Vision Engineered for the Mission

Daylight and thermal imaging built as complete sensor-and-optics systems — delivered in the interface your platform actually uses, and feeding AAD's onboard AI directly. No adapters. No added latency.

AAD Cameras — UCMOS and UTIM camera lineup

One Imaging System. Any Interface.

No two platforms integrate the same way. AAD cameras ship in the connector your airframe needs — same imaging quality, the interface that fits.

AAD camera interface options — analog, digital, and MIPI connectors

Analog

Direct video output for legacy FPV links and analog transmission chains where simplicity and low latency win.

Digital

Networked digital video (Ethernet / RTSP) for high-resolution streaming and multi-consumer architectures.

MIPI

MIPI CSI-2 for the lowest-latency, board-level connection straight into the onboard compute module.

Glass That Holds Up Edge to Edge

Detection only works if the image is clean everywhere — not just in the center of the frame. AAD cameras use high-quality lenses engineered for minimal distortion and consistent image quality across the entire field of view.

Minimal Distortion

Low-distortion optics keep straight lines straight and geometry true across the full field of view — essential for accurate detection and targeting.

Edge-to-Edge Sharpness

Consistent resolution from center to corner. A target at the edge of the frame stays as clear and detectable as one in the middle.

Sensor-Matched Optics

Every lens is matched to its sensor to extract maximum usable resolution and light — sensor and optics engineered together as one system.

Daylight or Thermal — Built for Both

Choose the sensor that matches the mission profile. Both feed the same onboard AI pipeline.

Daylight · Visible Spectrum

UCMOS

High-resolution visible imaging for clear-condition detection, identification, and situational awareness.

  • Maximum detail and color information
  • High-resolution daylight target detection
  • Streams directly to EdgeNode AI
Thermal · All Conditions

UTIM

Thermal imaging that sees through darkness, smoke, and adverse weather — when the visible spectrum can't.

  • Detection at night and in low visibility
  • Sees through smoke and obscurants
  • Pairs with thermal-trained AI models

AAD Camera Systems

UTIM-D640Thermal

Uncooled LWIR

UTIM-D640

Gigabit Ethernet · UDP / RTSP

H.264 / H.265 stream over Ethernet. Full HTTP REST API — configure via browser or script. Boot-to-stream in ~8 s.

Resolution
640 × 480
Latency
< 50 ms
Interface
Gigabit Ethernet
Codec
H.264 / H.265
View Product
UTIM-A640Thermal

Uncooled LWIR

UTIM-A640

USB-C (UVC) · CVBS Analog

Same LWIR core — USB-C form factor. UVC plug-and-play or CVBS drop-in to VTX. One output at a time.

Resolution
640 × 480
Latency
< 30 ms
Interface
USB-C (UVC) · CVBS Analog
Codec
UVC · Analog CVBS
View Product
UCMOS-DDaylight

Low-Light CMOS

UCMOS-D

Gigabit Ethernet · UDP / RTSP

High-res daylight camera streaming H.264 / H.265 over RTSP. Integrates with EdgeNode AI and any RTSP-capable GCS.

Resolution
1920 × 1080 (2.4 MP)
Latency
< 50 ms
Interface
Ethernet
Codec
H.264 / H.265
View Product
UCMOS-MIPIDaylight

Low-Light CMOS

UCMOS-MIPI

MIPI CSI-2 · Direct to SoC

Uncompressed frames direct to SoC. Zero compression overhead — built for tight AI pipelines where Ethernet latency is unacceptable.

Resolution
1920 × 1080 (2.4 MP)
Latency
< 10 ms
Interface
MIPI CSI-2
Codec
RAW Uncompressed
View Product

Camera, Compute & AI — Built to Work as One

AAD builds the camera, the EdgeNode AI compute, and the Wield-AI models together. Every frame streams straight into on-board detection — no adapters, no cloud — and because the models are tuned to our own cameras, accuracy is higher straight out of the box. One integrated pipeline, maximum performance.

  1. Camera

    UCMOS / UTIM

  2. EdgeNode AI

    Onboard Compute

  3. Wield-AI

    Detect & ID

  4. Maximum

    Performance

Camera

UCMOS daylight and UTIM thermal cameras, with sensor and optics engineered together and output in the interface your platform uses. Clean, low-latency frames built to be read by a machine — not just a pilot.

EdgeNode AI

The onboard compute module that ingests every frame and runs inference at the edge — no ground link, no cloud. Scalable headroom to run AAD's detection models in real time.

Wield-AI

AAD's detection and identification models, trained in-house and tuned to the exact camera profile — so the system doesn't just see motion, it classifies the target, day or night.

Maximum Performance

Because one team builds all three, there are no integration gaps between sensor, compute, and model — delivering higher detection accuracy and lower latency than any mix-and-match stack.

Cameras can ship with a representative training dataset matched to your targets and conditions — available on request, so your models start closer to mission-ready.