Cyber Robotics · Practical tool · English

Walk an inspection mission through interruption and recovery

Define a bounded inspection route, interrupt it deliberately and review the conditions for a controlled restart. This is a planning walkthrough: no robot is connected and no safety result is established.

Free to explore. Worksheet inputs stay in this browser tab; no account or paid API is required.

Conceptual title illustration for Cyber Robotics
Conceptual title image. Any real observation data is identified separately below.
Illustrative route and state model.

The layout, numbers and events are fictional planning inputs. Site hazards, stopping performance, sensing, controls and applicable requirements need assessment on the actual platform and environment.

Define the mission boundary

Illustrative indoor inspection routeA route runs from a supervised start point around two equipment areas to two inspection points and a return point. A marked crossing and a wireless weak spot require review.Equipment AEquipment BSupervised start / returnHuman crossingWeak linkInspection 1Inspection 2

Before a trial, name the permitted route, task, people-exclusion rules, stop conditions and operator authority. Distinguish a planning drawing from a surveyed map. A route that looks clear in a diagram may contain reflective surfaces, moving people, steps or connectivity gaps.

Walk the interruption and recovery states

IDLE

Recovery review conditions

Reconnection alone does not resume inspection. In the review state, all three conditions are required for the simulated reset. This models an explicit decision boundary; it is not a certified control implementation.

No simulation events yet.

Inspect an idealised stopping-distance model

The model adds travel during a delay to constant-deceleration braking: d = v × t + v² / (2a). It assumes level travel, constant speed during the delay and constant positive deceleration afterwards. It excludes slope, traction changes, controller limits, jerk, localisation error and safety margins.

No model calculation yet.

A real acceptance test requires measured traces and a justified limit for the actual operating envelope. Do not use this model output as a safe separation distance.

What evidence should the pilot retain?

For lifecycle context, see the ROS 2 managed-node design. A software lifecycle does not by itself establish robot safety.

Take the next step with context

Keep the question, assumptions and evidence together. These tools help organise a review; an exported record does not approve a design or certify a system.