Metrolinx: Presto Fare Card

[Tuscani01]

That wouldn't work in the case of short turns, scheduled or unscheduled.

Yea, noticed that in Nfitz's response afterwards.

A time based transfer could be the big push Metrolinx needs to get people in Toronto switching to Presto. I really hope that is the case!

 

[Megaton327]

That wouldn't work in the case of short turns, scheduled or unscheduled.

I suspect they will do one of two things with short turns: A) say you don't need to tap on the second streetcar, as fare inspectors/the transfers to a subway station would likely be able to work fine with your original tap, or B) have drivers give paper transfers to everyone getting off the car.

Alternatively, yes, it might be a simple timed transfer. It will be very, very easy to figure this out definitively as soon as the readers go live.

 

[rbt]

Not sure how you block transferring to same route - that's a short turn.

If it's in the backend (Presto V1 had rules coded into the machines) then disallow/charge upfront and pop up exception events to staff for review and possibly refund when > X number of people do it. A 3 to 5 person trigger might be enough.

When GO has equipment leave service today they move you to the new vehicle and skip the tap-in/out process of the new vehicle (effectively a free that remaining piece of the trip).

 

[nfitz]

If it's in the backend (Presto V1 had rules coded into the machines) then disallow/charge upfront and pop up exception events to staff for review and possibly refund when > X number of people do it. A 3 to 5 person trigger might be enough.

I've been short-turned enough times, when I'm the only one, or one of two people who bother waiting for the next car, rather than just walking home. Certainly late at night, there's no guarantee there are multiple people effected.

 

[rbt]

I've been short-turned enough times, when I'm the only one, or one of two people who bother waiting for the next car, rather than just walking home. Certainly late at night, there's no guarantee there are multiple people effected.

Sure, but after there's a hand-full of overrides you make it a permanently allowed transfer in the system; and the unordinary or unexpected but recurring situation becomes a part of the rule set after a small number of events.

I'm just saying there are very simple monitoring policies that could be put into place that would rapidly find these oddities that the rule makers did not originally consider.

A "rare event" or "change" identification mechanism for a dataset like this would probably table about 2 days work for someone skilled in SQL and R (look for tap combinations with a 1 sigma signal above or below normal) and maybe another couple days for the UI guys to make it pretty. Handling them in a click-able way will take the bulk of the time but you add to these systems over time and often even automate handling certain exceptions.


It's not uncommon in ecommerce to use these techniques to find pricing errors (sudden sales spike in product X) or even competitor sales (sudden decreased referrals from 3rd party website Y for product X) and fix or respond to the challenge a few minutes after it exists.

 

[nfitz]

Sure, but after there's a hand-full of overrides you make it a permanently allowed transfer in the system; and the unordinary or unexpected but recurring situation becomes a part of the rule set after a small number of events.

Might work - but one stop from that transfer location, I or my wife routinely get off the streetcar go pick up/drop off child at school, and pay another fare. Presumably they won't have the location identification restricted to the exact stop, or else they'll have other issues.

I'm just saying there are very simple monitoring policies that could be put into place that would rapidly find these oddities that the rule makers did not originally consider.

And do you anticipate Metrolinx/Accenture having the interest or ability in fine-tuning it to this degree?

We are all theorizing. It's going to be interesting to watch how this rolls out.

 

[rbt]

And do you anticipate Metrolinx/Accenture having the interest or ability in fine-tuning it to this degree?

Well, if they hired anybody who's done anything in big retail or ecommerce over the last decade they would.

It's not exactly new. This kind of thing was the main reason to get a SAP system in place whih have been in enterprise companies for decades. Walmart took it to the next level in the early 90's by integrating sales at register with security camera feeds (follow customer flow around the store and see what they actually bought; adjust store design to accordingly to increase revenue). Now if you don't do it you're dead; Hudson's Bay is a great local/recent example magically turned around with very little actual change. Minor details have massive impact financially.

Accenture won't do anything without being told; they make their money in contract changes when the purchaser realizes they've missed something; like Presto V2 killing the long dated before it was even tendered Presto V1 design.

Metrolinx and TTC should have at least one person with some commerce data-management experience. If not that, they at very least experience with the successful fare-card platforms (Octopus or Oyster). London does a ton of this kind of thing with their Oyster data; it drives their entire fare-fraud division, ensure 3rd party contracts are being adhered to for service delivery, among other things.

The inhouse SQL/R implementation is becoming more common today due to substantial cost savings and improved flexibility over prior SAP style platforms.

So yes, I think Metrolinx could be found incompetent on a government inquiry scale for not having and using detailed data analysis on their Presto tap data; but given the existence of Presto V1 I wouldn't be surprised if they didn't.


I don't think TTC has much of a choice regardless of how transfers actually get implemented. Route management, fare-fraud detection, among other things will require 95% of what I described and pay for itself several times over.

 

[nfitz]

Brad is tweeting <s>that you won't tap on the second or third streetcar of your journey, only the first one. And if going from subway to bus (in the fare-paid area) to streetcar, you just tap on at the subway, not when you transfer to the streetcar from the bus.</s> Edit - ignore this ... he answered a "you don't tap" with a Yes, meaning Yes you do tap, ... the joys of double negatives. They don't seem to have thought out short-turns.

He notes there is some flexibility and time restrictions in the transfer to subway (can handle streetcars going to different subway stations, etc.) but will be some hard restrictions - like not tapping back onto the same route.

Should be interesting.

Are there any reasonable trips that involving entering two subway stations? One I've done when travelling from Davisville to Dupont station, the most direct and fastest trip is taking the subway to St. Clair, 512 to St. Clair West, and then the subway the other direction to Dupont. But there's no need to tap once you enter the station. Are there any examples you do?

 

[Napoleon]

Well in the short-turn example on a PoP route, there's no reason to tap a second time. You already have a proof of payment for the streetcar route you're on.

The one thing I'm a little bit worried about is what happens when there's a change of plans that causes me to need to take a bus. I've had times when I'm on a streetcar route and something happens such that the route is diverting or an accident or something. For example, if I'm at College and Dufferin and heading to College and Yonge, if there's a problem with the streetcar, I might just decide to take the dufferin bus (as awful as it is) up to the Bloor line and reach my destination that way. If you haven't gotten a paper transfer when you first tapped, that's a bit confusing. That actually brings up another confusing bit of the transfer logic...what happens when you use a bus to go between a streetcar and a subway station. I imagine you should show the paper transfer when you get to the subway?

 

[dowlingm]

According to the Metrolinx report posted on stevemunro.ca, they are partnering with Gateway Newsagents in TTC stations to sell Presto cards...

 

[Jonny5]

Are there any reasonable trips that involving entering two subway stations? One I've done when travelling from Davisville to Dupont station, the most direct and fastest trip is taking the subway to St. Clair, 512 to St. Clair West, and then the subway the other direction to Dupont. But there's no need to tap once you enter the station. Are there any examples you do?

This is perhaps a whole other scenario, but some people take the subway to Union, transfer to the GO Train and ride to Kennedy, then transfer to the subway/RT there. I have done it perhaps 10 times in the past 15 years.

 

[Jonny5]

Are there any reasonable trips that involving entering two subway stations? One I've done when travelling from Davisville to Dupont station, the most direct and fastest trip is taking the subway to St. Clair, 512 to St. Clair West, and then the subway the other direction to Dupont. But there's no need to tap once you enter the station. Are there any examples you do?

This is perhaps a whole other scenario, but there are people who take the subway to Union, transfer to the GO Train and ride to Kennedy, then transfer to the subway/RT there.

 

[nfitz]

This is perhaps a whole other scenario, but some people take the subway to Union, transfer to the GO Train and ride to Kennedy, then transfer to the subway/RT there. I have done it perhaps 10 times in the past 15 years.

That's a good one. If at King station. In fact, I've walked to Union, caught the GO to Kipling, and then taken the Airport Rocket before - saves about 30 minutes in rush hour! If my trip started at King Station, I might take subway - would be two taps into subway station to catch bus!

I guess the question is Presto supporting TTC Times Two transfers.

 

[ksun]

It's always been allowed. But it does, of course, cost an extra fare under the current rules.

That would require a policy change - one that the board has cleary voted not to do. And of course $20 million. Where do we get that? We could raise all fares a nickel.

Metrolinx committed in the agreement with TTC that they could and would the current TTC transfer rules. If they can't meet their contractual requirements, then perhaps they should be coughing up the $20 million a year.

I think you are right, that they've implemented a time-based transfer. And I'm guessing, they don't want to say that very loudly. Do they think that users won't notice very quickly!?!

Not sure how you block transferring to same route - that's a short turn.

Sometimes I've even transferred to vehicle going in the opposite direction - perfectly legally. For example if taking westbound 506 from Coxwell to Dundas and Parliament, I often change at Broadview to a 505 if one is coming; but once I do so, I could just as easily jump on a 506 heading west, if it's short-turning at Broadview (as they often to), and then continue to Dundas and Parliament. A valid 506 west to 506 east transfer.

Charging a high fare to long distance riders will bring more than just $20 million. Makes perfect sense, works in many cities, all problem solved. Unless one will whine about "punishing poor people" again as if those who live at Bayview and Lawrence or those who transfer from Richmond hill are really poorer. I know more than a couple of UofT students who walk 30 minutes to campus in the winter because Ttc is too expensive.

 

[LNahid2000]

I wonder how Presto will work with the 512 St. Clair, which operates on a 2 hour transfer. I use that route quite often since it's nice to be able to grab some groceries and go back home on one fare.