From another thread, I got thinking about public transit systems. Public transit is only a facet of the transit system; it is another part of moving people (and things) from one place to another. Cars are a big part of that: people are 30 miles from work, so we have roads that allow that mass of people to get from home to work and back by driving their cars. Public transportation serves the same purpose: allows the mass of people to get from Point A to Point B and back, where (A) and (B) are near home/work.

This means that we must treat the transit system as one object. More bike infrastructure? Better public transportation? Grow up; it's all useless without a real plan for how to knit it together. What's bike infrastructure going to get you? More 50 mile trips nobody's going to go on. And what about subways? Either start-stop-start-stop every 2 miles (and who's going to walk 2 miles anyway?), or a 20 minute bust trip 5 miles to the nearest station, or whatever other horribleness that'll keep people off the rails.

We have to look at the transit system as one big project with multiple pieces that we use to compliment each other. To start, we have to think about what we have available, and what to use it for:

• Roads to carry cars, bikes, and public-transit buses and taxis.
• Sidewalks to carry pedestrians
• Multi-Use Paths to carry pedestrian and bicycle traffic away from the road
• Buses for local road mass transit
• Taxis for local road express transit
• Light Rail for high-efficiency mass local transit
• Sub/Rail for high-speed, longer-distance mass transit

We also have to consider that some of these merge into others:

• Roads and Sidewalks merge into Multi-Use Paths when appropriate, for example when a bridge over a main road crossing would move pedestrian and bicycle traffic more efficiently. For example, here's a sidewalk.
• Junction transfers between public transit, for example a bus or light rail service that travels to the local subway station.

There is also the consideration of operation for various types of transit:

• Trains and light rail only operate on routes fixed by construction, and only board and discharge passengers at stations.
• Buses, cars, and taxis follow the roads, and thus have some discretion for paths to take due to the more connected infrastructure; however, they are also mutually affected by traffic.
• Pedestrians and cyclists follow either the road (sidewalks) or dedicated MUPs. Because of crossing with the road, road traffic affects cyclists and pedestrians; additionally, cyclists ride in the road with traffic. Both of these are speed-restricted and distance-restricted by the physiology of persons being transported: these transports function under human power.

Because of the interaction between cross road traffic and pedestrians/cyclists, it makes sense to build cheap overhead MUPs to cross busy roads, with stairs and ramps (merge-off and then over-the-sidewalk where possible, or spiral for compact), to move cyclists and pedestrians. Physically separating the bike and walk lanes on these paths seems like a good idea, two-way each. Commanding that cyclists walk across short street-crossing bridges also seems feasible; you can walk across the street, and the bike ramp becomes a wheelchair/EPMAD ramp as well.

Because trains only operate on stop-start routes governed by stations, it makes sense that fewer stations allows for faster rail transit. However, because pedestrians take forever to walk five or ten miles, it seems that spacing the stations very far apart would limit pedestrian access. Thus, light-rail or bus service intended specifically to bring pedestrians to a station is key. Due to the delay caused by many starts and stops, a larger service area is more efficient for the sub-rail; but smaller service area is more efficient for light-rail or bus service. Thus, a balance must be struck between these.

Further, cyclists can move much faster than pedestrians; however, cyclists also cannot simply leave their bicycle somewhere. Even given ample secure storage, the cyclist's advantage is lost on one end of the system: if the cyclist is 5 miles from a station, it's likely that his destination is lain out similarly and thus he has a fairly high chance of being 5 miles from his destination upon discharge. Thus there is an increase in need for bicycle transport as we move through larger-area mass transit systems.

In other words, the light rail system could service an area 5 miles wide centered on a sub-rail station: subway or train stations are 10 miles away from each other, so a person is never more than 5 miles from a station. A pedestrian will find a 5 mile walk infeasible; but a half mile walk to the light rail is feasible. A cyclist will find a 5 mile ride a normal impedance at best, similar to a 1 1/3 mile walk if the cyclist takes a level path at a continuous speed of 15mph, and somewhat easier if any coasting or downhill sections are involved due to short rests. Thus the cyclist will tolerate the ride to the train station if he can't carry a bike on the bus or light rail; but he won't tolerate leaving the bike behind for his 30-50 mile ride on the subway, unless he finds the light rail at the destination satisfactory.

As for actually doing the major train transit, I have no idea.

The only way I can think of to efficiently connect train systems so they're equidistant from each other is an overlapped hex grid. Unfortunately, this has a transitional problem: you would have to get off every 10 miles and transfer. However, a recursive system presents an interesting solution.

Let's say each train follows a hexagonal ring: each hex has 12 trains, one going in each direction at each vertex. So for stations A through F, there is one train departing to travel clockwise and one departing to travel counter-clockwise at any given time. Similarly, one arrives just after. The train departure times are staged such that as each train departs, all other trains arrive; in the case of a late departure, the arriving train stops before the station and waits; the entire ring is assumed desynchronized; and it creates artificial delays to synchronize itself.

Each hexagon shares each side with one of six other hexagons, which means it must supply track capacity for two more trains. In other words, there are four parallel tracks lain at any given point: one each way for one hexagon, one each way for another. This also means that a desynchronized hexagon makes for interesting arrival anomalies, as one at a given station (it's only desynchronized in one direction) will be involved in resynchronization delays.

Additionally, because of the offset grid, the center of the hexagon (which is the same distance from every vertex on the hexagon as every other vertex is from its connected vertex), each station has not three, but rather four hexagons it services. each vertex is shared between three hexagons PLUS one whose vertexes include the centers of those three hexagons as well as one point on each of the three hexagons.

Thus for this system, each stretch of track must carry four trains, two going each way; and each station must service 24 trains (draw it, you'll see it).

To scale this, repeat the whole thing, and in fact reuse it: a certain subset of these stations will service 48 trains. If pedestrians will go 0.5 miles to a light rail to go 5 miles to a sub-rail, then they'll ride a sub-rail 50 miles (with a maximum of 5 change-overs) to board an Express Rail that follows a hex grid with each station 100 miles from any other station. The next expansion from here is, of course, 500 miles to the nearest 1000-mile cross-continental; in which case you may care to make those stations fall every 500 miles instead, with the obvious odd effect of having what amounts to twice as many bigger stations, but really doesn't demand any greater per-station capacity.

The stations in that system would handle 24, 48, or 72 trains, each in isolated areas of 24 ("Local," "Express," "Continental" for 10 mile, 100 mile, and 500 mile stops). To go anywhere within 100 miles would require at most 10 change-overs. to go anywhere within (below) 500 miles would require at most 14: 5 on the Local, 4 on the Express, 5 on Local at destination; if you need to go about 500 miles, you're going to hit the Continental route before then. And of course 5000 requires 23: 5 Local, 2 Express, 9 Continental, 2 Express, 5 Local.

Be mindful that, at 120mph, the Local change-overs happen every 5 minutes; Express happen every 50 minutes; and Continental every 250 minutes or roughly 4 hours.

It's also possible to do exchange routing: a common route would exchange trains between hexes, so you could have a "New York to LA" trip that involves getting on in New York, sleeping, and waking up in LA. This could even exchange trains between major circuits: a Local train could swap into an Express ring, as long as that Express ring swapped one of its trains into the Local ring as well. Similarly between Express and Continental. Obviously, all Express stations are also Local stations; and all Continental stations are also Express stations.

I never said this made business sense due to start-up cost and requirement for a national infrastructure; however, this sort of mass-transit system seems like it would work well. It may prove annoying for custom routes; but no more than booking flights or chartering buses (which have set routes as well). The selection of a 5 mile major sub-rail maximum distance places train stations well within bicycle reach; thus, providing for bicycle transportation on trains and providing for bicycle transit is largely important to solving the last-mile problem for this sort of system. As well, localized mass-transit designed to collect pedestrian traffic and move it efficiently to the nearest station would prove highly advantageous with this sort of system deployed nation-wide. Thus, transportation system design would have to address motorist, cyclist, and pedestrian traffic efficiently, as well as taxi and bus service needs, in order to eliminate the last-mile issue.

It's cute, but it'll never happen. Still, thoughts?

There's so much information here, I don't know where to begin! You obvious put a lot of though into this idea, maybe too much. There is so little rail transit in this country and all of these ideas you put forth will require hundreds of billions in new investment. I don't see this happening and rail transit has been reduced all over the nation. So we are going in reverse, not forward.

When it comes to transit, lightrail gives you the biggest bang for the buck. Much cheaper than subway and just as expensive (if not more) than commuter rail. However, lightrail if done correct, can bring in hundreds of million if not billions in investment capital around the rail line. Subway and high speed rail will do this but at a much higher cost and commuter rail may not do this at all!

I don't believe we will ever see a national train network other than Amtrak.

Yeah but light rail is not good for moving masses of people extremely long distances. It'll shuttle you around town, slowly; but it has the same problem as a bus system: it's very slow. Cars are far superior to buses because cars go directly from where you are to where you want to be, not on a route that eventually gets there; light rail as well, though the only light rail I've seen acts as a shuttle between mid-town and downtown. Very limiting.

This is why I have express rings and train swaps to establish complete routes without change-overs and with few stops. Mass transit has trouble getting you from one place to another, except over long distances: planes are like trains and cars slapped together, freedom of movement (cars) with no traffic and a pretty direct pre-selected route (trains). You don't need to lay track to get a plane to a particular point. Roads approximate this; trains less so.

Yes, I know this will never happen. Also, as for a lot of thought, I shot this from the hip; 20 minutes, tops.

dude, I stopped reading after seeing the poster quoted himself... getta blog bro!