Services

CUAS Training Data

Counter-UAS systems are classifiers. Their field performance depends on whether they've seen the adversary's current platform, in the current environment, at the current range, under the current countermeasures. We build the datasets that close that gap.

Collection Dimensions

• Target class: Group 1, 2, and 3 UAS captured across the classes your program cares about
• Flight profile: trajectory (looming, crossing, receding, diving), velocity, formation
• Environment: 9+ environment types — urban, rural, open field, forested, littoral, and more
• Lighting & obscurants: daylight, dusk, dawn, low-light, haze, precipitation
• Pixels-on-target: across the range of detection distances your deployment expects
• Scene activity: clean background through cluttered, multi-object scenes
• Horizon interaction: sky, mid-horizon, terrain-background target framing

Typical Use Cases

• Develop and benchmark counter-drone detection algorithms
• Train and validate computer vision models
• Support airspace security research and development
• Stress-test existing algorithms and CUAS systems
• Expand training datasets
• Accelerate Counter-UAS product development
• Protect U.S. warfighters

Custom Collection Operations

Your program has a spec. We fly it. Target drone class, flight profile, environment, lighting, specific edge cases — if you can specify it, we can capture it. If you can't specify it yet, we help you figure out what the data plan should look like.

How We Work

• Requirements call with your ML leads and program managers
• Storyboard of every scenario with frame counts and variance targets
• Platform and payload selection per scenario, not one-size-fits-all
• Range coordination and regulatory compliance (FAA Part 107)
• On-site QA with field retakes before leaving the capture site
• One-time delivery or recurring subscription with weekly cadence available

Delivery Timeline

Typical setup averages 3–4 weeks depending on location and mission complexity. Data delivered as frequently as weekly once the mission is established.

Labeling

Labeled data — bounding boxes on each frame identifying target drones — is available on request for specific engagements. We default to delivering tagged, metadata-rich clips and layer on labeling where the program needs it.

Advisory & Integration

Not every engagement starts with "we need 50,000 clips of X." Sometimes it starts with "we're building a new CUAS stack and we have no idea what our data plan looks like." We work those problems too.

Typical Advisory Deliverables

• Data strategy document — what to collect, in what volume, at what fidelity
• Scenario matrix design aligned to your model architecture and evaluation harness
• Licensing and IP structure for program-owned vs. shared data
• Vendor landscape review — when to build, when to buy, when to partner
• Integration plan — how captured data lands in your training pipeline

Retainer Structure

Month-to-month or quarterly retainer with defined hours and deliverables. Not a surprise-fee consulting engagement. We tell you what the scope is, what it costs, and what you're getting.