Skip to content
Technology How it works Breeder — Hyperion Burner — Aegis Burner — MetroVolt AI-Native Architecture Magnets Fuel cycle Safety Roadmap
Solutions AI & Data Centers Defense & Government Grid & Baseload Neutron Detection Quantum
Learn Technical Library
Proof Publications Whitepapers Technical Library Open Science & Reproducibility The Honest Gates
Company About / Mission Leadership Environment Health & Safety Investors Careers Press Contact
3D Model
AI Architecture › Real-Time Control & Safety
Real-Time Control & Safety

Heartbeat and Watchdog Supervision

A watchdog turns silence into safety: any controller that stops proving it is alive causes a trip, so a hung process fails safe instead of failing quiet.

STRATEGY / SLOW ▲ ▼ MICROSECOND REAL-TIMEL7Ecosystem & Strategytelemetry ▲ control ▼open ▸L6Experience & Visualizationtelemetry ▲ control ▼open ▸L5Applications & Copilotstelemetry ▲ control ▼open ▸L4Orchestrationtelemetry ▲ control ▼open ▸L3Twin Modeling & AItelemetry ▲ control ▼open ▸L2Data Fabrictelemetry ▲ control ▼open ▸L1Control Planetelemetry ▲ control ▼open ▸L0Foundationtelemetry ▲ control ▼open ▸PHYSICAL S.M.A.R.T. GENERATOR PLANTBREEDER · HYPERION1R0 1.2 m · A 2.5 · 16.84 T · δ −0.30BURNER · TANDEM MIRROR2317 T throat · 26.49 T plug · fₙ 5.44% · DEC1 center stack + plasma · 2 high-field plug · 3 expander → direct converterCOLOR GRAMMAR strategy AI-workflow infra/data models reactor/DECLINE SEMANTICStelemetry (µs)controlKRONOS FUSION ENERGYAI-NATIVE S.M.A.R.T. GENERATORMASTER BLUEPRINTSHEET 01REV. 2026-08L0-L7 · 2 MACHINES
The AI-Native S.M.A.R.T. Generator Master Blueprint — eight layers (L0→L7), one control stack, wired to both machines. Telemetry rises in microseconds; control descends the same path.

Liveness as a safety signal

A controller that crashes and stops sending commands can look identical to a controller that is holding steady. The watchdog removes that ambiguity: every reflex and supervisory node must periodically prove liveness by petting a hardware watchdog. If the heartbeat stops arriving within the window, the watchdog assumes the worst and trips to the safe state.

python
def watchdog(last_pet_ns, now_ns, window_ns):
    # miss the window by any margin -> trip; silence is not safe
    if (now_ns - last_pet_ns) > window_ns:
        return 'TRIP'
    return 'ok'

def pet_deadline(loop_period_ns, misses_allowed=2):
    # allow a small number of missed ticks before declaring dead
    return loop_period_ns * (misses_allowed + 1)

Design rules

Windowed watchdogs also catch the opposite fault: a heartbeat arriving too early can indicate a runaway loop. A good watchdog therefore trips both on too-late and too-soon petting, bounding the loop period from both sides.

Watchdogs are layered like everything else: node-level watchdogs guard individual controllers, and a plant-level watchdog guards the aggregate heartbeat so that a coordinated failure of several nodes is also caught. Each watchdog is independent of the process it supervises and defaults to trip, so the failure of a watchdog itself does not leave a node unguarded. The result is that no controller can silently stop mattering; silence anywhere in the reflex tier resolves to a safe trip.

Watchdog supervision composes with the ML-independent failsafe: loss of heartbeat is one of the conditions that drives a hardwired trip. It also feeds the liveness input consumed by tier separation when the supervisory link goes silent.

Content reviewed August 2026 · design-and-simulation stage