A monitored stop is a controlled halt in a robotic or autonomous system triggered by supervision logic when sensors, spacing, or operating conditions suggest risk. Unlike a sudden emergency cutoff, it aims to stop motion safely, keep the system stable, and preserve enough state for recovery or inspection.
In cyber security, monitored stops matter because attacks on sensors, control channels, or runtime software can push a machine into unsafe behavior. If a camera feed is spoofed, a proximity sensor fails, or a policy begins acting outside expected limits, the control stack may choose to stop before damage occurs. Defenders use monitored stops with runtime checks, override paths, and event logging to catch tampering, anomalies, and unsafe model outputs in physical environments.