Assignment is more than distance
With one mobile robot, task assignment can look simple. A job appears and the robot carries it out. With several robots, the same logic can produce congestion, avoidable empty travel and poor handling of urgent material movements.
The nearest robot is not always the best choice. Its battery state, load-handling capability, traffic conditions and remaining assignments can make another vehicle the better option. Priorities should be connected to production needs rather than configured only for convenient software behaviour.
Traffic belongs in the process design
AMRs share passages, intersections, charging points and transfer stations with other robots, vehicles and people. When a single narrow connection links production with a storage area, adding more robots can increase waiting time instead of throughput.
I would examine where tasks are released, where vehicles queue and which intersections become shared constraints. Some sites benefit from separated routes. Others need controlled access or redesigned transfer points. The answer depends on the building and operating process.
Utilisation does not equal useful work
High utilisation may simply mean that robots spend their time waiting, travelling empty or queuing at handovers. A more useful picture includes mission waiting time, completed deliveries within required deadlines, empty movements, exception rates and delays by location.
Fleet performance also depends on charging and maintenance strategy. It is not realistic to assume that every vehicle will always be available for transport.
Keep the operational objective visible
Fleet management should optimise material flow, not create the appearance of constant movement. I would validate the strategy against representative demand and real site constraints before deciding whether the facility needs more robots.
Keeping robots busy is not the same as delivering material when production needs it.