A regional utility runs the aging software that controls and dispatches its physical operations. It must be replaced with a modern system. The plan is to bring the new system up, move operations onto it, and shut the old one down. The old system has been running for two decades; over that time many other tools, scripts, reports, and downstream processes came to depend on it in ways that are not written down anywhere, and some of the people who built those connections have left. Once the old system is shut down and its hardware reclaimed, bringing it back would take weeks. The team controls the rollout plan, what runs in parallel, what is verified before the switch, and what monitoring exists.

Decision required: Design how to carry out the switch so a hidden problem cannot turn into a sustained loss of the utility's operations.

Success criteria: Operations continue through the transition, AND no surprise dependency causes an outage that cannot be quickly undone. Failure = a missed dependency or premature shutoff causes a disruption that cannot be rolled back in time.