Vigil Autonomy
Table of Contents

⚠ Disclaimer: This entry may be incomplete, out of date, or inaccurate. It is AI-maintained on a best-effort basis. Do not rely on it as a sole source — verify claims independently using the sources listed below.

Summary

Vigil Autonomy is an Austin, Texas startup, founded August 2025, that builds benchmark and perception infrastructure for counter-UAS (C-UAS) teams — not a detection or interceptor product itself, but the data, hardware, and evaluation layer that other C-UAS companies use to train and validate their perception models before fielding them. Its flagship offering, CommonDefense, is a multi-spectral (RGB/LWIR/MWIR) counter-UAS training and benchmark dataset with RTK-grade ground-truth positioning.

Key Facts

  • HQ: Austin, Texas
  • Founded: August 2025
  • Founder & CEO: Matt Goodman (also styled “Founder & Principal” on his personal site)
  • Type: Perception/data infrastructure and benchmarking provider — supplies tooling to C-UAS integrators rather than selling an end-user detection or interception product
  • Flagship product: CommonDefense — multi-spectral C-UAS training/benchmark dataset (v3.2 as of April 2026)
  • Funding: Not publicly disclosed as of this review; no verified funding announcement found
  • Status: Early-stage, active; partner program opened June 2026

What It Is

Vigil Autonomy positions itself as the “shared perception layer” for the C-UAS industry rather than a competitor to detection hardware or interceptor vendors. The company’s stated thesis is that adversary drones cost roughly $500 while the kill chains built to stop them cost millions, and that the bottleneck limiting cheaper, faster C-UAS response is perception infrastructure — training data, evaluation, and model baselines — rather than hardware.

Its product line has three parts:

  • Field Data (CommonDefense): Real-world drone capture data with calibrated timing, RTK position, and scenario metadata across RGB, LWIR (long-wave infrared), and MWIR (mid-wave infrared) imagery, intended for training and evaluating detection/tracking/segmentation models.
  • Latency Benchmarks: Measurement of detection timing and continuity against deployment constraints, not just raw detection accuracy — the company argues “toy demos” of detection quality alone are not a reliable indicator of real-world performance.
  • Model Baselines: Reference computer-vision models and failure reports showing performance against the CommonDefense corpus, plus fine-tuning services using edge-collected data specific to a customer’s deployment region.

Vigil also offers its own field collection hardware — a Jetson-based “Ground Station” platform (camera, RTK positioning, and collection rig) — for customers who want synchronized data collection before their own production hardware is finalized. Benchmarks are designed to run offline on a customer’s own Linux-based compute/perception stack, including NVIDIA Jetson edge hardware.

Notable Developments

  • 2026-06: Opened its partner program — data licensing, model evaluation, and custom collection programs for prime contractors, system integrators, government agencies, and research institutions.
  • 2026-04: Released CommonDefense v3.2, expanding multi-spectral (RGB/LWIR/MWIR) annotation coverage with RTK-grade positioning and full camera intrinsics per annotation.
  • 2026-01: Ground Station (Jetson-based field collection platform) entered active field collection in the Texas Hill Country.
  • 2025-08: Vigil Autonomy founded in Austin, TX by Matt Goodman.

Key People

Matt Goodman — Founder & CEO

  • LinkedIn: linkedin.com/in/matt-goodman-89b76989
  • (https://x.com/goodmattg) · GitHub · Personal site — handle goodmattg consistent across all three; GitHub profile explicitly cross-links the X account and personal site, confirming same person.
  • Previously: Staff Software Engineer, Setpoint (2023– ); Head of Engineering, Cercle AI (2022–2023); Computer Vision Research Engineer, ByteDance AI Lab Computer Vision Group (2021–2023)
  • Education: M.S. Computer Science (computer vision), University of Pennsylvania (2019–2021), including research in the UPenn GRASP Lab Computer Vision Group; B.S. with honors in Electrical Engineering, University of Pennsylvania
  • Per Vigil Autonomy’s own About page, Goodman founded the company “after consulting on computer vision for C-UAS companies based in Austin, TX” — no specific prior C-UAS employer named.

People — Last Reviewed: 2026-07-03

Claim Verification

Claim: 94.2% mAP@0.5 model baseline performance

Status: Unverified. Supporting sources: Stated on Vigil Autonomy’s own homepage as a headline metric; no test set, methodology, or third-party benchmark disclosed. Summary: Self-reported figure with no independent corroboration or documented evaluation protocol.

Claim: 500K+ annotations in field dataset; 12+ latency/timing metrics tracked

Status: Unverified. Supporting sources: Company homepage only. Summary: No independent source confirms dataset size or benchmark metric count.

Claim: “$500 average adversary drone cost" / "$3M+ average interceptor cost” / “400% increase in UAS incidents, 2019–2024” / “1B+ consumer drones produced annually”

Status: Unverified as company-specific research; directionally consistent with cost-asymmetry figures cited elsewhere in this knowledge base (e.g., Allen Control Systems’ ~$10-per-kill claim), but Vigil does not cite primary sources for these industry-wide statistics on its site. Summary: Treat as marketing context rather than sourced statistics pending an independently cited methodology.

Sources

← prev next →