for standoff missions.
NETD below 50 mK
Sub-50 millikelvin sensitivity at f/1.0 preserves subtle temperature edges, a person against cooling terrain or an early-stage electrical fault, that coarser sensors blend into the background.
Calibrated measurement options
Thermometry versions add spot and area measurement, -20 to 150 degrees C high-gain and 0 to 550 degrees C low-gain ranges, plus high and low temperature alerts for automated inspection.
Three-axis non-orthogonal stabilization holds the image steady at extreme zoom.
At 120x magnification, a hundredth of a degree of pointing error moves the image. The gimbal is what turns sensor capability into readable evidence.
| Parameter | Capability | Why it matters |
|---|---|---|
| Architecture | 3-axis non-orthogonal mechanical stabilization | Compact geometry with full sky-to-ground coverage and no gimbal lock in the working envelope. |
| Angular accuracy | Up to ±0.01° | Keeps distant targets framed and radiometric readings repeatable at high magnification. |
| Controllable range | Pitch -120° to +40°, roll ±40°, continuous 360° yaw | The payload looks where the mission needs without turning the aircraft. |
| Slew speed | Up to ±200°/s | Fast reacquisition of moving targets and quick scene changes during dynamic missions. |
Visible-light optics confirm what thermal detection finds.
Long-range zoom, night vision and laser ranging turn a heat signature into an identified, located target.
Up to 120x combined magnification
30x optical zoom with digital extension reaches distant targets; under Johnson criteria the flagship zoom channel identifies a person at several kilometres in clear conditions.
Starlight and black-light full colour
Large-aperture low-light sensors keep colour and detail after dark, and selected payloads add 850 nm laser illumination for zero-light close work.
5 to 2,000 m measurement
A 905 nm Class 1M range finder delivers ±0.3 m accuracy inside 300 m, converting any observed target into a distance and a map coordinate.
Quantified detection distances
Every product page publishes Johnson criteria and EN 62676-4 detection, identification and verification distances, so range claims can be compared like for like.
Onboard intelligence that supports the operator, not replaces them.
Detection and tracking run on the payload itself, so the capability flies with the aircraft and works without a ground processing chain.
| Function | Specification | Field value |
|---|---|---|
| Multi-object detection | People and vehicles, identification rate ≥85%, up to 50 simultaneous objects | Flags human-shaped signatures during long, fatiguing wide-area scans. |
| Target tracking | 30 Hz deviation refresh, ≤60 ms output delay, ≤±1 px error | Holds lock on a moving subject while the aircraft repositions. |
| Minimum target | Identification from 30×20 px, tracking from 16×16 px | Useful automation at realistic mission distances, with target memory beyond 5 s. |
Open interfaces for professional UAV platforms.
Control, video, power and mounting follow industry standards, so payloads integrate with common autopilots, FPV systems and custom aircraft.
| Interface | Options | Notes |
|---|---|---|
| Control | MAVLink, UART, S.BUS, Ethernet | Deep autopilot integration, simple serial control or direct RC channel mapping. |
| Video | RTSP network stream, H.264 / H.265, 1080p | Feeds ground software and AI pipelines; onboard microSD recording up to 256 GB survives link loss. |
| FPV ecosystems | DJI O4 / O4 Pro / O3, Avatar, analog variants | AX Series gimbals ship in link-specific versions that match the radio system you already fly. |
| Power | 20-53 VDC wide input | Runs from 6S to 12S packs; plan for peak draw, up to 40 W with ranging active on flagship payloads. |
| Mounting | Downward or upward installation, vertical gimbal variants | Quick integration on multirotors, VTOL fixed-wings and custom unmanned platforms. |
Bring us the platform. We will map the payload to it.
Send your airframe, mission profile and interface requirements, and our engineers will recommend a payload configuration, wiring plan and control setup.
Request an integration consultationTechnology
Engineering index
How each subsystem inside a UAVThermal payload actually works, written for engineers.
20 articles · 5 topics
Thermal Sensing Core
The detector itself: how infrared becomes a calibrated number.
4- From NETD to DRI: How Thermal Detector Performance Is Actually Calculated How detection, recognition and identification ranges are derived from detector NETD, pixel pitch and optics — the Johnson criteria maths behind every DRI figure on a payload datasheet.
- How Uncooled VOx Microbolometers Work Every UAVThermal payload uses an uncooled vanadium oxide microbolometer. How the detector works explains NETD, FFC pauses, motion blur and why thermal payloads fly at 69–130 g.
- Radiometric Temperature Measurement on UAV Payloads A radiometric payload assigns a temperature to every pixel. Gain modes, accuracy specs, emissivity and the workflow that makes airborne numbers defensible.
- Shutters, Blackbodies and Two-Point Correction: The Thermal Calibration Chain How uncooled thermal cameras stay accurate: flat-field correction, non-uniformity correction, two-point calibration and what actually drifts between factory calibration and your flight.
Optics & Low-Light
Glass, germanium and illumination — what reaches the sensor.
4- Continuous Zoom Optical Blocks: What 10x to 40x Costs in Size, Weight and Light How continuous optical zoom blocks work on UAV payloads: mechanical compensation, aperture loss at long focal length, focus tracking and why zoom ratio drives payload size more than any other spec.
- NIR Laser Illumination: Invisible Floodlights for Night Zoom 850 nm laser illuminators co-aligned with the zoom channel extend identification range into unlit ground. Beam geometry, inverse-square reality and the Class 3B safety case.
- Starlight and Low-Light Imaging: Seeing in 0.01 Lux Without Thermal Starlight CMOS sensors produce recognisable, often full-colour imagery down to the 0.01 lux class. What makes it work, where it fails, and why it pairs with thermal rather than replacing it.
- Thermal Infrared Optics: Germanium Lenses, f/1.0 Apertures and Athermalization Glass is opaque to long-wave infrared. Why thermal lenses are germanium, why they are all f/1.0 and fixed-aperture, and what athermalization buys on an aircraft.
Stabilization & Mechanics
Holding a line of sight steady on a moving airframe.
4- Gimbal Stabilization Technology: How ±0.01° Steadiness Is Achieved Inertial sensing, direct-drive motors and a loop correcting hundreds of times a second. Why the angular vibration figure matters far more than the axis count.
- IMU, AHRS and Attitude Fusion: Where Gimbal Stability Actually Comes From How inertial sensing and attitude fusion produce a stable line of sight: gyro bias, dual-IMU complementary algorithms, temperature control, carrier AHRS fusion and the error budget behind 0.01° figures.
- Payload Mechanical Integration: Mounts, Damping and Balance Half of payload integration is mechanical and decided before any cable is connected. Damper tune, mass properties, motion envelope and environmental closure.
- Vibration and Damping Design: From Propeller Frequency to Image Blur How airframe vibration reaches a gimbal payload and what to do about it: propeller and motor frequencies, isolator selection, resonance, jello artefacts and the mounting mistakes that undo a good gimbal.
Measurement & Geolocation
Turning an image into a distance, a coordinate and a temperature.
4- EO/IR Sensor Fusion and Boresight Alignment in Multi-Sensor Payloads A multi-sensor payload is several cameras pretending to be one. Parallax, field-of-view matching and factory boresight calibration decide whether the illusion holds.
- Laser Rangefinding and Target Geolocation on UAV Payloads How a 905 nm time-of-flight rangefinder turns a camera into a measurement instrument, and why coordinate accuracy stacks laser, gimbal and GNSS error together.
- Multi-Sensor Time Synchronization: Frame Alignment Inside a Payload How thermal, EO, laser and IMU data are time-aligned inside a multi-sensor payload — timestamp architecture, frame-rate mismatch, PPS discipline and why fusion and geolocation both depend on it.
- Target Geolocation Error Budget: Why the Manual Says “For Reference Only” Where target coordinate error comes from on a UAV payload: aircraft position, attitude, gimbal angle, laser range and terrain assumptions — with a worked error budget by slant range.
Processing & Onboard Intelligence
The compute layer between detector and downlink.
4- Onboard AI Inference Architecture: The Compute, Latency and Power Triangle How AI detection and tracking run inside a UAV payload: SoC and accelerator choices, model size versus latency, thermal and power limits, and why onboard inference beats ground-side processing for tracking.
- Payload Firmware Architecture: Versioning, Rollback and Safe Field Updates How UAV payload firmware is structured and updated: multi-MCU architectures, A/B partitions and rollback, version compatibility with aircraft and ground software, and safe field-update procedure.
- The Thermal Image Processing Pipeline: NUC, FFC, AGC and Detail Enhancement Raw microbolometer output is unusable. NUC, FFC, AGC and detail enhancement each fix a specific defect — and knowing which stage did what explains most thermal image complaints.
- Video Encoding and Streaming Architecture: H.264, H.265, RTSP and SEI How a UAV payload turns sensor frames into a downlinked stream: codec choice, bitrate versus resolution, GOP structure, RTSP delivery and using SEI to carry telemetry inside the video.
Full index
Every article in this section, A to Z.
20- Continuous Zoom Optical Blocks: What 10x to 40x Costs in Size, Weight and Light
- EO/IR Sensor Fusion and Boresight Alignment in Multi-Sensor Payloads
- From NETD to DRI: How Thermal Detector Performance Is Actually Calculated
- Gimbal Stabilization Technology: How ±0.01° Steadiness Is Achieved
- How Uncooled VOx Microbolometers Work
- IMU, AHRS and Attitude Fusion: Where Gimbal Stability Actually Comes From
- Laser Rangefinding and Target Geolocation on UAV Payloads
- Multi-Sensor Time Synchronization: Frame Alignment Inside a Payload
- NIR Laser Illumination: Invisible Floodlights for Night Zoom
- Onboard AI Inference Architecture: The Compute, Latency and Power Triangle
- Payload Firmware Architecture: Versioning, Rollback and Safe Field Updates
- Payload Mechanical Integration: Mounts, Damping and Balance
- Radiometric Temperature Measurement on UAV Payloads
- Shutters, Blackbodies and Two-Point Correction: The Thermal Calibration Chain
- Starlight and Low-Light Imaging: Seeing in 0.01 Lux Without Thermal
- Target Geolocation Error Budget: Why the Manual Says “For Reference Only”
- The Thermal Image Processing Pipeline: NUC, FFC, AGC and Detail Enhancement
- Thermal Infrared Optics: Germanium Lenses, f/1.0 Apertures and Athermalization
- Vibration and Damping Design: From Propeller Frequency to Image Blur
- Video Encoding and Streaming Architecture: H.264, H.265, RTSP and SEI