Just a Bill wrote:
The fact that we never see anybody waiting for a cab
Not so. There is at least one time that I can remember and it is from "The Naked Time".
Well, I did qualify that "never" with a 99.9% ... but yes, I should have said "almost never." (I wanted to say "rarely," but that seemed to understate the near-perfect availability we consistently see.) Maybe it's only 99% or 98.5%, but rather than quibble about the decimals, can we agree that we almost always
see immediate cab availability?
And the few cases where we do not see this usually seem to have something unusual going on. The "clear the tube" scene from The Naked Time is in a context where the entire ship is screwed up
, and the WHMHGB scene is from a pilot episode in which many details are different, and in some cases of debatable canonicity.
Regardless, my point was not to make an absolute statement that nobody ever has to wait. My point was, based on the observation
that folks almost never
have to wait, the system must be designed to replace cabs almost immediately. The fact that there are sometimes cases where the system cannot provide 100% on-demand service is to be expected no matter how the system is engineered.
Or Kirk and Spock in WNMHGB? There was definitely a gap between pressing the button and opening the doors.
I'm not entirely convinced Kirk is really "pressing a button." Those don't look at all like buttons; they look like indicator lights. And the shoulder- or head-height placement is impractical: children, folks in wheelchairs, diminutive aliens — heck, even just short people
— wouldn't be able to reach them. They present much more like a red-alert claxon (at the same height) or maybe "cab ready" lights than like any kind of control element.
Maybe Kirk is just leaning on the wall, or maybe Shatner or Goldstone decided on the fly to add some "action" to the scene by pretending he was slapping a giant, vertical button-that-looks-like-a-lamp ... but I highly doubt the set designer intended
those to be buttons: it would be a pretty stupid ergonomic design, we never see it again, and we don't even see the light change color or make a sound as a result of being pressed/leaned on.
I don't think we can conclude from this anomalous half-second of footage (which never occurs again) that turbolift cabs are waiting to be "called."
Robert Comsol wrote:
Actually, in my first TOS Enterprise
deck plan drafts I had individual turbo lift access points next
to the main turbo shafts. So in most cases, there's a cab waiting and just makes a quick turn to move into the main turbo shaft.
It's only when another turbo lift cab arrives
that the standby cab has to move out of the way.
So basically you have depots, but they are decentralized and very small; essentially a one-cab depot at every terminus. I don't know whether that's more or less efficient than shared, localized multi-cab depots, but it's an interesting thought. Let's take it one step further:
In computer programming, the concept of "stacks" comes from the stacks of dishes one sometimes sees in diners and buffets, held in place by vertical guides and supported underneath by springs. You can remove a plate from the stack for use, or drop new clean plates onto the stack, and the stack self-adjusts its height so the (new) topmost plate is always at the same height, the "delivery point."
We could minimize delays from cab deficits/surpluses by "stacking" them in much the same way. Suppose each lift terminus has a two-cab stack; something like this in the abstract:
Again, this is abstract. Maybe the cabs are stacked vertically, or maybe one is stacked on either side, or maybe the stacking is not immediately adjacent to the access point; but I think this visualization will suffice for the moment.
Base scenario, let's say everyone's asleep and no cabs are moving. Every terminus has two cabs: one ready and one standby. The overflow position is empty.
Use Case 1: Departing occupants.
Ready cab leaves with occupants; Standby cab immediately moves to Ready position (this answers the redshirt/Lazarus question)
. An empty replacement cab is dispatched (from "somewhere") to restore this terminus to its two-cab state, automatically queueing up a pending Use Case 1a.
Use Case 1a: Empty replacement cab arrives.
Ready cab moves back to Standby; replacement cab becomes new Ready cab. Default terminus state (two cabs) now restored.
Use Case 2: Arriving occupants.
Ready and Standby cabs simultaneously shift to Standby and Overflow positions, respectively. New cab arrives and occupants disembark. Doors close and new cab, whether occupied or not, is immediately sent away (to "somewhere"); Standby and Overflow cabs simultaneously shift back to Ready and Standby positions, restoring terminus to its default two-cab state.
Use Case 3: Extra cab to/from "somewhere."
Presumably there is still a depot or spur or twin-track area where extra cabs can be sent to and retrieved from; although in this scenario it's okay for such areas to be both few and distant (since each terminus has its own mini-depot, it can usually afford to wait a bit for a replacement). Or, perhaps there is not depot at all, and each terminus with a cab deficit simply waits for the nearest surplus to occur. (With twice the number of cabs on board as termini, this might work out fine under most circumstances.)
Use Case 4: Routing efficiencies.
In some cases, some occupants will leave a terminus while other occupants are en route to that same terminus; in this case, no surplus or deficit need be corrected as the arriving cab will restore the terminus to its two-cab state. The routing algorithm of course gives high priority to these situations. In many other cases, when a replacement cab is needed or an extra empty one must be sent away, there will be another terminus not too far away that is in the opposite state. The routing algorithm will optimize these to pair one terminus' surplus with another's deficit as efficiently as possible. There is usually plenty of time for these adjustments to take the "slow route" (letting occupied cabs have right-of-way priority in the network) since, again, each terminus has its own mini-depot that allows two successive departures or two successive arrivals before a delay will occur.
Use Case 5: High- and low-priority termini.
A high-priority terminus such as on the bridge may have more than one Overflow position. (Perhaps they are wrapped around the bridge in the dashed-line space surrounding the studio set, or stacked below the bridge.) A low-priority terminus such as near a seldom-use storage area may omit the Standby and/or Overflow positions (and in fact may not even maintain a Ready cab if the anticipated use is infrequent enough: having to wait 30 seconds for a cab once every month or two is certainly no hardship).
Greater routing efficiency could be achieved by hooking up more shafts to the "overflow" pipes so that extra cabs can be added/deleted without getting in the way of arrivals and departures (a service entrance, so to speak), and there's also the idea of "continuous loop" routing a la
blood circulation, but both of those require a lot more shafts and may make it impossible to produce a believable blueprint.
Anyway, if space-efficiency is the highest goal, then this scheme may or may not be optimal; but if minimizing arrival/departure delays is the goal, then I think this design has a good chance to win the contract.
My apologies if this has already been discussed to death before I came along.