Skip to content
FOR SAFETY AND CONTROLS ENGINEERS

Deterministic.
Formally verified.
Standards-mapped.

This page is the engineering contract. No marketing, no hand-waving.

ARCHITECTURE

Five layers. One trust boundary.

Every command flows top-to-bottom: agent intent, trust scoring, deterministic verification, graph state, capability-gated dispatch. The capability boundary between layers 3 and 5 is the only path to hardware. Agents cannot bypass it.

Each layer can be operated in isolation for unit tests. The kernel is the only layer with a formal correctness argument.

command flow · top → bottomtrust boundary
L5
LLM Agents
per-robot intent · typed input
L4
Trust Engine
score → fast / slow path routing
L3
Deterministic Kernel 3-replica · 2-of-3 quorum
invariant enforcer · two-tier verifier (microseconds, then formal solver bounded at 200 ms) · graph-pattern checks for joint, velocity, force, and zone safety
L2
Knowledge Graph + Protected State
typed graph of cell state · SHA-256 hash-chained audit log, replicated 3×
capability boundary · signed, short-lived approval (500 ms)
L1
HAL · Robot & Simulator
only path to physical actuation
VERIFICATION PIPELINE

Two paths. One contract.

fast path · rules on KGhot
38μstypical · <2 ms worst case

Pre-compiled deterministic rules evaluated against the live knowledge graph. No solver, no allocations on the hot path.

routed when agent confidence ≥ 0.85 and trust ≥ 0.50

slow path · Z3 SMTbounded
200mshard timeout · reject on expiry

Z3 SMT encoding of joint and velocity constraints, run against an isolated copy of the live graph. A solver timeout rejects the command and steps the cell down to a safe stop. Never fails open.

routed when low confidence or trust < 0.50

STANDARDS MAPPING

Architectural support, not certification.

SafeKernels is not certified against these standards today. The table describes the architectural controls that map to each standard's requirements, and the path to certification.

StandardSafeKernels controlStatus
ISO 10218-1:2025Capability boundary, audit trail, software safety argumentsArchitectural support
ISO 10218-2Cell-level coordination, mutual exclusion patternsArchitectural support
ISO/TS 15066Force-and-contact graph patterns align with power-and-force-limitingArchitectural support
IEC 615082-of-3 voting, deterministic kernel, hash-chained audit (SIL 2/3 techniques)Architectural support
ISO 13849-1Performance level Pld/Ple control structure mappingArchitectural support
IEC 62443OT cybersecurity controlsRoadmap
FAILURE MODEL

Each failure has an explicit degradation path.

  1. 01

    Replica crash

    Counted as a rejection. Quorum still achievable with two healthy replicas.

  2. 02

    Network partition

    Same as a crash. Reject-by-default until the partition heals.

  3. 03

    Z3 timeout

    Reject the command and step the cell down to a safe stop. Never fail open.

  4. 04

    Sensor tick lag

    Verify against the last-known KG. Threshold-driven degradation if sustained.

  5. 05

    WAL hash break

    Detected on the next audit-chain check. The cell drops to a hard E-stop and won't accept a software reset.

  6. 06

    LLM hallucination

    Command rejected, trust decrements. After three rejections in ten seconds, the agent loses the fast path; every subsequent command goes through the formal solver.

THREAT MODEL

Adversaries in scope.

AdversaryDefense
Compromised agentKernel-mediated verification. Agent cannot call HAL directly. Capability boundary.
Compromised vendor controllerShort approval lifetimes, hardware-layer validation, audit-log reconciliation of dispatched vs. received commands.
Network MITM (agent ↔ kernel)Signed approvals. mTLS · Roadmap
Rogue operatorRBAC on policy reload and operator-reset. Audit trail in the WAL.
Physical tamper with WAL filesHash-chain detection drops the cell to a hard E-stop.

Out of scope today: firmware-level compromise of the host running the kernel, side-channel attacks against Z3, hardware supply-chain compromise.

Stress-test the safety argument.

Bring your hardest scenario. We'll walk through fast-path rules, Z3 encoding, and graph patterns end-to-end. See pricing if you're scoping commercials.