Oh No not again
Uncle Joe
Tcard: here we go again
Linton Besser Transport Reporter
July 3, 2008
THE dumped Tcard has been revived in a secret cabinet decision that could
make public transport more expensive for many long-distance commuters but,
if successful, should lead to seamless travel on train, bus and ferry.
Less than five months after terminating a $367 million, 10-year deal to
produce a Tcard - sparking a $200 million lawsuit - cabinet has agreed on
another attempt to provide an integrated smartcard for commuters.
Crucially, cabinet's decision also included a requirement that the
Government pursue changes in the way fares are structured. Experts say the
lack of reform of Sydney's complex and antiquated ticketing system was a big
reason the Tcard project failed.
As the Herald revealed last week, the Government's mishandling of the Tcard
has earned Sydney worldwide notoriety and may threaten its ability to find
international partners for transport projects.
In the contract awarded to ERG Limited in 2003, there was no overhaul of the
500 fare permutations that exist across the city's bus, rail and ferry
networks. In Melbourne there are about 54 variants.
The cabinet decision on Monday last week will be seen by some as an
admission that it shares the blame for the ticketing fiasco that has so far
cost the state $95 million. Bureaucrats had repeatedly warned the Government
that fares had to be simplified if the Tcard was to succeed.
It was also a bold move because the jumbled fare system has been seized upon
by ERG in its damages claim.
The call for tenders is likely to be made well before new fares are
finalised. Consultants have been working on a number of scenarios, but the
Herald understands the budget committee has demanded a vastly simplified
fare structure.
The new Tcard will operate as a stored-value card, much like an EFTPOS card,
from which the fare is deducted.
The Transport Minister, John Wankins, told the Herald last night: "The
Government has agreed in principle to proceed with an expression of interest
in the next couple of months to test the market for electronic ticketing
options."
He said the Independent Pricing and Regulatory Tribunal would take a lead
role in developing a fare structure. "Building on IPART's recent CityRail
fares discussion paper, the Government intends to expand IPART's reference
and ask that it advise on the fares that should apply on any new product for
all transport modes."
Any changes in fare structures will be politically difficult. Under almost
every model, long-distance commuters will face larger fare increases than
those who live closer to the city. The impact will be most severe in
marginal Labor-held seats in western Sydney and on the Central Coast.
Last month, the pricing tribunal urged sweeping reform to both the CityRail
network and the way fares are structured. It criticised the substantial
discounts built into weekly tickets and Travelpasses. "The size of the
discount for journeys longer than 35 kilometres is not consistent with
promoting economic efficiency of rail services," its report said.
The Opposition transport spokeswoman, Gladys Berejiklian, said: "The State
Government needs to accept that its cumbersome fare structure and its lack
of political will in fixing it was a huge impediment to the project."
At least five companies are believed to be competing for a Tcard contract
worth several hundred million dollars. They include the defence giant
Thales, Hong Kong's MTR, and Cubic Transportation Systems.
Gettamulla Tupya
> Thales, Hong Kong's MTR, and Cubic Transportation Systems.
>
Isn't MTR in Hong Kong associated with ERG??
Fish Womper
>
>
>Tcard: here we go again
>
>Linton Besser Transport Reporter
>July 3, 2008
>
>
It's beyond me why they wouldn't put the readers on the trains rather
than at station entrances. Then a fare scale could be developed
according to distance travelled over a period of time and travellers
would not need to "tag in". Make it a sliding scale by all means, just
not sliding as much as it does now.
Everyone knows that short haul trips make money (eg Bondi Junction to
Central) but long haul ones don't. How can they, when a weekly Central
to Bondi Junction costs 40% as much as Katoomba to Central although
costing only 10% (and possibly less) of the amount to provide the
service.
fish
Sylvia Else
>
> Crucially, cabinet's decision also included a requirement that the
> Government pursue changes in the way fares are structured. Experts say the
> lack of reform of Sydney's complex and antiquated ticketing system was a big
> reason the Tcard project failed.
Who are the alleged experts who say this? How hard can it be to
implement a set of fare rules in a computer program, even if there are
500 permutations?
Sylvia.
terryc
Search time and data storage all lead to increase compexity and power
requirements.
Sylvia Else
Shouldn't amount to much unless someone tries to implement it as a
monumentally complicated SQL query.
Sylvia.
Kevin Martin
I agree with you Sylvia.
Tcard are whinging about something that existed when they signed the
contract. Surely the information on the number of permutations must of
been part of the preliminary information available to all potential
tenderers (is this the right term) right at the start?
Perhaps they didn't bother reading it properly? Or didn't even try the
current system out. How hard would it be to send 50 people out for a few
days and buy various tickets for a variety of destinations and try to
find out what they were trying to improve on?
A petty cash proposition compared to the value of any new contract. They
probably tried to do it based on supplied info or (worse) computer
modelling.
For what its worth, the initial tests on Melbourne's Myki system is
showing similar lack of success during trials and is already 2 years
behind. Melbourne has already significantly simplified the system by
dropping Zone 3.
These changes were not done for Myki's benefit, but for political
reasons. The Liberal Opposition promised it at the last state election,
the Government matched and has since carried it out.
Kevin Martin
David Johnson
I agree. It shouldn't be that hard. I think the problem is that
currently, the end user chooses the ticket that best suits them, but
they want the new system to automatically select the best option.
For example, someone at Penrith might select a Purple travelpass. But
under the new system, they tag on at Penrith, then off at Meadowbank. A
charge is deducted. They then tag on a ferry at Meadowbank wharf and
then tag off at McMahons Point. Another charge is deducted. They then
tag on a bus at McMahons Point and off again at Chatswood. The system
then needs to have the intelligence to realise that a daytripper is now
cheaper and deduct accordingly. This would continue for a few days
until the system realises that the purple travelpass is now cheaper and
stop deducting when it gets to that level.
The trick to the whole system is comparing all the available ticketing
options to ensure that the end user gets the cheapest option, which
would change as time progresses.
--
---
Every time I make my mark, someone paints the wall.
---
David Johnson
usenet.at.trainman.id.au
http://www.trainman.id.au
------------------------------------
These comments are made in a private
capacity and do not represent the
official view of RailCorp.
Sylvia Else
Really, of all the things that could go wrong with the implementation of
such a system, calculating the fare is the one I think I'd worry least
about. I'd be much more concerned about the coal-face harware and its
connection to the central system, including failure tolerance.
Sylvia.
terryc
> Really, of all the things that could go wrong with the implementation of
> such a system, calculating the fare is the one I think I'd worry least
> about. I'd be much more concerned about the coal-face harware and its
> connection to the central system, including failure tolerance.
AFAIUI, that was where it failed.
The new system sounds like it is going to overcome that by.....
requiring you to preload a credit card (woopeee). just what I need,
anther bit of plastic with some of my money attached.
C. Dewick
>On Thu, 3 Jul 2008 06:54:53 +1000, "Uncle Joe" <be...@over.com> wrote:
>>
>>
>>Tcard: here we go again
>>
>>Linton Besser Transport Reporter
>>July 3, 2008
>>
>>
>It's beyond me why they wouldn't put the readers on the trains rather
>than at station entrances. Then a fare scale could be developed
>according to distance travelled over a period of time and travellers
>would not need to "tag in". Make it a sliding scale by all means, just
>not sliding as much as it does now.
Putting 'mobile' devices on trains requires 100 percent 'live' wireless
technology to get it working accurately and properly, but that doesn't exist
at present. Even the existing CityRail train radio system does not have 100
percent coverage. There are some key places where drop-outs are frequent and
where signal coverage falls off to a point where an actual conversation
could not be carried out, even though the radio on the train can get enough
signal from the nearest base-station.
>Everyone knows that short haul trips make money (eg Bondi Junction to
>Central) but long haul ones don't. How can they, when a weekly Central
>to Bondi Junction costs 40% as much as Katoomba to Central although
>costing only 10% (and possibly less) of the amount to provide the
>service.
This is why the fare-structure is so badly skewed IMHO.
Craig.
--
Craig Dewick - HO-Scale Railway Modeller and Professional Train Manager at
http://lios.apana.org.au/~craigd or cra...@lios.apana.org.au if you're game!
RailZone Australia - http://www.railzone.org - No Fundies! No RailCorp CodeCon!
http://lios.apana.org.au/mailman/listinfo/aus_rail_safety for Oz Rail Safety
Matthew Geier
> x@x.x (Fish Womper) writes:
>
>>On Thu, 3 Jul 2008 06:54:53 +1000, "Uncle Joe" <be...@over.com> wrote:
>
>
>>>
>>>Tcard: here we go again
>>>
>>>Linton Besser Transport Reporter
>>>July 3, 2008
>>>
>>>
>>It's beyond me why they wouldn't put the readers on the trains rather
>>than at station entrances.
What's the difference ?. You can only join/leave the train at a station,
so why add a reader to all many 1000's of train doors in the system
instead of the much lower number of station entrances.
The only thing i'd possible change is the range of the readers. Why 'tap-
n-go', when the reader field could be designed to read anything it finds
in a several square metre radius. (Although this stops the carrying of
multiple cards, which their may be legitimate reasons to do - say a work
card and a private one).
> Putting 'mobile' devices on trains requires 100 percent 'live' wireless
> technology to get it working accurately and properly,
The smart card systems currently deployed do not require 'live' access
to the back end. That's the major 'smart' bit about it - the ability to
run disconnected. (on a bus mainly).
Data is written to the card AND stored by the vehicle reader. In daily
use the system uses the data on the card to work out what's happening 'on
the spot'. On a bus for example, when the bus returns to the depot in the
evening, the data is downloaded and transmitted to the back end
processing facility that tallies it all up.
At the end of the day, the result the back end comes up with should
match the data stored in the users card. It will have multiple data
points collected during the day of the users card. Every thing should
match at all times.
If it doesn't, the card will be flagged for further investigation.
(possible card tampering fraud for example, although 'computer error' is
possible too).
David Bennetts
system:
Mention was made of Hong Kong MTR being a prospective tenderer. I
successfully used the Octopus card there, it worked well, you could even use
it to make vending machine purchases at some locations.
The bus fare system is interesting in Hong Kong, the fare reduces if you get
on toward the bus destination.
Here's an example:
http://tinyurl.com/5f8k3f
Regards
David Bennetts
Matthew Geier
> The trick to the whole system is comparing all the available ticketing
> options to ensure that the end user gets the cheapest option, which
> would change as time progresses.
As far as I know, no one has attempted that past converting a number
individual trips into a day pass. Anything further, the user is expected
to load a weekly or periodical 'product' into their card.
ERG did however agree to implement the unimplementable. They shouldn't
have even tried.
And as far as I can determine, London Oyster 'fare capping' isn't done
by the back end, as it works on the buses. Each 'station' (or vehicle)
controller must have the logic in it, other wise if the system couldn't
'phone home' it could deduct the wrong fare and the user might
unexpectedly run out of credit in the middle of a multi-stage trip.
I still can't figure out why the Transport Dept/ERG didn't start with
something simple, like weeklies and travel 10s and add features
incrementally as they got each product bedded down. Instead they tried to
do every thing at once and spectacularly failed to get anything at all to
work.
Dik T. Winter
...
> The only thing i'd possible change is the range of the readers. Why 'tap-
> n-go', when the reader field could be designed to read anything it finds
> in a several square metre radius. (Although this stops the carrying of
> multiple cards, which their may be legitimate reasons to do - say a work
> card and a private one).
Overhere in Europe many people have multiple RFID cards in their pockets.
And quite legitimate. In the Netherlands for instance the ID-card. It
interferes with the public transport card.
--
dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland, +31205924131
home: bovenover 215, 1025 jn amsterdam, nederland; http://www.cwi.nl/~dik/
Paul Westcott
> C. Dewick wrote:
>> Fish Womper writes:
>>> Uncle Joe" wrote:
>>
>>>> Tcard: here we go again
>>>>
>>>> Linton Besser Transport Reporter
>>>> July 3, 2008
>>>>
>>> It's beyond me why they wouldn't put the readers on the trains rather
>>> than at station entrances.
>
> What's the difference ?. You can only join/leave the train at a station,
> so why add a reader to all many 1000's of train doors in the system
> instead of the much lower number of station entrances.
The reason that Myki is being trialled first on Geelong buses is that
the technology required to handle the ticket information emanating from
a mobile source is the hardest to get right.
Kamco have been testing for six months in Geelong and have most recently
reported a 90% success rate.
Paul Westcott
john...@nospam.com.au
Cityrail's ticket machines have been satisfactory for me, but like ticket
machines around the world, they are a nightmare for visitors.
Other jurisdictions view public transport differently and charge a nominal
fee for access rather than distance travelled.
Alan in Darwin
> Cityrail's ticket machines have been satisfactory for me, but like ticket
> machines around the world, they are a nightmare for visitors.
>
> Other jurisdictions view public transport differently and charge a nominal
> fee for access rather than distance travelled.
>
the Northern Territory (Yes, I said the NT). Alice Spring bus service is
a part of DarwinBus, so if you time your plane trip right, you can get
from Fannie Bay to Alice Springs for $2.00 plus air fare.
Regards
Alan in Darwin
ten
>
> The only thing i'd possible change is the range of the readers. Why 'tap-
> n-go', when the reader field could be designed to read anything it finds
> in a several square metre radius. (Although this stops the carrying of
> multiple cards, which their may be legitimate reasons to do - say a work
> card and a private one).
>
>
Not a good idea at some stations where the T-Card readers were located
in public thoroughfares and could thus be activated when not required to be.
An example of this is Berala where the readers are located in the narrow
subway that links the north and south sides of the suburb. With your
range of 1 sq metre it would be almost impossible to walk through the
subway without activating one of the readers on either side.
john...@nospam.com.au
Ignorance + Hubris is the explanation. They read too many management textbooks
and believed them.
Sylvia Else
ERG would have agreed to whatever the government wanted. The only issue
would have been the price. If they'd said it was too hard to do all at
once, the government would simply have gone to another vendor who would
undertake to deliver it in one hit.
It's an unfortunate fact of life in the IT business that telling
potential customers the truth is a sure way to avoid them becoming
actual customers. The usual approach is to be less than honest about the
timescale and features, and leave the customer having either to walk
away from something they need, or wear the time/functionality detriments.
It's rare indeed for a supplier to get sued. If it happened more often,
suppliers might not feel such a great need to lie.
Sylvia.
David Johnson
> x@x.x (Fish Womper) writes:
>
>> On Thu, 3 Jul 2008 06:54:53 +1000, "Uncle Joe" <be...@over.com> wrote:
>
>>>
>>> Tcard: here we go again
>>>
>>> Linton Besser Transport Reporter
>>> July 3, 2008
>>>
>>>
>> It's beyond me why they wouldn't put the readers on the trains rather
>> than at station entrances. Then a fare scale could be developed
>> according to distance travelled over a period of time and travellers
>> would not need to "tag in". Make it a sliding scale by all means, just
>> not sliding as much as it does now.
>
> Putting 'mobile' devices on trains requires 100 percent 'live' wireless
> technology to get it working accurately and properly, but that doesn't exist
> at present. Even the existing CityRail train radio system does not have 100
> percent coverage. There are some key places where drop-outs are frequent and
> where signal coverage falls off to a point where an actual conversation
> could not be carried out, even though the radio on the train can get enough
> signal from the nearest base-station.
There is a move to coincide with the introduction of the PPP trains to
have a permanent data link back to base using GSM technology.
--
---
Captain please, not in front of the Klingons.
Matthew Geier
>> Putting 'mobile' devices on trains requires 100 percent 'live' wireless
>> technology to get it working accurately and properly, but that doesn't
>> exist at present. Even the existing CityRail train radio system does
>> not have 100 percent coverage. There are some key places where
>> drop-outs are frequent and where signal coverage falls off to a point
>> where an actual conversation could not be carried out, even though the
>> radio on the train can get enough signal from the nearest base-station.
>
> There is a move to coincide with the introduction of the PPP trains to
> have a permanent data link back to base using GSM technology.
There wouldn't be enough bandwidth on a mobile channel to deal with
several hundred transactions in a sort period of time - think of the
crush alighting and boarding at Town Hall in the peak.
But anyway, 'smart card' systems DON'T require a live link to the back
end. That's the 'smart' bit of the system - the secure data store on the
card itself.
You really only need intermittent connections to download collected
data. You could do that with 802.11 wireless in 'terminal' stations where
trains are stopped for a little while. They come into wireless range and
do a download.
The guys I work with have a vehicle tracking system like that - it's
intended for areas with poor to no phone coverage. The vehicles up load
their logs when the come back into 802.11 radio range automatically.