The only autonomous swarm platform that can formally prove its governance architecture is immune to adversarial collapse — while every other system's health dashboard shows green.
Edge-deployed AI for autonomous course-of-action generation. 22-module authority stack at 50 Hz. Self-healing C2 — commander failover in <20ms. TAK-integrated via CoT adapter.
4-gate message validation with intrusion classifier. 5 Byzantine adversary classes defeated. Autonomous jam/spoof queuing. Consensus-gated technology denial.
TDOA/FDOA onboard via cooperative passive bearing intersection. Multi-INT fusion. Pattern-of-life on RF emitters. 67 runtime invariants enforced every tick.
60% packet loss with <7% mission degradation. 6-level survival protocol (SVI-driven). System performs better at 30% packet loss than at 0% — stochastic resonance.
No GCS required. Byzantine-tolerant consensus election. Role-fluid doctrine with 6 tactical roles and 4 playbooks. Single-operator control of multi-platform formations.
LVC-ready digital twin with PX4 SITL + Gazebo + UE5 4K rendering. Kill-chain tracking. 171-event forensic log for AAR analytics.
Every module has a single, unambiguous authority tier. Higher always overrides lower. 67 invariants validate this ordering at 50 Hz. This total ordering is why SOVEREIGN EDGE is provably immune to false viability.
Standard spectral analysis tells you a governance system is stable. Pseudospectral analysis reveals the truth: under adversarial perturbation, the real robustness margin can be a fraction of what eigenvalues predict.
of governance operators have fragility >2× the spectral prediction
of the MRGC governance margin is illusory — invisible to health dashboards
constructive proof precision — exact to machine epsilon
Every other vendor claims resilience via health telemetry — heartbeat, link quality, battery state. These metrics can all show green while governance fidelity collapses to zero. The failure is structural, not observational.
Any distributed governance system whose module authority is not a total order admits adversarial liveness failures invisible to standard monitoring. This is a constructive theorem — we build the explicit perturbation that breaks it.
SOVEREIGN EDGE enforces a strict total authority ordering (L1 through L5.5). No two modules at any tier can issue conflicting commands. 67 invariants validate this every tick. Provably immune to false viability.
This is the SOVEREIGN EDGE gateway running a doomsday scenario — 20 agents, compound faults, commander kills, cascade containment — through the real WHACO governance stack with PSC certification. Not a recording. Not simulated state.
Real-time swarm governance • 20 agents • Doomsday scenario • PSC drift monitoring • systemd service
SOVEREIGN EDGE is a governance brain, not an airframe. It integrates with any UAS that speaks MAVLink — from a $200 Crazyflie to a Group 3 tactical platform.
pip install blackridge-whaco[edge]
Python 3.10+, numpy + scipy. No GPU, no cloud, no license server. Air-gappable.
sovereign-edge gateway --web
Full governance engine + CoT adapter + web dashboard in one command.
sovereign-edge certify -o cert.json
PSC pseudospectral certification. Signed JSON certificate. CI gate exit codes.
sovereign-edge demo --scenario doomsday
Run the full doomsday scenario through real governance modules.
Every governance state change is a CoT event. Per-agent tracks with governor mode, hierarchy role, ASDP level, threat classification. UDP multicast on standard TAK SA channel (239.2.3.1:6969) or TCP to TAK Server with TLS. ATAK plugin with side-panel swarm dashboard.
SOVEREIGN EDGE addresses 6 of 7 AT27 focus areas. TRL 5, IRL 6. 28,000+ simulation runs. Zero governance failures. Platform-agnostic. Air-gappable. Runs on a Raspberry Pi.