I have started work again in Melbourne's CBD. The trains during "peak hours" are much busier than when I previously commuted regularly about 6 months ago. It is great to see so many people using public transport, but the system is performing poorly.
The "stopping all station" trains along the Box Hill line used to go direct to Flinders Street, now all trains on the line go through the loop. There are hundreds of people swapping trains (from Loop train to Flinder St and vice versa) for every train that arrives at Richmond. This delays trains departing and obviously inconveniences many people. There has to be a better way.
The Myki ticketing system is a debacle. After costing around $1.2b (way over budget), and possibly up to $1.5b, it has critical performance issues that are impacting commuters, lots of bugs in the billing, and serious problems with the card readers. The review commissioned by the Baillieu government has not been made public, but they have decided to throw good money after bad and try and make it work.
They are now decommissioning the Metcard readers and many stations and forcing commuters to use Myki, so the real issues and failings are becoming very obvious.
Here is a summary of issues I have encountered:
- Some card readers don't read my card, but the one next to it does (malfunction, slows passengers)
- I was charged a Zone 1-2 trip even though I only travelled in Zone 1 (I got a refund)
- I was charged a penalty fare for "not touching off" at Surrey Hills, when I carefully did. I contacted the Myki Call Centre - they told me a faulty reader had immediately touched me on (twice!) when I touched off at it (I got a refund).
- The Myki Call Centre cannot see transactions for the current day and they are not visible via the Myki website. You have to wait for an overnight computer batch run to see them!
- The refunds did not appear in my account. The Call Centre informed me that "they send the refund to every card reader in the system (store and forward!?) because they don't know where I will touch on. When I touch on I then get the credit. So it seems the card stores the value, not they central system. If this is true it is a bizarre and very complex IT architecture.
- The card readers are far too slow. Huge queues form at all exits of busy stations such as Flinders Street and Southern Cross while everyone waits around 1 second (or more) for the device to touch off. This should take milliseconds. The sensitivity of the card readers and the response time is woeful.
- The system would not have passed basic performance and load testing (if this was done) and is now failing user acceptance testing.
- The cause of the very slow response of the card readers is not certain. They may be just substandard, or the network latency might be causing it. There is also a "rumour" that the sensitivity of the readers has been "turned down" to get them to work in close proximity with the existing Metcard readers. There may be some validity to this as it has been announced that many of the Metcard/Myki readers and gates are about to be replaced by new gates with Myki readers only. This will be very expensive. If it fixes the poor read response and high bad read problems, this should have been done BEFORE commuters were forced to use the Myki system
- There is still no Myki solution for tourists or country people who visit Melbourne. This is a basic requirement that should have been met BEFORE any roll out.
So how did this disaster happen? It seems that Transport Minister Peter Batchelor in the previous Labor Government decided they would "develop a world class ticketing system that they could sell all around the world" and ignored the option of buying a system that was already available and working well (such as Brisbane's, or London's Oyster Card or Hong Kong's Octupus card).
Then the government awarded the contract to the lowest tender and the Kamco consortium that knew very little about ticketing systems.
Then Kamco bought cheap and inferior "end of life" reading devices and designed a system that has serious architectural points of failure.
Many of the problems with Myki were evident over a year ago, as I wrote in t
his post about Myki in February 2010. I still think that the Myki experiment should be canned. We should start again with a system that is proven and works, or stop issuing tickets or employ conductors again. Governments should not engage in research and development for this type of infrastructure. They should select systems that are proven, and conduct due diligence on companies and consortiums before awarding them contracts.
There is a whiff about this whole debacle that I just don't like.
Let's have the "open and transparent government" that you promised Premier Baillieu. Please release the Myki review report, and tell me which of the issues I list are fixable.
To top matters off, the Auditor general calls for annual spending of $3 billion to cope with growing demand and fix problems due to chronic under funding. Then Ian Dobbs, the new head of newly formed Public Transport Victoria, rejects the call for $30 billion and states that "The biggest opportunities are in making best use of what we've got at the moment, rather than believing that all the solutions are building, or buying new kit".
He, and members of our government, have obviously not traveled by public transport recently. They probably never do.
I am riding by bike to and from work. It is faster, more reliable and much more pleasant, even taking into account the traffic I have to negotiate.
Links