If you’ve ever visited a water park, you’ve probably seen a “Lazy River” ride like Castaway Creek at Disney World’s Typhoon Lagoon or Rambling Bayou at Adventure Island in Tampa, Florida. They are shallow channels of water with a strong current. You can sit in an inner tube and just ride the current, or you can swim with the current, and get a boost that makes you feel like an Olympic swimmer. Some parks are completely encircled by a Lazy River that serves as a kind of transit system. The river is the guideway, and the stations are steps or ramps that lead into the water. In a Lazy River, traffic jams can actually be fun!
Automated Transit, in its simplest form, can also be configured as a single loop, with more complicated single-level networks comprising multiple intersecting loops. In the same way, a Lazy River could be expanded with additional loops. Of course the stretch of canal after a merge must have enough capacity to accommodate the combined water flow of the two canals that feed into it. Similarly, it must have enough flow to maintain the current in the two channels of the upcoming diverge. That means that the segment of canal between a merge and a diverge must have twice the cross-sectional area of the canals that feed into, or out of it. I refer to the part between a merger and a diverge as a double-density segment, and the parts leading into or out of it as single-density segments.
It is well known among automated transit engineers that single-level Automated Transit Networks (ATNs) also have single- and double-density segments. If two guideways that lead into a merge are each fully loaded with pods running at the minimum headway, the following segment of guideway cannot accommodate the traffic without violating the minimum headway requirement. This is why it is considered good ATN design practice to alternate merges and diverges. If there were two successive merges before a diverge, then you would have a triple-density segment.
This phenomenon is the source of the specious claim of automated transit critics that ATN systems cannot operate at more than half their capacity. They reason that in order for a transportation system to be economical, every part of it must be saturated with traffic at peak times. In the world of automobiles, this is known as gridlock. So by the flawed logic of the detractors, the only transportation system that is “practical” is one that is gridlocked.
It is true that a single loop can be packed with vehicles along its entire length. And that’s fine if all the places you want to go are arranged along that loop. But as soon as you add the opportunity to turn left or right, you are faced with the need for extra capacity. That’s because sometimes all the vehicles will want to go left, and at other times they will all want to go right. This is true for any kind of vehicle that does not follow a fixed route, including automobiles.
Think of one of those puzzles with the tiles that you are supposed to rearrange in numerical order by sliding them into the one open space. By the critic’s reasoning, that one empty space is wasted. But if you were to add one more tile, then nothing could move at all. How efficient is that? Mobility requires room to move. An adequate amount of unused capacity is not a design flaw – it’s a necessity.
Moreover, the only reasonable measure of a transportation system is customer satisfaction, which by definition includes everything that matters to people. No one chooses a mode of transport because it has better capacity utilization. What people care about are things like comfort, safety, reliability, simplicity, cost, and travel time (including any walking, waiting, parking, transferring, etc.). By that standard, automated transit runs circles around anything else.