CIT route optimisation begins by distinguishing contractual frequency from actual service demand. Build a baseline for stops, volumes, crew hours, vehicle capacity, emergency calls, failed visits, waiting time, and custody exceptions. Without reliable location demand, an optimisation engine only automates assumptions.
The routing model must respect operational constraints: cash-on-board limits, security rules, time windows, depot capacity, crew policy, vehicle compatibility, mandatory route variation, and customer SLAs. These constraints should be explicit and versioned so dispatch teams can understand why a route was proposed.
Start with a controlled route group rather than the whole fleet. Compare cost per completed stop, utilisation, travel and wait time, emergency activity, SLA performance, and manual overrides. A route that looks shorter but creates more exceptions is not an operational improvement.
The strongest design closes the loop between location demand, smart-safe or device telemetry, the control room, chain of custody, and customer reporting. Targets should be based on the operator's baseline and approved only after the pilot demonstrates stable service and security.