AI & Mission Autonomy
We build the embedded system of mission autonomy layer that turns sensors, optics and unmanned platforms into systems that can operate with — or without — a human in the loop. Our stack runs end to end: perception → sensor fusion → target recognition → planning → motion control → multi-agent coordination, all executing on low-SWaP edge hardware so the platform stays effective in comms-degraded and GNSS-denied environments.
AI auto-formation
AI Target Recognition
AI Perception & Automatic Target Recognition
Deep-learning models for detection, classification and tracking of personnel / vehicles / vessels / UAS across EO/IR, low-light and radar inputs.
Multi-Sensor Fusion
Tightly coupled fusion of EO/IR, LiDAR, radar, IMU, RF with Kalman and factor-graph back-ends, delivering robust state estimation under jamming, smoke, dust, glare and night conditions.
GNSS-Denied Navigation
Visual-inertial odometry, LiDAR/terrain SLAM and map-matching keep positioning accurate to 30 mover over 120 km of autonomous operation when satellite navigation is unavailable or spoofed.
Autonomous Planning & Motion Control
Model-predictive and learning-augmented controllers for UGV cross-country mobility and USV station-keeping in Sea State 6.
Swarm & Multi-Agent Cooperation
Distributed task allocation and consensus algorithms enable UAV–UGV–USV teaming; >50 vehicles/UAV coordinated by a single operator, with graceful degradation when links are lost.
Human–Machine Teaming
Adjustable levels of autonomy (LOA 1–10), intuitive single-operator multi-vehicle control, and human-in-the-loop authorization for any engagement decision。