AUTHREX is a safety governance layer for AI-controlled systems, a virtual force field that catches autonomous mistakes before they cause damage. When a self-driving car, a drone, or an aircraft is about to make an unsafe decision, AUTHREX intervenes, pauses the action, degrades authority, and recovers control.
Your home has circuit breakers. When something goes wrong with the electricity, they cut power before the house burns down. AUTHREX does the same thing for autonomous systems.
When an AI system is about to make an unsafe decision, AUTHREX catches it, forces a pause, strips the system's authority to act, and hands control back to a human. The AI keeps the intelligence. Humans keep the authority.
AUTHREX continuously checks whether what the AI is seeing matches reality. If a GPS signal is being jammed, a camera is glare-blinded, or radar data is corrupted, the system knows it cannot trust itself.
Before any irreversible action, a mandatory pause happens. The system reviews the evidence, checks whether humans should weigh in, and only proceeds if the confidence bar is high enough. No split-second catastrophes.
When the system detects it can no longer operate safely, AUTHREX doesn't crash, it degrades gracefully. Full autonomy becomes supervised, supervised becomes hold-position, and humans regain control in a structured way.
Every autonomous disaster in the last 40 years, friendly fire, misidentified airliners, drone strikes on civilians, self-driving crashes, follows the same pattern: a system acted on bad information, too fast, with no authority check.
AUTHREX is the engineering layer that prevents it. Not by making AI smarter. By making sure AI never acts without verified trust, authorized intent, and a recovery path.
Three approaches to autonomous safety have shaped the field. AUTHREX adopts what works in each, and adds the missing layer: the system reasoning, in real time, about whether its own inputs and decisions can be trusted.
Three identical sensors, take the majority. Three identical computers, vote on the answer. Used in commercial aviation since the 1970s.
All three sensors can be wrong the same way. Cosmic rays, GPS spoofing, glare. Identical voters share identical blind spots.
Watch the autonomous system. If it tries to do something unsafe, override with a known-safe controller. Simplex architecture. Used in aerospace.
Binary thinking. Either the safe-controller takes over or it doesn't. No gradient between full autonomy and full intervention.
Different sensor modalities (camera, radar, GPS, INS, celestial). Continuous trust assessment per source. Authority allocated in proportion to trust, with formal recovery when trust collapses.
The system reasons about its own inputs. Authority is graded, not binary. The fallback is structured, not last-resort.
AUTHREX does not replace redundancy or RTA. It composes with both. The novelty is in treating authority itself as a graded, trust-proportional resource governed by a formal lifecycle, rather than a binary on/off held by either the autonomous system or the safety override.
Autonomous systems fail the same way across industries, bad sensor data, rushed decisions, no safe fallback. Here's what AUTHREX does in the seven domains where it matters most, in plain English.
A high-energy particle strikes a memory cell inside the autopilot. A single bit flips. The computer now has corrupted sensor data, but it doesn't know it's corrupted. It commands an uncommanded pitch-down. The aircraft drops 190 feet in 4 seconds. Passengers hospitalized.
Treats corrupted ELAC-1 data as untrusted. The computer knows it cannot trust its own reading.
ELAC-1 says "dive," ELAC-2 says "hold." Byzantine vote rejects the corrupted command before actuators move.
Autopilot drops to supervised mode, crew regains authority in a defined state, no cascading failure.
Same governance layer · Seven domains · Zero retraining of the AI itself
Every DARPA program is evaluated against the Heilmeier Catechism, eight questions developed by former DARPA Director George Heilmeier that cut through jargon and force a researcher to explain the what, the why, and the so-what in plain language. Here are our answers for AUTHREX.
We are building a safety layer for autonomous systems. When an AI-controlled system, a self-driving car, an aircraft, a ship, a power grid controller, is about to do something unsafe, our layer catches the error before damage happens. It pauses the action, strips the AI's authority to act, and hands control back to a human in a controlled way. The AI stays smart. Humans stay in charge.
Today, safety for autonomous systems is handled three ways: (1) testing to try to catch every failure case before deployment (impossible in the real world), (2) watchdogs that shut everything off when something looks wrong (expensive, kills productivity), or (3) rule-based safety rails that only work for scenarios the designers imagined.
None of these handle the real problem: AI systems are asked to act on sensor data that might be wrong, at speeds where humans cannot supervise every decision, in environments where an adversary may be actively lying to the system. The result is that when something goes wrong, there is no graceful path back to safe operation. You get Iran Air 655, Patriot fratricides, Kabul drone strikes, Tesla Autopilot crashes, Colonial Pipeline shutdowns, all different failure modes, same missing layer.
What is new: AUTHREX is the first integrated framework to treat authority itself as an engineered lifecycle, computed in real-time from sensor trust, verified against formal logic, and enforced at the hardware boundary. Instead of building more rules on top of the AI, we built a governance layer that sits between the AI and the actuators. The AI can still think whatever it wants; it just cannot move a motor, fire a missile, or issue a pipeline command without AUTHREX's authorization.
Why it will succeed: Every piece is built on mathematics and formal verification, not heuristics. Sensor trust uses Dempster-Shafer evidence theory. Multi-agent agreement uses Byzantine fault tolerance. Authority state machines are proven correct in TLA+. The approach is domain-independent: the same pipeline works on a drone, a car, a ship, and a power grid. We have 17 documented incidents the framework explicitly addresses.
Defense: The DoD Replicator Initiative and the Collaborative Combat Aircraft program are fielding autonomous systems faster than they can be supervised. AUTHREX is the governance layer that lets commanders delegate more authority because the boundaries are hardware-enforced, not because the AI is trusted implicitly.
Commercial automotive: The 467 crashes and 14 deaths in the NHTSA Tesla investigation are not a Tesla-specific problem; they are a structural problem that will repeat in every ADAS/ADS system until manufacturers add governance. AUTHREX provides that layer.
Critical infrastructure: Colonial Pipeline, Ukraine grid, and dozens of other industrial control system compromises force operators to choose between contaminated operation and full shutdown. AUTHREX provides graceful degradation so you can keep critical functions running while containing the breach.
Technical risk: Moving governance to the hardware boundary requires FPGA or ASIC integration at the actuator level. We have the FPGA governance bitstream designed with BOM-specified components but not yet tested on live silicon. This is where SBIR Phase II funding would validate the design.
Adoption risk: Integrators may resist adding a layer between their AI and their actuators. The counter is that AUTHREX makes AI systems more deployable, not less, because legal and certification risk drops dramatically when the boundaries are hardware-enforced.
Adversarial risk: An adversary who understands AUTHREX may try to manipulate the sensor trust calculus or the authority handoff conditions. We address this through ADARA (adversarial deception detection) but require red-team evaluation, which is part of the research roadmap.
Research phase (current, internally funded): 14 DOI-verified publications, 7 governance frameworks, 4 provisional patents, 14 browser-based simulations, 4 BLADE hardware platform designs (BOM-specified, $12K – $139K per platform).
Phase I (SBIR, ~$300K over 6 months): FPGA bitstream commissioning on Zynq UltraScale+ development board. Hardware-in-the-loop validation of SATA-FLAME pipeline. Red-team evaluation on the Rover testbed.
Phase II (SBIR, ~$2M over 24 months): Full BLADE platform integration, one defense domain (suggested: BLADE-EDGE directed energy or BLADE-AV autonomous ground). Independent verification campaign. TRL 4 → TRL 6.
Q2 2026 (current): Foundation complete. All 7 frameworks published, 4 BLADE platforms designed, 4 patents filed, Rover and UAV testbeds documented.
Q3-Q4 2026: BLADE-EDGE prototype assembly · FPGA governance bitstream RTL commissioning · UAV testbed flight validation · Provisional-to-utility patent conversion for all 4 applications.
Q1-Q2 2027: SATA-FLAME on FPGA (TRL 4-5) · SBIR Phase II submission · BLADE-MARITIME hardware integration · Rover testbed governance validation campaign.
Q3 2027+: TRL 6 target across multi-framework governance on physical hardware · Utility patents granted (projected) · Research partnership or CRADA engagement (planned) · BLADE-AV autonomous vehicle integration testing.
Mid-term exam (Phase I end, ~12 months): SATA-FLAME pipeline running on FPGA hardware. Red-team evaluation under six attack vectors (sensor spoofing, authority hijack, Byzantine node compromise, jamming, credential theft, physical tampering). Proof of hardware-enforced governance that cannot be bypassed in software.
Final exam (Phase II end, ~36 months): Full BLADE platform, one defense and one civilian domain, demonstrated under independent evaluation. Success = the governance layer correctly prevents action in adversarial or low-trust scenarios AND correctly allows action in nominal scenarios, measured against the 17-incident evidence table.
Commercial exam: One OEM adoption in automotive ADAS or maritime USV, with measurable reduction in false-positive disengagement and false-negative incident rate. Independent safety certification (ISO 26262 ASIL-D pathway for automotive, MIL-STD-882E for defense).
Original Heilmeier Catechism: darpa.mil/about/heilmeier-catechism
Autonomous systems are making decisions faster than humans can supervise. The industry is optimizing for intelligence while the governance layer remains absent.
Without structured authority governance, systems operate with unconstrained delegation. No mechanism for degrading authority when trust erodes, no protocol for recovering control when autonomy fails.
AUTHREX addresses this as an engineering problem, not a policy aspiration.
The DoD's Replicator Initiative is scaling autonomous mass across every domain. The Collaborative Combat Aircraft program is fielding AI wingmen alongside human pilots. Both demand governance infrastructure that does not yet exist, the gap between DoDD 3000.09's safety mandates and operational autonomy at scale is widening with every deployment cycle.
AUTHREX SYSTEMS is a research program developing authority governance infrastructure, frameworks, hardware designs, and simulations operating under a single integrated architecture.
A single integrated research architecture combining seven governance frameworks, four hardware platform designs, and fourteen browser-based simulations, providing end-to-end authority lifecycle control for autonomous systems across defense, maritime, infrastructure, autonomous vehicle, and robotics domains.
Between 1983 and 2026, documented incidents involving misidentification, sensor-trust collapse, rushed escalation, and coordination failures have caused hundreds of casualties.
AUTHREX is designed to reduce the probability of exactly these classes of failures.
| SATA | HMAA | ADARA | MAIVA | FLAME | CARA | ERAM | ADV. LEVEL |
● = primary defense ◐ = contributing defense Adversary capability: sophistication level required to execute threat class
Sources: CENTCOM, ICAO, GAO, NTSB, NHTSA, FAA, CISA, CNAS, DoD investigations. All publicly documented. ALIGN = framework alignment to documented failure mode (HIGH = strong match to 3+ frameworks; MED = partial match).
Publicly documented incidents globally across three governance-relevant categories. Counts are lower-bound estimates derived from NHTSA SGO reports, CSIS Significant Cyber Incidents database, ICAO/ASN aviation records, and national investigation releases (GAO, NTSB, DSB, NATO). The upward trend reflects both rising deployment of autonomous and automated systems and improved incident reporting infrastructure after 2021.
Every catastrophic autonomy failure in the record follows one of six recurring patterns. AUTHREX maps each class to a specific combination of frameworks that prevents, contains, or recovers from it.
Weapons, vehicles, or actuators engage targets based on corrupted, spoofed, or incomplete sensor data. Historically the single largest category of fratricide and civilian-harm incidents.
Automated engagement chains compress decision timelines below the threshold at which meaningful human judgment or cross-check is possible. Risk compounds in multi-agent and swarm contexts.
Systems retain full operational authority even as their epistemic foundations collapse, no graceful degradation, no automatic authority reduction under trust decay. The default bias is optimism rather than caution.
Multiple autonomous agents or redundant computers reach incompatible conclusions and act on them. Without fault-tolerant voting, a single compromised node can cascade into systemic failure.
Systems execute irreversible actions before confidence thresholds are met, without a deliberation window, and without a forced pause for evidence review. Particularly acute under hardware-level radiation or jamming.
Adversarial jamming, spoofing, ransomware, or physical environment effects (cosmic particle SEUs, electronic warfare) corrupt the inputs or control infrastructure the autonomous system depends on.
Traditional autonomous systems focus on what the AI can do. AUTHREX adds the missing layer: governance that decides when action is safe, under what authority, based on real-time trust, threat, and context. That shift matters in six concrete ways:
When sensors become untrustworthy, the system reduces authority or blocks action, it does not continue optimistically. The default state is safe / no actuation unless governance explicitly allows it.
Enforcement lives at the hardware boundary, a normally-open relay between AI and actuator. Compromising the AI stack would not automatically let an attacker command the weapon, motor, or controller.
Autonomy is not binary. AUTHREX uses tiers from emergency stop through restricted, standard, and full autonomy, matching how real operations demand different control levels under different threat conditions.
Most systems know how to stop. Few know how to safely restart. AUTHREX treats recovery as its own governance problem, requiring evidence that the threat has cleared before restoring authority, not just a manual reset.
The same governance pipeline applies to defense UCAVs, drive-by-wire vehicles, maritime autonomy, and critical infrastructure controllers, making it governance infrastructure, not a single-domain product.
Operators can authorize more aggressive autonomy precisely because the boundaries are hardware-enforced. Without governance, you throttle the AI out of caution. With it, routine decisions happen at machine speed while risky ones escalate by design, faster operations with bounded failure modes.
AUTHREX is currently a research-stage architecture at TRL 2–4. Simulations validate governance logic. Hardware platforms are BOM-specified but not yet built. Independent peer review and red-team evaluation are planned as part of SBIR Phase I team formation. The idea is important, the architecture is coherent, but it is not yet proven deployment-ready infrastructure.
End-to-end pipeline governing trust, authority, constraints, consensus, deliberation, recovery, and escalation.
Live computational demonstrations of all seven AUTHREX frameworks operating independently, showing the math, the logic, and the real-time behavior of each subsystem.
Continuously tests whether each sensor reading matches what other sensors are saying, what the world should look like, and what known-good baselines report. When a sensor lies, SATA knows.
Computes in real-time how much authority the AI should have given current trust, situation risk, and operator availability. Authority is not a static permission, it is a dynamic variable that rises and falls with conditions.
Detects when inputs have been manipulated, GPS spoofed, cameras blinded, cyber intrusion into sensors, and distinguishes honest confusion from hostile attack. Actively tests hypotheses, not just failure modes.
For drone swarms, redundant computers, and multi-sensor fusion, uses Byzantine fault tolerance (the same math cryptocurrencies use to reach consensus) so one compromised or faulty node cannot corrupt the whole system.
Before any irreversible action (firing a weapon, commanding a grid shutdown, committing to a high-speed maneuver), FLAME imposes a mandatory delay window scaled to the consequence severity. No instant catastrophes.
When trust collapses, CARA walks the system down through defined safe states: full autonomy → supervised → safe loiter → human control → powered down. Never a cliff, always a staircase.
Monitors whether a local incident is likely to cascade (single drone compromise → swarm failure → mission failure → international incident) and auto-escalates to appropriate command echelons before the situation owns the operator.
Below: live interactive demonstrations of each framework's underlying computation. [ SIMULATED SUBSYSTEM COMPUTATION ]
Standalone browser-based simulations demonstrating AUTHREX governance frameworks operating across strategic command, tactical engagement, multi-domain operations, and distributed consensus environments.
[ ALL SIMULATIONS RUN CLIENT-SIDE, ZERO EXTERNAL DEPENDENCIES, SEEDED PRNG ]
Quantitative cross-domain escalation risk framework for AI-enabled C2. Six scenarios, 600 Monte Carlo runs, formal property verification, and Merkle provenance chains.
Full seven-stage governance pipeline operating in a joint all-domain command and control theater environment with Byzantine fault detection and Merkle audit trails.
HMAA/CARA governance across Air, Maritime, and Kinetic domains. Demonstrates authority handoffs and Byzantine fault isolation in a multi-domain operations environment.
CARA fratricide prevention demonstration. HMAA governance intercepts a compromised UCAV targeting a friendly naval asset via common operating picture integration.
Unified JADC2 architecture with distributed Web Worker nodes, tactical COP, ERAM analytics, and Merkle provenance.
Decentralized actor-model testbed. Each node runs in an isolated Web Worker with MAIVA PBFT consensus and emergent CARA interlock.
Zero-dependency kinematic engine. Proportional Navigation guidance, CARA Flight Termination System, HMAA governance. Pure HTML5 Canvas.
AUTHREX is designed to be evaluable against the safety and assurance standards that govern airworthiness, defense system safety, and formal-methods software. The mappings below describe how each AUTHREX framework relates to the relevant clauses of these standards. Mappings are research artifacts; they are not certification claims and do not constitute an audit or DER finding.
FAA-recognized standard for safety-critical airborne software. Defines five Design Assurance Levels (DAL A through E) based on failure-condition severity, from catastrophic to no-effect.
Defines how formal methods may be used to satisfy DO-178C objectives. Recognizes formal verification as an alternative to testing for many objectives at higher DALs.
DoD's framework for managing safety risk across the system lifecycle. Defines hazard severity, probability, and risk-acceptance levels for defense systems including autonomous platforms.
DoD handbook of airworthiness criteria for fixed-wing, rotary-wing, and unmanned aerial systems. Used by USAF, USN, USA, and USMC airworthiness authorities to evaluate fitness for flight.
These mappings are intended to position AUTHREX within the certification landscape and to guide future work toward a fielded artifact. They are not certification claims, and they do not represent findings by an FAA Designated Engineering Representative, a DoD airworthiness authority, a DER, or any system safety review board. Formal evaluation against these standards requires a target platform and a System Safety Program Plan that AUTHREX does not yet have. The TSD risk register, formal TLA+ specifications, and architecture documentation are designed to support such a program when one is initiated.
Six operational scenarios across air, ground, sea, infrastructure, and orbital domains, each showing what happens without governance vs. with AUTHREX authority control.
[ ALL SCENARIOS ARE SIMULATED ENVIRONMENTS, NOT FIELDED SYSTEMS ]
Every component backed by published research, reproducible simulations, and documented engineering specifications.
All metrics are simulated values from browser-based validation environments. Hardware-validated metrics pending BLADE platform assembly.
Each governance framework undergoes a four-stage verification pipeline designed to meet MIL-STD-882E safety-critical requirements, progressing from computational validation through formal mathematical proof to physical hardware execution.
This research program exists because the gap between autonomous capability and authority governance is widening. Current approaches treat control as a policy overlay. AUTHREX treats it as an engineering problem.
The governance architecture provides the operational mechanisms for assigning, monitoring, degrading, revoking, and recovering authority in high-speed autonomous environments.
This is not AI safety in the abstract. This is control engineering research for real systems operating under real constraints.
The same governance pipeline that prevents catastrophic failures in military systems directly addresses high-liability scenarios in commercial autonomous operations.