Quote:
Originally Posted by fusili
Rant away. You have understandable frustrations. But you do have some hints of ways to improve. Perhaps transit can work to better align frequencies, most importantly, ensuring that routes are running at similar times each hour.
|
Noooo, CT shouldn't bother with actually improving and optimizing conventional transit... An urban gondola will fix it instead! (Not directed at anyone)
Quote:
Originally Posted by suburb
The frequency is based on the length of the route divided by number of buses.
|
While this is true mathematically, the demand, therefore capacity required, therefore frequency required is what determines (or should) the number of buses, not the other way around. Perhaps I just misunderstood.
Quote:
Originally Posted by MalcolmTucker
Calgary's system is largely a resource optimization thing. No point adding pointless waits at the station or along the route to syncronize schedules. Better to have the bus out being productive.
|
This doesn't necessarily have to be an either/or situation. This can be changed in route/network design, making it longer or shorter (thereby fitting to the cycle time) - goes for the first part of the post below too.
Quote:
Originally Posted by ByeByeBaby
Not to sound like a CT apologist, but clockface scheduling with round multiples like you describe does mean buses have to spend longer waiting at timepoints. For example, consider a totally made up route that takes 77 minutes to drive, including recovery time. To fit in the 10/15/20/30 paradigm, there need to be 3 buses assigned to the route, and an additional 13 minutes built into the schedule somewhere, so the route takes 90 minutes and has 30 minute headways. That means that 14% (13/90 minutes) of the cost of operating the route is just in keeping the schedule neat and tidy. That also means that a trip across the timepoint is a lot longer; on a looping route this is especially difficult. (Consider a rider on a looping route like the 72/73 going from, say, Ogden to MRU -- some of the additional time added to keep to schedule would likely be spent on waiting at Chinook, so this makes the trip longer.) And I don't know if it's happened to you, but one of my pet peeves is getting to Lion's Park heading to Foothills, getting on the 91 that arrives, only to sit while the driver waits to go back on schedule and a 40 comes and goes.
|
I'm guessing you chose those times as an example purposely. That is sort of a rule of thumb recovery time (15%.) For a loop line I would say it is ideal that any extra terminal time above the 15% (to fit schedule) is of course done only at that one terminal and not, or very slightly so, prorated throughout the route stops. Again though, this can also be minimized by route design. Secondly the 15% is a rule of thumb. If a route cycle time, especially on one that is proven to have low variability is larger, this 15% can be lowered giving for a better terminal time to revenue operation rate ratio. However, it can't be forgotten that driver's need breaks as well and therefore, this constraint, and it is largely dictated from the union, is to be heeded. No idea what CT has to adhere to here. Anyone? It can also make run-cutting more difficult.
I guess the point I'm trying to make is that routes can be optimized so that there isn't an overly high proportion of non-operation time
and that they can adhere to a clearer, more user-friendly schedule. Not saying this is easy though, as the two act circuitously - ie, change route, change demand, change capacity, change cycle time, change frequency, change...
The pulsing, or time transfer system vs stagerring bus arrivals is another interesting thing though...
But back to the first, and yes, critical point about Urban gondolas and their applicability in Calgary. To be frank, that applicability is essentially next to zero with the odd exception that perhaps sometime in the future a Brentwood/University node is linked with a Westbrook one.
I don't want to say that it isn't worth investigating or knowing more about, but this may quickly become a case of a technology fascination.
Combining comments from the CBC article and the podcast, I believe that Steven Bates is not really getting the "Last Mile" problem right. This could potentially act as a feeder
if the O-D stations are serving very high demand nodes, which I don't think Foothills counts as, especially with all that parking. Not too mention, that while although cheap in capital for a transit system, it isn't exactly a cheap feeder service.
Let's, however, say that Foothills or wherever is a high demand node, then the 1 minute waiting time also quickly starts to disentigrate - so too does it in a non-high demand case for that matter. Yes, while you may wait only 1 minute to get on from the Foothills, all that has been done is that that waiting time has been allocated to the one after that trip - to the one for the LRT in our case. Perhaps not a big deal when the LRT is at a frequency of 5 mins. This however would become more evident if, people that are frequent users understand when tranfer times (based on the LRT schedule) occur, they would tend to all want to ride the same transit unit, in this case a gondola, at the same time. This then becomes impossible and waiting times occur. The same thing in reverse, as someone had already mentioned would also occur.
In essence a near linearly continuous capacity is serving or feeding one that, on a disagregated level, supplies a capacity at discrete time steps. It makes any sort of timed transfers impossible and, as the service is foreseen to be as such, isn't necessarily an effective feeder service in way of making transfers efficient (or theoretically efficient). To that point, and thanks to the LRT's large variability in schedule, this possibility might actually be diminished though. A case of two wrongs making a right?
I further doubt that the max capacity one could provide is even close to the current actual demand. Underutilization? With that, I also don't believe the claim that only 1 operator is needed for the system. One at each station, perhaps. What happens when they need a break? There is also a greater need for mechanical supervision and operations, as a technical break can prove much more fatal (and scary for the public in general) if something goes wrong - compared to a bus, let's say. This was the case in one of the championed (tourist system nonetheless), namely the Ngong Ping Cable Car in HK.
Lastly, its ability to go over everything, touted as an advantage, at the same time can also work against it,
especially in the case of solving the last mile problem. It might solve this problem only if your last mile happens to be the node it ends/starts at and is thus a very poor collector service. In fact, if the span of the stations happen to be the literal mile, it is the very definition of the last mile problem itself. The way to remedy this would be to add stations of course. This costs next to nothing for a bus. However, in this case adding stations destroys the exact two main advantages it has. Average speed and capital cost.
I don't want to come across a hater of this technology, but I see it quickly becoming the next monorail (admittedly probably a bit more effective.) It looks cool, effective and cheap on the onset and especially, dare I say, to those with a more limited understanding of PT.
That's not to say that Steven Bates should stop supporting and pushing this system, but I think the flat, topographically easy, high rate of existing road infrastructure city that Calgary is, is about the last place this system could work. Even Edmonton probably looks more applicable. DT to Whyte Ave? This is especially the case when simple, as we've just discovered, things such as making the bus system a bit more optimal can largely fix many problems. How about bus actuated garaunteed signalling?
Prev
Linear Mode
