OT Resilience

Show disturbances in a controlled way without endangering real infrastructure

The resilience demo changes only simulation and presentation state. It shows how a factory remains controllable when connections are lost, the historian is unavailable or proposals are invalid: degrade, buffer, restore, validate, prove.

NORMAL

All connections healthy, telemetry flows without backlog.

CONNECTION LOST

Broker, historian, OPC-UA connection or edge gateway fail in the simulation.

EDGE DEGRADED MODE

Edge buffers locally, keeps allowed local logic running and closes unnecessary return paths.

BUFFERING

Samples are counted and replayed reproducibly later.

BACKFILL

After reconnection, data is backfilled and validated.

VALIDATION

The system checks gaps, order, timestamps and production impact.

NORMAL

Recovery complete; ideally Data Loss remains 0.

Resilience Lab

Controlled disturbances can be triggered on the start page. They serve as evidence of architecture behaviour: not as a chaos test against real plants, but as a reproducible simulation for discussion, acceptance and training.

MQTT Broker disconnect: shows buffering and recovery without direct controller access
Historian unavailable: shows local continuation and later backfill
OPC-UA connection lost: shows asset status, degraded mode and alerting
Invalid AI proposal: shows policy rejection before any execution
Machine offline: shows production impact and restart metrics