Green Line LRT | ?m | ?s | Calgary Transit

Go Elevated or try for Underground?

  • Work with the province and go with the Elevated option

    Votes: 52 75.4%
  • Try another approach and go for Underground option

    Votes: 14 20.3%
  • Cancel it altogether

    Votes: 1 1.4%
  • Go with a BRT solution

    Votes: 2 2.9%

  • Total voters
    69
The gates also default to open, and close only if it detects someone passing without scanning. It really speeds up the lines during commute times.
This is a great example of small innovations that we need more of here - seemingly minor things that don't necessarily cost a whole lot but make an incrementally important distance to speed transit up. This type of thing only happens when transit agency cultures are based on a relentless focus on transit efficiency and customer speed improvements.
 
If you've used Compass/Presto, which are mostly fine now, but the Japanese gates are very fast. I'm not sure how they do it but their response time is unbelievably fast
It really depends on whether a record is being updated in one place (on the card alone) or two places (card, reader/offline datacentre) or three places (card, reader, live data centre) and whether you trust the record on the card alone, or if you wait to reconcile with the stored record either in the reader, or back in the data centre. In many systems, the reader may also have data that it wants to push to the card, such as a new balance from an auto-reload the night before, a new balance due to reconciliation and fare-capping, or a new pass product. So how good of a computer are you putting in each reader, with how good of a data connection? How much trust do you extend to users in case something goes wrong? Is it worth it to extend credit to wait for that nighttime update if someone has an auto-reload flag on the account, or if the balance drops during the course of a day, do you reject someone at a fare gate? Do you do a soft warning for one day if something isn't looking right then reject on the second day?

It is really easy for a transit agency that doesn't understand the tradeoffs on this IT system to put in requirements that are very hard to deliver, or that degrade the ability to meet other requirements.
 
This is a great example of small innovations that we need more of here - seemingly minor things that don't necessarily cost a whole lot but make an incrementally important distance to speed transit up. This type of thing only happens when transit agency cultures are based on a relentless focus on transit efficiency and customer speed improvements.
For this one specifically you'd need a very law abiding population, which isn't always the case with North American transit. But agreed with the general point. I noticed in Seoul, their transit map has different colour station circles based on door opening position. Yellow circle opens on the right, black circle opens on the left. It helps when the train is crowded to know ahead of time which way the door is opening. Not sure why that isn't the case for Toronto/Vancouver/Calgary where we have one train per line so almost no situation where it'll open on a different side.
 
It really depends on whether a record is being updated in one place (on the card alone) or two places (card, reader/offline datacentre) or three places (card, reader, live data centre) and whether you trust the record on the card alone, or if you wait to reconcile with the stored record either in the reader, or back in the data centre. In many systems, the reader may also have data that it wants to push to the card, such as a new balance from an auto-reload the night before, a new balance due to reconciliation and fare-capping, or a new pass product. So how good of a computer are you putting in each reader, with how good of a data connection? How much trust do you extend to users in case something goes wrong? Is it worth it to extend credit to wait for that nighttime update if someone has an auto-reload flag on the account, or if the balance drops during the course of a day, do you reject someone at a fare gate? Do you do a soft warning for one day if something isn't looking right then reject on the second day?

It is really easy for a transit agency that doesn't understand the tradeoffs on this IT system to put in requirements that are very hard to deliver, or that degrade the ability to meet other requirements.
I can't speak to the technical details, but form a user perspective, I'm able to reload the card from my iPhone and use it immediately, there is no loading "wait period". There's custom fare zones with different prices and multiple operators (Tokyo Subway, Toei Subway, JR East, and compatible with every transit agency in the country). When I enter the station, my Apple Wallet gives me a notification "trip in progress" and when I exit, I see the charges on the screen on the machine and on my phone. This all happens in less than 1 second (they literally advertise it on the reader, that IC is sub 1 second lol). Not to mention the card also work in many convenience stores, fast food restaurants, processing transactions in a country of 100 million plus.

I recognize technical challenges, but system like Presto definitely have a design and execution problem. It is impossible that a transit system with a fixed fare is more technically challenging than these foreign systems like Tokyo, Hong Kong, London, etc. They picked a terrible contractor (Accenture) instead of a proven system like Cubic then kept complaining about how it's just such a complicated system.
 
is more technically challenging
Classic procurement issues for Presto, over spec'd, low bid, focused on capital costs not lifecycle costs (even worse, the province was paying for the IT capital, and the municipalities pay for the operating), so end up with huge mismatch and lack of a single decider. For Toronto especially things were missed--the old system was mostly mechanical, and they had to run electricity and data connections to all the turnstiles as only a small number of turnstiles had had a very basic retrofit to run a magnetic strip reader for monthly passes.

The larger systems counterintuitively are easier to do, since you have a higher volume of cash flowing that you can fund the system with. They also started with simple systems and built up, having used stored value cards for (it must be, similar to Paris, which introduced magnetic stored value and pass cards in 1975) close to half a century now, adding features relatively slowly. A key place to start was stored value, and trip tracking, to do distance based fares. Build the system up with as much being done on the card itself.
 
Last edited:
I can't speak to the technical details, but form a user perspective, I'm able to reload the card from my iPhone and use it immediately, there is no loading "wait period". There's custom fare zones with different prices and multiple operators (Tokyo Subway, Toei Subway, JR East, and compatible with every transit agency in the country). When I enter the station, my Apple Wallet gives me a notification "trip in progress" and when I exit, I see the charges on the screen on the machine and on my phone. This all happens in less than 1 second (they literally advertise it on the reader, that IC is sub 1 second lol). Not to mention the card also work in many convenience stores, fast food restaurants, processing transactions in a country of 100 million plus.

I recognize technical challenges, but system like Presto definitely have a design and execution problem. It is impossible that a transit system with a fixed fare is more technically challenging than these foreign systems like Tokyo, Hong Kong, London, etc. They picked a terrible contractor (Accenture) instead of a proven system like Cubic then kept complaining about how it's just such a complicated system.
The QR code system used in Calgary is far superior to a reloadable card as it doesn't require access to kiosks to distribute or reload cards, doesn't impose a delay on waiting for funds to show up on a card and doesn't have to deal with overdrawn cards. Most importantly, it uses low cost hardware that is easy to update so it isn't frozen in time.

As long as the Calgary region has a single transit agency, the need to track origins and destinations is low as there is no need to calculate revenue shares.
 
Last edited:
All the cards are interchangeable, there is a very, very small list of operators that only take 1. All can be loaded from your credit card directly within Apple Pay. If you've used Compass/Presto, which are mostly fine now, but the Japanese gates are very fast. I'm not sure how they do it but their response time is unbelievably fast, I had to do double takes sometimes to make sure it actually scanned. The gates also default to open, and close only if it detects someone passing without scanning. It really speeds up the lines during commute times.
As darwink said it depends on how many places you're sending data during a tap. For Suica, the response time is 0.2 seconds, which is achieved by storing information on the card itself while the gate handles the processing locally. The gate then sends and receives information with servers at 1 and 24-hour intervals (also allowing readers onboard buses to be just as fast). On top of that the read distance is a small radius around the reader so the motion of tapping the card (~0.52s) is part of the response time, essentially allowing the entire process to be instant. When you "reload" the card on your phone or a fare machine you genuinely are reloading it, not just topping up a balance on a server. Fraud is prevented by the servers checking the data and if discrepancies are detected a card can be automatically disabled.
Screenshot_20250324-164635.png

This allows the system architecture to be decentralized and avoid cascading or system-wide failures. In the event of a failure gates and various fare machines can store data for up to 3 days and once the communication or system failure is fixed backlogged data is uploaded and processed.

1742886033136.png

I believe MyFare does processing locally which is why you can validate the QR tickets offline and the response time is quite fast (if you nail the scan). If anyone is interested: Development of Suica Autonomous Decentralized IC Card Ticket System
 

Attachments

  • Screenshot_20250324-164635.png
    Screenshot_20250324-164635.png
    75.5 KB · Views: 6
Last edited:
Oh no!

The incredibly beautiful and walkable 10th Avenue will be completely ruined for those poor businesses. The view of the 1960's concrete parkade might be obstructed. Plus...you could have 20,000 - 40,000 C-train passengers per day having easy access to businesses that could drive away the regular customers from the Mustard Seed.
There are a good amount of popular bars and restaurants along the route. I get why they're pissed. Bottlescrew Bills in particular is gonna get fcked by this. Also gonna negatively impact Barbarella for sure.
 
I can't wait until I can just use the app with auto reload and fare capping.
Funny thing with fare capping and cross agency trips. Presto hasn't figured out how to do it using machine logic within their system. So they run a manual report and fix the accounts every day. These things are hard!
 
When you "reload" the card on your phone or a fare machine you genuinely are reloading it
This is what makes the system work. It doesn't (or didn't) have accounts with cloud account integration, and features like automatic top-ups in the background, protection against card loss (being able to move a pass product or balance between cards using the cloud), moving balances between cards without a specialized card reader/writer.

They recognized a limitation of their system and made the system as functional as possible while accepting the limitation. In North America, features get added to IT projects like this without an understanding of how it fundamentally changes how the system will need to work.
 
There are a good amount of popular bars and restaurants along the route. I get why they're pissed. Bottlescrew Bills in particular is gonna get fcked by this. Also gonna negatively impact Barbarella for sure.
Construction will suck for Bottlescrew Bills but I don't actually think the line will rob them of anything that Gallery and Tenth wouldn't, namely sunlight. There will of course be the noise but I do think that should be quite minimal as the noise will be more of an issue for the... checks notes... above ground parking. I will concede it is a different story for Gallery and Tenth.
1742916809860.png


Barbarella is in the same situation as Bottlescrew Bills, what are they losing that Bankers hall hasn't already taken away?
1742916970033.png


I think there is an opportunity to do something that could be an asset under the track too that could make that space quite attractive. Imagine a multi-use pathway under the track that connects Victoria Park right into the downtown. You could do a spiral ramp in the parking lot they'll have to buy on 10th and 1st that allows the multi-use pathway to get over the existing rail corridor and winds down where the existing ramp to the 10th Ave parkade is. Not to mention being able to integrate the line into the +15 network. This is what I want them to be concentrating on; seeing this alignment as an opportunity to build the best elevated line they can.
 
The QR code system used in Calgary is far superior to a reloadable card as it doesn't required access to kiosks to distribute or reload cards, doesn't impose a delay on waiting for funds to show up on a card and doesn't have to deal with overdrawn cards. Most importantly, it use low cost hardware that is easy to update so it isn't frozen in time.

As long as the Calgary region has a single transit agency, the need to track origins and destinations is low as there is no need to calculate revenue shares.
For a system the scale of Calgary, it probably doesn't make sense to run a reloadable card with all the associated infrastructure costs. The need to track origin and destination isn't just a cross-agency thing. Their networks are very large, so one could travel 2 stations vs 25, which should probably cost different amounts.
I can't wait until I can just use the app with auto reload and fare capping. Would never think about it again and just tap. Maybe one improvement would be adding support for NFC from the phone rather than QR only
The ticket system works well. I'm curious if MyFare will eventually have a "card" feature where you get a QR code linked to a credit/debit card or account balance and scanning that will deduct an amount for a ride, a transfer, etc. automatically It probably can't be physical for security reasons (card with a printed QR), but available on the app.
 
I use Clipper a lot in the Bay Area and it is a disaster. Kiosks are out of service or won't read credit cards, the tap terminals are broken, adding funds online imposes a several hour lag, transferring balances between cards or to Google Wallet seems to fail more often than it works, leading to time wasting Customer Service calls. I had one incident where the tap off was broken at Berryessa BART so I couldn't tap out. An employee was there to manually let people out and hand them vouchers to credit back the fares as the system would charge a maximum because it would think you didn't tap out. It ended up consuming twenty some dollars off my card and BART employees didn't know how to process the voucher so I gave up. Smart cards are a 90's technology. Back then I worked on a smart card ecash system that stored balances on a card for use in vending machines, payphones, parking meters etc. with the intent that it would expand to other vendors. It was a joint venture between Nortel, MasterCard and the big banks. Nortel made the hardware that read the cards. The banks walked away from it because customers kept losing cards and maintaining all of that custom hardware was too challenging.
 
It makes me wonder about tap for credit and debit cards. Prepandemic, the cards would have a certain balance on it that would be trusted for tap, a certain dollar amount, or number of transactions before the card would refuse the tap. I know some cards massively increased their tap limit during the pandemic.

I do appreciate being able to use tap when abroad. Singapore supports credit cards, same with New York. Chicago has for a decade.

The question I would raise is, is it just cheaper to eat the fare loss if someone cannot pay on boarding, versus building this infrastructure for buses? The fare system on the LRT is pretty easy.

As long as there is good wifi at the airport for loading the fare app, plus a fare vending machine.
 

Back
Top