Valet Breakdowns · Multi-Point Operations
Several doors.
One inventory.
A campus operation may accept a car at one entrance, store it near another, and return it to a third. Once the curb becomes a network, local memory stops being an operating system.
An unnamed archetype
The campus is a distribution network with guests inside it.
This breakdown does not describe a named resort or claim a relationship with one. It examines the common geometry of a campus: multiple arrival points, parking areas at different distances, and guests whose desired pickup point may change during the stay.
The important shift is conceptual. There is no longer one stand with overflow. There are nodes, routes, and a shared inventory that must stay legible to every node.
The network
Local optimization can make the whole campus worse
One queue per door creates conflicting truths
If each stand sees only its own requests and keys, no one can tell whether the nearest car, runner, or open handoff belongs to the campus's highest-priority need.
Separate stands
Local memory- Simple when every car returns to its first door
- Transfers become phone calls and favors
- The same inventory can look available twice
Campus network
Shared state- Every car has one current location and owner
- Pickup destination is explicit
- Dispatch can account for route and node capacity
→Shared state does not require a central dispatcher to make every move. It requires every local decision to update the same truth.
Route the request before routing the car
A changed pickup point is cheap while it is still information. It becomes expensive after a vehicle has been staged at the wrong door.
→Campus rules, pedestrian areas, and allowed vehicle routes are property-specific. This sequence is descriptive, not routing advice.
The minimum record for a distributed inventory
Every node needs enough context to continue the service without asking the originating stand to retell it.
Vehicle
Current physical position
A location precise enough for someone at another node to act on.
Key
Current custody
The key and vehicle remain linked even when they travel separately.
Guest
Requested pickup node
The destination is part of the request, not a note added after dispatch.
Work
Current owner and state
Every transfer says who owns the next move and when that ownership ends.
→The network works when a shift can change and a request can move without forcing the operation to reconstruct either history.
Where Valletto changes the math
A campus needs one operational picture, even when it has many front doors.
Shared session state keeps vehicle location, key custody, pickup node, and active assignment together so each stand can act locally without fragmenting the inventory.
The takeaway: multi-point valet is a distribution problem. The product is not merely finding the car; it is delivering the right car to the right node without losing the chain in between.
