Doggy-livery

Dogs can be trained to follow a scent from point A to point B. Dogs also urinate along their trodden path, leaving scents that other dogs will find. If we construct a grid of pathways, with trained dogs receiving commands from centrally controlled stations at the intersection of each path, we can use this trainable scent-leaving/following behavior to create a new dog-based delivery service. Each station would also dispense water, food and treats to keep the dogs happy and well-fed. Certain stations would be further elaborated into delivery hubs, where the dogs would pick up or drop off their packs, and human labor would take over to deliver the packages to their precise final locations.

So, here is the basic layout of such a system:

The pathways (in theory at least) are laid out in a grid along the cardinal directions. Each station, then, has paths leaving in four directions. And each station is controlled by a central computer that manages traffic on the grid, tracking the movement of each dog and telling the station in which direction to send a dog next. Each dog has an RFID tag in its collar, so it can be uniquely identified upon arrival and departure from a station.

Two classes of dogs work in this system, scent-leaving dogs and scent-following delivery dogs. Scent-leaving dogs are trained to follow a single path for the duration of their assignment -- always walking back and forth between a set pair of stations on the grid. Their natural urine-marking behavior distinctly marks an assigned path with a scent which the delivery dogs will follow.

The KEY to this system is the way in which the stations manage the dogs and the scented pathways. Because there are four distinct paths leaving from each station, the station must manage four distinct scents to mark those paths. It does this using non-toxic scent-marker chemicals that are given to the scent-marker dogs in their water bowls. Uniquely identifying each arriving dog by their RFID dog tag, the station knows which path the dog is assigned and gives it water laced with that path's scent. That laced water is converted to urine which the dog dutifully, and naturally, uses to mark his assigned path. The delivery dogs are given normal, un-laced water.

When a delivery dog completes an assigned path from one station to the next, he is rewarded with food and water. The station then communicates with the central control computer, alerting it to the dog's arrival and requesting the direction in which the dog should be sent next. The central computer recalculates the dog's route -- taking into consideration any new obstacles between his current point and ultimate destination -- and responds to the station with the next direction in which the dog should be sent. The station then emits an appropriate scent from an aerosol canister, which the dog has been trained to know is his next assigned route. He sniffs the ground, picks up the scent (left by a scent-leaving dog) and off he goes. At each subsequent station, the dog is fed, watered, and his next direction issued by the central computer; and so on, until he arrives at his final destination station.

Despite all the moving parts, it's a pretty simple system. As long as only four routes emerge from any given station, only four types of scent are needed to supply the entire network. North-south routes alternate between scent A and B at each station, and east-west routes alternate between scent types C and D. The central computer tracks dog movement along such a grid, and can dynamically recalculate dog routes at any time to avoid bad weather, malfunctioning stations, or overturned trucks of cats. The system for rearing and training dogs will be an added complexity, but one providing fun and rewarding jobs to displaced postal workers.