“Whatever you can do, or dream you can do, begin it. Boldness has genius and power in it.” – Goethe.

A city-wide network topology

Long-distance rail lines need to be reasonably fast and they need to have a higher throughput than simple cables above the street. Teleport pods are designed to be carried singly up elevator shafts, singly by Teleport one-pod trucks that hang below cables or below rails, and in larger groups on Teleport-compatable train systems.

Below I’m going to be describing an ideal topology for four-pod (the number "4" is rather arbitrary for now, capacity may vary) above-street Teleport trains.

My current ideal for four-pod Teleport train coverage of a flat, rectangular city combines a number of L-shaped four-pod Teleport train tracks. Your urban district may have hills and crooked streets.

Not shown in the diagram below are the secondary above-street feeder slack cable lines that solve the last mile problem from people's houses or from street-level elevator doors to the four-pod trains. Also not shown are connector tracks at every place where two tracks cross, and various sidings at the terminuses of the rail lines.

Two of these tracks cross at the very center of the city. Those two tracks are straight.

All tracks have both of their terminuses out in the city's suburbs. The network allows four-pod trains to travel from a particular suburban terminus directly through a city to one other suburban terminus. On rare occasions a train will travel between these two points without stopping.

The network is designed so that every track crosses over or under every other track at least once. At these transfer points, entire four-car trains might drop individual pods off for automated transfer of the pods, or an entire train running express from one terminus to a certain other terminus might switch tracks on a siding.

The completely automated network will try its best to pack cars onto a four car train so that the train can run rather directly toward its destination. The ultimate destinations of each pod was earlier set when the owner got into that pod or when they otherwise programmed the pod's trip, so the scheduling computer has had a few minutes notice beforehand for all pod trips.

The other eight tracks on this topology (or twelve or sixteen tracks will also work) all bend in an L shape around the designated center of the city. Each track has its ell in one of the four quadrants around the center, northwest, northeast, southwest or southeast.

In this configuration, all eight ell-shaped tracks run within two units, two square blocks, of the city center. The other side of each ell swings wide of the city center. I’ll expect that one square equals ten city blocks by ten city blocks, or about 1/2 mile square. And so, every track somewhat services the populous city center.

Every suburban terminus has multiple sidings (not shown) where idle 4-car trains can sit and wait to be filled with cars, preferably all going toward one specific area of the city but occasionally trains can be packed willy-nilly. Every suburban line is a single track with regular sidings (not shown) where traffic traveling in one direction can pull off the main line to let traffic traveling in the other direction through. Pulses of 4-car trains can travel in opposite directions on the suburban spurs of each track, passing each other at these sidings. A typical siding connects to the main rail line at both ends of the siding. If one switch fails, the other siding switch allows the system to continue functioning, even if the throughput will be slower until the switch is fixed.. Also not shown, a slow network of cables allows pods to be passed around any out-of-operation section of track. In the event of a track failure at rush hour the network computer can swarm the area with extra cable trucks to reduce the jam-up.

At every intersection where a track crosses above another track, both tracks have a siding. A connector track runs between the two sidings, allowing individual 4-car trains to switch tracks. 4-car Trains are pre-loaded with cars all going to approximately the same place, and then individual car-trains can make only one single track switch to get to their destinations. For the most part, trains always travel forward toward their general destinations. The exception would be near the edge of the city. A car-train trying to get from one suburban terminal to an adjoining suburban terminal would most likely be better off switching two tracks rather than going all the way through the city’s center twice.

A ten track system gives us a total of 2 x 10 = 20 terminal sections of rail heading out into the suburbs. In an idealized situation, the 20 suburban rail spurs would radiate out in a 360 degree circle, all ending in terminals about 15 degrees around the circle from the next terminals in the circle.

In this idealized network north-south tracks are always above east-west tracks, or else east-west tracks are always higher.

All sidings are equipped with the ability to transfer Teleport pods between a car-train and Teleport one-car carriers that are capable of traversing a pair of slack wire cables hung above individual streets. Individual Teleport cars should be able to speedily go anywhere in a city to within a block of any destination. Many stores will have indoor Teleport stations, and many buildings will have Teleport-ready elevator shafts.

Also, individual cars can be hoisted from one siding on cables to a nearby siding to wait for another car-train to come by, one that is headed in the right direction for that car. Teleport is akin to (fully automated) hitchhiking in a sense.

For Teleport car-trains, right now I’m thinking in terms of only four cars per train. The track will only have to hold 4 cars at a time, because trains won’t be allowed to get closer. 16 car larger trains are always possible. Intercity railroad trains capable of carrying perhaps 100 pods are likely.

With a fully automated and complex network, we only need one track per line, not two. We need regular sidings to move trains off the main track so that opposite trains can pass.

Rush hour and after midnight system performance:

These are for the most part single-track lines. During rush hour a number of 1-way tracks will naturally form around the city center. Many 4-car trains will need to exit onto 1-way tracks going their way.

Remember that the entire trackage is automated. Almost as soon as a patron steps into a Teleport pod, the system knows specifically where that pod is headed. For this reason, the server already knows how many pods are arriving, exactly when they are arriving and where they need to go. The network is programmed to minimize the sum of the squares of the waiting times in seconds that individual patrons are forced to endure. Note that the waiting times of high priority vehicles, police, fire and ambulances, have 100 times the weight of individual cars, so that they get through in excellent time. Teleport can also overweight full pods during rush hour and it can underweight freight pods if the shipper wants to save some money. Empty 4-pod trains are only a priority because more suburbanites are going to want to ride with the morning or afternoon rush.

Given one pod on the system that is needed to complete a four-pod train and then the train can be sent down the track, that local pod will either be hurried by the local system or the four-pod train will be sent off partially empty. It's not good to keep the other three pods waiting.

After midnight the coast is clear and 4-pod trains, often carrying only one pod, can often speed to their destinations without stopping. Not stopping is rather energy-efficient and it's certainly time-efficient.

Caravan Ordering

A colleague once told me that high speed rail tracks are at the mercy of balky rail switches. The next car or train coming down the track needs room and time to stop in case the switch ever gets stuck halfway between the main track and the siding.

My first fix would be to have longer multiple-car trains, in order to limit the number of switching motions that the switch needs to perform. Beyond that I recommend caravan ordering of all cars and mini-trains before each caravan enters the main track. The ordering is based on where each sub-caravan of trains in each caravan will exit the main rail.

For the following example, I'm going to assume that two medium-speed one-way rail lines merge into a high-speed rail, and then at the far end of the high speed rail are a number of rail switches to medium-speed rails. For input line A I'm going to always sort each inputting caravan into five sub-caravans numbered 1, 2, 3, 4 and 5. On input line B I'm going to always sort each inputting caravan into sub-caravans numbered 5, 4, 3, 2 and 1 in reverse order. At the merge point a caravan from line A merges onto the main line, then the rail switches over, then there's a pause of time, then a caravan from line B merges onto the main line, then the rail switches back, then there's a second pause of time. So, on the main line the cars or trains are sorted into the following order: 1, 2, 3, 4, 5, a gap, 5, 4, 3, 2, 1, a gap, 1, 2, 3, 4, 5, a gap, 5, 4, 3, 2, 1, yet another gap, and this series repeats forever.

The next siding happens to go to station 5. Certain trains in the mega-caravan slightly slow down on the track to create the following caravan gaps: 1, 2, 3, 4, a gap, 5, 5, a gap, 4, 3, 2, 1, 1, 2, 3, 4, a gap, 5, 5, a gap, 4, 3, 2, 1, 1, 2, 3, 4, and this series repeats forever. Now the 4, 3, 2, 1, 1, 2, 3, 4 caravan can continue down the main track, we've left time for the switch to safely flip over to the station 5 siding, the 5,5 caravan goes down the siding, the switch flips back, and this series repeats forever.

The next siding leads to a spur line that goes out to stations 3 and 4. The remaining trains form into the following caravan groupings: 1, 2, a gap, 3, 4, 4, 3, a gap, 2, 1, 1, 2, a gap, 3, 4, 4, 3, a gap and this series repeats forever. The spur line switch again has time to switch back and forth safely.

This caravan pre-sorting system can extend backward and forward for quite a large network topology.

Cellular transit networks

Breaking a large network into cells reduces the entire transit system to a more easily handled level. It also reduces the chance of system-wide blackouts.

Each cell is given the times that new cars will enter the cell from neighboring cells. The cell tells the system control computer the estimated times that cars will leave the cell for an adjoining cell. The local cell controller must also anticipate what local patrons might walk up to local stations. If a cell controller fears that it may become overloaded and jammed, it negotiates with the central computer. Naturally there are backup control systems if any particular primary control system goes on the blink.