Autonomous Distributed Multi-Camera System

Occam Vision Group’s Autonomous Optical Sensor (AOS) system is a self-contained, distributed camera network designed to autonomously capture ultra-high-speed missile events in harsh, infrastructure-limited environments. Developed during Phase I, AOS consists of 10 lightweight Pods, each integrating 8 synchronized, high-frame-rate cameras capable of up to 32,000 FPS per Pod and a synthetic system-wide rate of 320,000 FPS. Each Pod includes onboard compute, solid-state storage, wireless communications, and a solar-capable power system, making the platform fully self-contained and deployable in austere field environments.AOS addresses key limitations of current range instrumentation, which relies on manually aligned, fixed-angle cameras tethered to centralized trailers. These legacy systems require significant setup time and fail to adapt when test articles deviate from their nominal flight path. In contrast, AOS leverages reinforcement learning (RL) to autonomously steer camera orientation and dynamically adjust region-of-interest (ROI) selection during flight. This enables real-time adaptation to non-nominal trajectories, tumbling behavior, and unpredictable impact locations—ensuring that critical test data is captured even under off-nominal conditions.Phase I established technical feasibility through a combined hardware/software development effort. Results included a ruggedized Pod with pan-tilt control, onboard AI processing, and real-time synchronization using sub-25 ns clock alignment across multiple cameras and Pods. Calibration, timing, and tracking functions were demonstrated in simulation and lab test environments, using data generated by the Trajectory Analysis and Optimization System (TAOS) from Sandia National Laboratories. These tools enabled scalable training and evaluation of AI models across a wide range of flight scenarios, and support for complex cases such as multi-stage events or intercepts was investigated for future integration.Phase II will complete system integration, validate the platform in live test events at White Sands Missile Range (WSMR) and Yuma Proving Ground (YPG), and deliver a complete 10-Pod system to the Government by Month 18. The final product will include hardware, software, deployment tooling, and documentation suitable for transition to Phase III procurement or follow-on Commercial Solutions Opening (CSO) contracts.Beyond Army test ranges, AOS is also being commercialized as a passive, infrastructure-light airspace awareness system for counter-UAS and drone detection at commercial airports. The first early-adopter deployment is planned for summer 2025, with follow-on beta installations under discussion with five U.S. airports through early 2026.