It's taken 10 years, but finally the penny has dropped in the dark recesses of Auckland Transport's brain box.
After a decade of bus passenger wailing, the public transport provider has conceded its $24.3 million "real-time" bus arrival indicator system is a disaster. It lies. It misleads. Instead of helping passengers, it just drives us crazy.
Hidden away in the depths of chief operations manager Greg Edmonds' report to Auckland Transport's last board meeting, he reveals that "evaluation of a replacement for the AT real-time service tracking system continues. The new solution will be implemented from September and with the roll-out of new ticketing equipment on buses, will provide significantly improved performance of real time passenger information from the third quarter 2013".
He says a "data warehouse" system will collect and analyse "real time performance against schedule, ticketing data and customer complaints".
It's not before time, but true to form, the transport boffins are not conceding the existing system is a dog, and always has been. Instead a spokesperson tells me "the system is now 10 years old, the technology is old and we now have the funding to update. [It's] time for an upgrade, like a TV, washing machine or fridge".
The only flaw in that argument is that in any normal establishment, no one would have put up with a television that for 10 years kept showing the wrong programmes, or a fridge that warmed the beer instead of keeping it cold.
The upgrade will be on the core prediction and calculation engine system and the on-bus tracking equipment. A trial is already under way on two New Zealand Bus vehicles. Existing road-side hardware will remain.
Amazingly my request for any reports on the failings of the old system and the need for a replacement drew a blank. I was told there was none. Just how they hope to improve on the existing disaster if no analysis of its various failings has taken place, I don't know.
Perhaps the following potted history will help them. The old Auckland City awarded the original contract to Saab ITS in March 2002 to develop and implement a combined signal pre-emption system to give buses priority at traffic lights, and a real-time passenger information system. The latter, to feed information on the arrival of approaching buses to electronic bus stop screens. The cost, $6.9 million.
Trialling began a year behind schedule, at Christmas 2003. By April 2004, Auckland City councillors voted to expand the system from a test area to bus routes across the city - that's despite many warning signs. At start-up, "missing services" were as high as 30 per cent. By April, the boffins were embracing a failure rate of about 3-4 per cent as some sort of triumph. This despite an external audit by Parsons Brinckerhooff warning against proceeding with Stage 2 and 3 until the system was running "with far fewer issues".
Even an in-house Saab ITS technical paper admitted that given the parameters involved in a bus prediction system, "it could well be argued that an accurate prediction is more an oxymoron than reality".
Two years later and with no improvements in the predictions, Auckland Regional Transport Authority project manager Mark Lambert - now AT passenger transport manager - argued the system was still a good buy in that both the software system and the hard were fine. The problem, he said, was the various interfaces between the two, the biggest culprit being drivers who failed to log on properly at the beginning of the journey. Mr Lambert planned to remove the human factor from the process.
He also said that now Arta was taking over the system from Auckland City - it bought the system from Auckland City for $1 in 2006 - it could lean on bus company operators to ensure all buses were properly plugged into the system.
Arta subsequently set about a $17.4 million expansion of the flawed system out across the region.
Despite on-going tinkerings with it, and the development of software to enable ever-suffering bus passengers to access the "real-time" information on their smartphones and computers, the underlying inaccuracy of the system continues to blight commuters' lives to this day.
It broadcasts the movements of a parallel bus system which is uncannily similar to Auckland's, but not the same. It announces buses that never turn up, and forgets buses that do arrive. It announces buses are due in five minutes that have already departed.
By September, Mr Edmonds is suggesting, the 10-year nightmare will be over.
I wish I could believe him.