Myki Shmyki (Because sometimes a catchy name just isn’t enough)
December 27, 2007
So there is some news about the implementation and roll out of the Myki public transport ticketing system here in Melbourne, with the US based company Kamco asking for more time and money to complete the project and the Victorian Premier John Brumby, considering the matter.
I’m not entirely surprised. The contractor might have bid higher that their abilities/resources, or agreed to an unreasonably short delivery time-line, or the government agency handling the tender might be operating in such a way that Kamco cannot keep it’s agreed targets. It is a simple and unavoidable property of the universe that software projects creep, and without inside information it is difficult to determine just where the blame might lie.
To be honest, the project delivery issues and accusations of irregularity in the tendering process have never really outweighed my basic opinion that the entire concept behind the new ticketing system is bogus.
Let me explain – in Melbourne we have an extensive public transport system consisting of a combination of rail, bus and tram services. This system is based around time based trips over two geographical zones. For instance, you can purchases a base fare (priced over a scale from standard users to children, and concession fares etc) and it will give two or three hours usage of the system within a zone. You can also purchase tickets that cover an entire day (or as in my case a ticket providing access for the entire year). All of this is administered via a simple disposable magnetic strip card that stores the relevant information electronically and also has human readable expiry information printed on the face of the card on validation.
The new system is based around an RFID smart card that is to be scanned on entrance and egress of each vehicle, at each stage of a trip. The information is stored, and when your trip is completed the various stages of your trip are aggregated and a charge is levied, based on some unknown criteria.
I see several basic problems with this:
- It is more complicated than the old system. I’m not talking about the complexity of the solution, but rather complexity to the user. The commuter has to remember to scan their key twice for each leg of their trip, and be physically able to do so. I know that I tend to run on auto-pilot in the mornings and it would be very easy for me to miss a scan. While you are indeed currently required to validate your card on each vehicle, this is only once and if you are unable to, or forget, you are still traveling with a valid ticket. This raises the second point:
- The failure case is disastrous to the user. What happens if you are carrying two cards on your trip and both are accidentally validated? What if you scan in at the front doors of a tram, and scan out by the middle doors, but one of the scanners is broken? How will your trip be charged? And how will you even know that something has gone wrong? People tend to get off public transport in groups, so how do you ensure that it was your key that made the scanner beep? Even if you do suspect that something might have gone wrong, will the system be transparent and traceable enough that you can find erroneous charges, and flexible enough that you can challenge them? The onus of detecting and handling failure cases is shifting toward the user, the person who is least capable of detecting or handling them. Madness.
I’m not against the new system because it’s new. But I think that any system that is costing that much money to provide the user with a more complicated and difficult experience has some serious issues.
Sure, change the system – but make it better. Change for the sake of change is just nuts.