Most people who aren’t transit enthusiasts probably won’t tell the difference between a VAL metro (as well as some other non-standard automated metro or people-mover technologies) and an automated metro using standard steel wheel technology. The VAL metros in Rennes, Toulouse and Lille are excellent and the service they provide is very frequent and reliable.
That doesn’t mean that a city that wants to build a new metro line should choose VAL technology; standard metro technology would probably be better in most cases. But in cities that have already built VAL systems, I’d say the user experience on these metros is quite similar to the user experience on standard metros.
Well said. Amtrak/Commuter Rail/LRT runs next to San Diego airport but it's separated from the terminals by the runway. Ideally, the airport would relocate the terminals to abut a new Amtrak/Coaster/Trolley station, but the airport refused.
The next best concept is then San Diego's latest plan: an APM providing a one-seat ride between Downtown and the airport with 2-minute frequencies. The airport terminals are on a geographic cul-de-sac, so the only alternative to an APM would be branching off an LRT trunk that already has woefully inadequate frequency.
Yet some people still prefer the airport LRT concept because they believe "unlike LRT, APMs can't be extended past the airport into dense urban neighborhoods" or "the LRT will attract more riders than the APM because it looks cooler than the APM."
Automation allows for a lot of flexibility: increase frequency on extra demand/events, long operating hours, low operating cost etc. That's far more important than vendor flexibility.
The other aspect is that elevated rail can be complex to build while APMs are often much more prefabricated: In North America people know how to build elevated highways and therefore elevated rail isn't built much different, meaning you first build it like a road, then you put tracks on top, then railings on the side and the power system. APMs (and monorails) are usually far more integrated, newer systems like TSB's maglev are all precast/fully integrated. A single vendor can make procurement much easier, too, and therefore reduce complexity, schedule and risk.
So I see APMs, monorails and even urban gondola lifts as a great entry point to automated transit.
They can work well, but vendor flexibility is important. The LAX people mover is costing as much as the Canada Line despite having 0 tunnel and will be locked into a single vendor!
I would attribute that more to TransLink's vs LAWA transit expertise.
Ultimately transit projects should be open to any bidder with any technology so that the provider can make the proper decision based on price, extension flexibility, vendor selection etc.
“Generally, I think transit problems need to be broken down along their major features — capacity, alignment, network structure — and then we can find a particular solution that best matches needs across all of those dimensions.”
Well put. And benefit vs cost is also part of the equation. As you state, if the solution is a people mover, fair enough.
Writing this as a Singaporean reader who has experienced all three feeder APM systems here, and regularly use one of them.
APMs are really quite inappropriate especially in the Singapore implementations (and I'm referring to BOTH the Sengkang/Punggol APMs that you highlighted, and the Bukit Panjang one elsewhere), because they really are not the best tool to play the feeder transit role they were designed to serve
First up, their capacity really sucks from the perspective of the individual vehicle, and it's not like they're running these particularly frequently (6 minutes intervals at the peak if your destination is unlucky). The Bukit Panjang system also operates a unidirectional loop outside the peak period, so sometimes taking the bus is actually faster, especially when there do exist multiple buses with a combined frequency better than the 5-minute off-peak interval. For the Sengkang/Punggol systems you highlight, these are located in highly populated towns which are much denser than the average public housing town in Singapore, which makes low-capacity, low-intensity transit like an APM very inappropriate for feeder service. It doesn't help that for Sengkang/Punggol there is branching involved, which means it's really difficult to freely add trains as you wish, since overloading the line simply slows the ride immensely (you can feel it when riding the APM into the central stations -- they have to stop and wait for the pod in front from the other branch to depart first before proceeding), and in the case of the Punggol lines they are a significant distance away from the depot (the pods have to pass through the *entire* Sengkang West Loop, proceed to the Sengkang-Punggol connecting viaduct, then travel the entire length of the Punggol West loop to reach Punggol central station)
The elevated nature of the APM lines makes it cheaper to build, but honestly the amount of infrastructure for it is overblown for such a low-capacity system with stop spacings equivalent to the parallel bus lines (especially for S/P APM). Also, this adds to *access penalty* which is the time taken to travel from street level to the platform. Factoring this in, it makes a feeder bus ride (which does exist for all 3 towns here with feeder APM systems) more preferable than riding the APM, especially if you're travelling to only the town center, say for getting groceries or eating out, and that's purely based on travel time, without considering other factors like comfort and convenience. Even if you're transferring to the MRT and going elsewhere, taking a bus and connecting to the MRT may be a more competitive option than a similar route via the APM systems instead. In Bukit Panjang, residents willingly wait longer for bus 976 which is a full duplication of the APM, as evidenced by its long snaking queues at Choa Chu Kang Bus Interchange in the evening.
Not all APM systems suck, the ones in the Greater China area are pretty decent, although you must take note that they are strictly for low-ridership, low-density applications that are more geared towards expanding catchment. It's just the way the Singapore ones are set up which make people movers a subpar choice for feeder transport here. Politically, they've also been an embarrassment for the government, which had to admit that they were built under pressure.
In any case, I highly encourage you to come visit Singapore and see the system for yourself to decide, and perhaps I could show you around too, my schedule permitting!
I mean I think a lot of the issues in Singapore have less to do with the mode than the implementation - 6 minute headways are bad on a fixed guideway system! IIRC the Bukit Panjang the alignment is truly awful. That being said I am skeptical that buses are really the solution for such extreme density as you see in that part of the city - and I like buses!
I only brought Singapore up as an example because it really did motivate my interest in the topic!
When APMs don't give you a significant capacity advantage over buses, and all the additional downsides that buses don't have, it doesn't make the bus any worse an option for said suburb (which is what HDB towns in Singapore are translated to NA context) frankly...
Some bus routes in these towns (both feeder and trunk routes, if you're wondering) operate as frequently as the APM and I dare say the combined capacity from having multiple buses down the same road, mostly operating double-deck/articulated fleets can outrival that of the APM systems. I tend to think of the APM systems as outvalves for the crush-load demand for existing bus services (which was indeed the case, so much so we had an entire rail line segment that exists really because of this issue), in cases where the option with more infrastructure appears rather overbuilt for what it can deliver.
Case in point: Downtown Line and its light-rail level capacity was built to release demand from Bus 190, so infamous for its crowds it made it to national news in 2013: https://www.youtube.com/watch?v=gY65MF6NwAo
Stage 2 of the DTL which parallels 190 opened in 2015, while they had to launch a second duplicating bus service (also equally frequent last time, but now with heavily reduced service after DTL2 opened) in the meantime.
And then there is the more extreme length versions of cable hauled people movers, like the new London Luton Airport DART (connecting the airport to the heavy rail train station, replacing a less than ideal bus link). And yes Geoff has a video.
And the DART will be extended further when the airport gets a new terminal (planned into the design). Built by Doppelmayer (and opening was delayed while some challenges with the cable hauling were resolved). Similar tech also used at Birmingham (UK) airport for a similar link, replacing what was the world's first passenger Maglev (short, end of life, and not that quick)
The DART had to be a cable hauled or rubber tyred for cost reasons--there is quite a height difference between rail and terminal area (which are about a 1.5km apart, so quite a gradient), and tunneling plus lifts would have been much more expensive, and steel wheels would have needed a quite indirect route and likely some significant earthworks.
(Luton is also one of a select few UK cities that has a guided Busway, but that's the other side of town, west towards Dunstable)
Most people who aren’t transit enthusiasts probably won’t tell the difference between a VAL metro (as well as some other non-standard automated metro or people-mover technologies) and an automated metro using standard steel wheel technology. The VAL metros in Rennes, Toulouse and Lille are excellent and the service they provide is very frequent and reliable.
That doesn’t mean that a city that wants to build a new metro line should choose VAL technology; standard metro technology would probably be better in most cases. But in cities that have already built VAL systems, I’d say the user experience on these metros is quite similar to the user experience on standard metros.
Oh it's absolutely similar and the difference to the user is negligible, but the cost for operators can be real!
For short journeys, with lower demand, people movers are good, for either longer or higher demand routes, they are not.
Those are some guardrails for their use for sure!
Well said. Amtrak/Commuter Rail/LRT runs next to San Diego airport but it's separated from the terminals by the runway. Ideally, the airport would relocate the terminals to abut a new Amtrak/Coaster/Trolley station, but the airport refused.
The next best concept is then San Diego's latest plan: an APM providing a one-seat ride between Downtown and the airport with 2-minute frequencies. The airport terminals are on a geographic cul-de-sac, so the only alternative to an APM would be branching off an LRT trunk that already has woefully inadequate frequency.
Yet some people still prefer the airport LRT concept because they believe "unlike LRT, APMs can't be extended past the airport into dense urban neighborhoods" or "the LRT will attract more riders than the APM because it looks cooler than the APM."
Link to San Diego's latest plan:
https://www.sandag.org/-/media/SANDAG/Documents/PDF/projects-and-programs/featured-projects/central-mobility-hub/atc-concept-evaluation-study-appendices-k-s-2023-07-14.pdf#page=212
How it looks is very dubious! I guess it could be both too! LRT could probably work with the right design
Could you explain what is dubious?
Automation allows for a lot of flexibility: increase frequency on extra demand/events, long operating hours, low operating cost etc. That's far more important than vendor flexibility.
The other aspect is that elevated rail can be complex to build while APMs are often much more prefabricated: In North America people know how to build elevated highways and therefore elevated rail isn't built much different, meaning you first build it like a road, then you put tracks on top, then railings on the side and the power system. APMs (and monorails) are usually far more integrated, newer systems like TSB's maglev are all precast/fully integrated. A single vendor can make procurement much easier, too, and therefore reduce complexity, schedule and risk.
So I see APMs, monorails and even urban gondola lifts as a great entry point to automated transit.
They can work well, but vendor flexibility is important. The LAX people mover is costing as much as the Canada Line despite having 0 tunnel and will be locked into a single vendor!
I would attribute that more to TransLink's vs LAWA transit expertise.
Ultimately transit projects should be open to any bidder with any technology so that the provider can make the proper decision based on price, extension flexibility, vendor selection etc.
“Generally, I think transit problems need to be broken down along their major features — capacity, alignment, network structure — and then we can find a particular solution that best matches needs across all of those dimensions.”
Well put. And benefit vs cost is also part of the equation. As you state, if the solution is a people mover, fair enough.
I think its really important to get rid of what are often one dimensional perspectives that look at transit as capacity or speed and nothing else!
Writing this as a Singaporean reader who has experienced all three feeder APM systems here, and regularly use one of them.
APMs are really quite inappropriate especially in the Singapore implementations (and I'm referring to BOTH the Sengkang/Punggol APMs that you highlighted, and the Bukit Panjang one elsewhere), because they really are not the best tool to play the feeder transit role they were designed to serve
First up, their capacity really sucks from the perspective of the individual vehicle, and it's not like they're running these particularly frequently (6 minutes intervals at the peak if your destination is unlucky). The Bukit Panjang system also operates a unidirectional loop outside the peak period, so sometimes taking the bus is actually faster, especially when there do exist multiple buses with a combined frequency better than the 5-minute off-peak interval. For the Sengkang/Punggol systems you highlight, these are located in highly populated towns which are much denser than the average public housing town in Singapore, which makes low-capacity, low-intensity transit like an APM very inappropriate for feeder service. It doesn't help that for Sengkang/Punggol there is branching involved, which means it's really difficult to freely add trains as you wish, since overloading the line simply slows the ride immensely (you can feel it when riding the APM into the central stations -- they have to stop and wait for the pod in front from the other branch to depart first before proceeding), and in the case of the Punggol lines they are a significant distance away from the depot (the pods have to pass through the *entire* Sengkang West Loop, proceed to the Sengkang-Punggol connecting viaduct, then travel the entire length of the Punggol West loop to reach Punggol central station)
The elevated nature of the APM lines makes it cheaper to build, but honestly the amount of infrastructure for it is overblown for such a low-capacity system with stop spacings equivalent to the parallel bus lines (especially for S/P APM). Also, this adds to *access penalty* which is the time taken to travel from street level to the platform. Factoring this in, it makes a feeder bus ride (which does exist for all 3 towns here with feeder APM systems) more preferable than riding the APM, especially if you're travelling to only the town center, say for getting groceries or eating out, and that's purely based on travel time, without considering other factors like comfort and convenience. Even if you're transferring to the MRT and going elsewhere, taking a bus and connecting to the MRT may be a more competitive option than a similar route via the APM systems instead. In Bukit Panjang, residents willingly wait longer for bus 976 which is a full duplication of the APM, as evidenced by its long snaking queues at Choa Chu Kang Bus Interchange in the evening.
Not all APM systems suck, the ones in the Greater China area are pretty decent, although you must take note that they are strictly for low-ridership, low-density applications that are more geared towards expanding catchment. It's just the way the Singapore ones are set up which make people movers a subpar choice for feeder transport here. Politically, they've also been an embarrassment for the government, which had to admit that they were built under pressure.
In any case, I highly encourage you to come visit Singapore and see the system for yourself to decide, and perhaps I could show you around too, my schedule permitting!
I mean I think a lot of the issues in Singapore have less to do with the mode than the implementation - 6 minute headways are bad on a fixed guideway system! IIRC the Bukit Panjang the alignment is truly awful. That being said I am skeptical that buses are really the solution for such extreme density as you see in that part of the city - and I like buses!
I only brought Singapore up as an example because it really did motivate my interest in the topic!
When APMs don't give you a significant capacity advantage over buses, and all the additional downsides that buses don't have, it doesn't make the bus any worse an option for said suburb (which is what HDB towns in Singapore are translated to NA context) frankly...
Some bus routes in these towns (both feeder and trunk routes, if you're wondering) operate as frequently as the APM and I dare say the combined capacity from having multiple buses down the same road, mostly operating double-deck/articulated fleets can outrival that of the APM systems. I tend to think of the APM systems as outvalves for the crush-load demand for existing bus services (which was indeed the case, so much so we had an entire rail line segment that exists really because of this issue), in cases where the option with more infrastructure appears rather overbuilt for what it can deliver.
Case in point: Downtown Line and its light-rail level capacity was built to release demand from Bus 190, so infamous for its crowds it made it to national news in 2013: https://www.youtube.com/watch?v=gY65MF6NwAo
Stage 2 of the DTL which parallels 190 opened in 2015, while they had to launch a second duplicating bus service (also equally frequent last time, but now with heavily reduced service after DTL2 opened) in the meantime.
I wouldn't say the DTL has light rail like capacity - its probably 20,000 PPDPH while most light rail is sub 10,000 PPDPH
No insurance is needed and Gulf Cart can use sidewalks and or Multi Use Paths MUPs
AGT Gulf Cart that holds 10 to 23 people is a good solution for short trips around Stations
And then there is the more extreme length versions of cable hauled people movers, like the new London Luton Airport DART (connecting the airport to the heavy rail train station, replacing a less than ideal bus link). And yes Geoff has a video.
And the DART will be extended further when the airport gets a new terminal (planned into the design). Built by Doppelmayer (and opening was delayed while some challenges with the cable hauling were resolved). Similar tech also used at Birmingham (UK) airport for a similar link, replacing what was the world's first passenger Maglev (short, end of life, and not that quick)
The DART had to be a cable hauled or rubber tyred for cost reasons--there is quite a height difference between rail and terminal area (which are about a 1.5km apart, so quite a gradient), and tunneling plus lifts would have been much more expensive, and steel wheels would have needed a quite indirect route and likely some significant earthworks.
(Luton is also one of a select few UK cities that has a guided Busway, but that's the other side of town, west towards Dunstable)