1. Vehicle autonomy is not fleet autonomy
A vehicle can localise and follow a route correctly while the overall fleet still performs badly. Mining autonomy therefore needs a supervisory control plane that maintains a current picture of machine state, route occupancy, work-zone availability and mission ownership.
2. Shared traffic state
The orchestration layer should maintain a site graph containing roads, intersections, parking or hold points, loader approaches, dump approaches and exclusion zones. Each vehicle publishes position, heading, speed, health and mission state. The control plane then uses that state to prevent incompatible reservations and coordinate right-of-way.
3. Geofences and reservations
Geofences are useful for defining where a machine may operate. Reservations are more specific: they define which machine may enter a constrained resource such as a one-lane ramp, dump pocket or narrow loading approach during a particular interval. This turns traffic management into an explicit state problem rather than a collection of local guesses.
4. Mission assignment
A mission may contain route segments plus semantic actions such as TRAVEL, QUEUE, SPOT, LOAD, DUMP and HOLD. Assignment logic can consider machine compatibility, route availability, work-zone capacity and current fleet state. The vehicle remains responsible for local obstacle response even when the fleet layer owns the mission.
5. Mixed traffic
Real sites rarely transition to full autonomy in one step. The system therefore needs to represent manually operated light vehicles, support equipment, temporary closures and supervised machines. A conservative design treats unverified actors as dynamic obstacles unless they participate in a trusted traffic-management interface.
6. Supervisory visibility
Operators need a real-time view of fleet status, blocked routes, active missions, degraded machines and exceptions requiring intervention. This is the difference between a fleet that merely moves and a system that can be supervised as an industrial operation.
7. Industry alignment
Hexagon describes autonomous management systems that orchestrate vehicle movement, while Komatsu describes central command, optimised assignments and route/spacing control. Those public descriptions support the same architectural separation between onboard autonomy and fleet-level orchestration.