One of the biggest frustrations we frequently hear about is the real time system for both trains and buses. At some times it’s like the system has a mind of its own with buses or trains appearing and disappearing at random, not turning up when they say they will and generally just being wrong. This is something picked up on by Campbell Live yesterday.
Despite what the Mayor said, my understanding is that the existing system is effectively run using obsolete technology. Things are so bad that it seems a complete upgrade is needed. Here is what the business report to the AT Board Meeting says about it.
An upgrade of on-bus equipment for the bus real-time system is being implemented in parallel to the AT HOP bus rollout. This will improve the tracking of vehicles and performance of the system. A focus on operational processes in May has resulted in an improvement of NZ Bus vehicle tracking from 94% to 96%.
An upgrade of on-road real-time signage to show both scheduled and real-time estimated departure is progressing with 382 of 400 signs completed.
A business case is underway for the end-of-life replacement of the bus real-time back-office system.
There is also piece about the rail system.
A major upgrade to the rail real-time service scheduling and information system, which has improved the tracking of services was completed in May. Work is also progressing with KiwiRail to secure service tracking feeds from the new signalling system into Auckland Transport’s real-time service information system that will provide further reliability enhancements.
Hopefully taking the feed from the actually signalling system will result in some much more accurate information, unlike what I came across this morning when trying to get to town.
And in the IT section of the report we get this.
Real Time System Replacement
- Workshops have been completed to finalise the business requirements and operating model for the replacement system.
- A request for proposal (RFP) has been prepared and is in final reviews prior to release.
- Negotiations with the current supplier have been held and are expected to be completed this month to ensure support during transition to a new system and for components of the existing system that will need to be retained, e.g. current street displays, for a period of time.
Hopefully these upgrades can finally fix a system which has been a dog for years. I’m sure it is something that many readers will be looking forward to.
Processing...
Interestingly using the system in Parnell with Link buses have very few problems. Use AT app on my phone to input stop number, and when I see bus in 3 minutes away leave my house. Very useful.
However at Orakei station last week said train was 5 minutes away, was about to leave station for few minutes until saw train coming across the causeway, and showed up still saying it was 4 minutes way.
My fitness tracker program does a better job, they should release the specs and crowd source the programming; its not new technology when its been around 5 years! crazy words from Mr mayor.
I agree that using an app tracking/ smartphone style system should do the trick, and we (the people) should all be able to watch where all the services are too, Cairns stylee: http://www.tmr.qld.gov.au/Travel-and-transport/qconnect/Cairns-Real-Time/Cairns-Real-Time-Bus-Tracker.aspx.
I hope we’ll see the customer satisfaction data moving over from paper survey collection to crowdsourcing via an app like this too: http://www.tiramisutransit.com/.
Updating RTI and customer feedback would allow AT and the operators to manage problems in real time and also get some handy stats on bus journey time vs car journey times. Bring on the digital revolution in Auckland 🙂
It was an interesting observation in the Campbell Live story that some buses were passing stops earlier than scheduled. Drivers should be constantly referring to their clocks and timetables and this sort of service failure should not be acceptable. If ahead of schedule they should wait until it’s time to leave, passengers on board or not. That’s what happens in Japan, which probably helps explain why everything there connects so remarkably well. This problem does as much to perpetuate the impression of random buses as late services do.
That is a big issue. I sent a complaint the other day regarding a NZ Bus service that departed New Lynn six minutes early (a nice change from my usual complaints about late running services!). The response I got from NZ Bus was that they had confirmed the specific bus I complained about did depart early and they were investigating the cause of that early departure. They seem to deal with early running only when someone sends in an complaint which I find frustrating. They should do an automatic report every day and quickly follow up on all cases of buses departing timing points early. Is the system is at the point yet where automatic reports can be generated with reliable data?
I think it depends on the service frequency. Clearly 6 minutes too early is not good, but if bus running every 10 minutes I don’t think it matters at all. Sitting on buses going nowhere is very frustrating, and puts people off.
How do they/we know where buses are? Do they have GPS units radioing their position to some sort of bus command center?
GPS unit on bus with cellular (Vodafone?) link back to HQ. I seem to recall the data costs of cellular meant the system skimped on position updates – only notifying once the bus was early or late by some threshold, and not that often. In contrast the CHCH used radio band and constant updates (for free). Does anyone know it this changed over Auckland’s decade long debacle?
IMHO Auckland’s lack of bus lanes, long traffic light phases and general congestion is a fundamental problem any prediction tech will struggle with. Let’s hope they solve the total failures though – the fictitious count downs, no shows and phantom buses that plague the current signs.
I wonder if they have changed their position on reduced position updates following the reduction in data charges. Remember when they were first rolled out it was 2004 and data charges back there were pretty high. The CHC system may be ‘free’ in terms of no cost per KB but there would still be a cost of maintaining the infrastructure. I would be interested to see what the cost differences are in the long run. Also I’m pretty sure AT would have some sort of arrangement with Vodafone where they pay a discounted rate — less than what consumers pay for data.
The fictionious countdowns will be a thing of the past when they finish rolling out the updates currently in progress. The only reason for the fictious countdown was to give people an idea of when their buses will arrive but that obviously backfired because they didn’t give people an idea of whether the data was estimated or actual. Whenever I encounter a stop that has yet to be upgraded I just revert to using the web/SMS service to get the RTI. Or in some cases I know the schedule so well and I can actually tell that the bus I want is showing as later/earlier than scheduled (which shows the bus is being tracked since it would just otherwise show it as being on-time).
No shows and phantom are operator-related issues which is getting better and better (though I only use NZ Bus services so I have no idea how bad/good it is with other operators). Has anyone seen any cancelled buses at all (which is supposed to show with a C under the time)? I suspect NZ Bus just operates as much buses as they can even if it means running a service massively late just to be able to say they did operate it that day!
Maintaining radio coverage in Auckland would be costly. CHC has the benefit of being largely flat, so a couple of moderately-powerful radio repeaters can cover the whole city. Auckland is not even vaguely flat, and getting adequate coverage would mean several repeaters in strategic locations – it’d be necessary to have as close to zero gaps as possible, because of the real-time nature of the information and the one-way nature of radio updates.
Also, radio ain’t free. Spectrum licences cost, and a system like this would require a few frequencies. The operating costs for transmission are zero, but the operating costs to support that transmission are pretty high. Cellular data for GPS tracking is so cheap as to be effectively zero per vehicle per month. A mate works for a company that produces GPS tracking systems for trucks, and they get away with a few dollars per unit per month; if AT were doing the buying, in bulk, they could probably get every bus reporting actual real-time for maybe $5/bus/month.
While I am certainly annoyed with the system as a user, I was equally annoyed with the TV segment which basically just moaned and wrung its hands. I would have liked to hear how the system works – or rather how it is supposed to work – and what the $$$ spent on improvements is being used for i.e. what is changing that will magically fix it and get it to tell the time correctly.
Couldn’t agree more with you Steve – and as per usual Brewer gets his 2c dig but offers no recommendations for improvements. On the whole its another example of lazy and tabloid style journalism that NZ suffers from.
Agree with other comments, no real revelation about why the system doesn’t work or how they’re going to fix it. Love that Christchurch has had a working system for the last 6-8 years, why not ask them how theirs works?
Actually going since around 1999. Provided by Connexionz a Christchurch based company http://www.connexionz.co.nz. It’s a dispatch based system as opposed to a driver based system. Untracked buses appear to be highly unusual which is much easier to manage with dispatch based systems. Helped of course by Ecan also having significant financial penalties on their bus companies if they let a bus run untracked.
I agree, it would be helpful to know how the system works and what the money is/was spent on. Another 1.5M doesn’t sound a lot if the whole system cost 40M, so we should be happy to spend it if it makes the system more reliable.
I’ve had so many frustrations with this real time board, sometimes the smartphone app gives you different times than the sign at the stop, then which one do you trust? Unfortunately the answer often is ‘neither of them’.
Once this “real time info” almost made me miss a flight and I had to resort to sharing a cab to the airport, when I was just looking at the display which announced several Airbus Express services. I eventually noticed a sign telling me that just for the Airbus Express the stops had been changed to Symonds St for that evening. Luckily I wasn’t the only one who fell for the display and the shared cab cost little more than the bus would’ve.
I’ve found the RTI to be nearly useless for Airbus Express services which is really disappointing as the Airbus is the very service that would benefit most from real-time information (not to mention being a visitor’s first impression of Auckland’s public transport system). It has improved in recent time with the occasional service now being tracked but it’s nowhere as good as NZ Bus is. Even Tranzit seems to track most of their buses with only the occasional ones not being tracked.
I once casually mentioned to a driver that they needed a sign at the new starting stop on Victoria Street (outside NZ Post) and she said it wasn’t likely as they weren’t putting in any new signs since the system was so broken.
Simon
The problem with the Auckland Transit (AT) bus system is not just that it’s unreliable (you Kiwis love to buy overpriced, ineffective foreign solutions over our proven and highly effective home grown solutions – supporting local industry, jobs and the economy), but that it also operates on a ‘deviation from schedule basis’, which has been considered largely obsolete throughout most of the developed world for many years now because they’re so unreliable. And the evidence clearly shows – if you buy a solution that delivers unreliable real time service information to the public, ridership numbers can actually fall as a direct result.
Radio based systems are by far the most most common globally for tracking buses in real time, and radio systems are more cost effective for larger agencies than cellular based systems. In fact, I can’t think of a single cellular based system anywhere in the world being used by agencies with over 80 buses. Support costs for radio based solutions are minimal, albeit the initial capital investment required is slightly more than a cellular based solution. As a general rule of thumb – for a 60 bus system, return on investment for a radio system in comparison to a cellular system can be realized following 2-3 years of operation.
There’s a good reason why a certain NZ company is taking the US by storm, delivering much better systems that that currently in use in Auckland and Wellington. Perhaps AT needs to have a good look at itself in the mirror and decide whether the people making the decisions there aren’t commercially/technology biased and perhaps lack some national pride….
Many good points but all assume that AT actually chose the current real time system. To be fair on AT the system was inherited with its origins in the old Auckland City Council and Saab, inherited by the former Auckland Regional Council and then inherited by Auckland Transport. I’m not even sure RTI was even a key purpose of the original system the city council instigated but rather traffic priority. I have heard from staff at AT a while back that they would have been rid of it years ago but for budget cuts which forced them to cancel a tender process a few years back and go back to the current system supplier and negotiate an extension to the system rather then total replacement. which was their favoured alternative.
Would be nice to see real time bus positions in AT smartphone apps like internationally and have the RTB system estimate based off that. E.g. Average arrival time to stop position from that distance. Could be done with the Google maps API and some better tracking equipment on bus.
The Christchurch system has worked reliably since 2001. Like an aircraft, there are regular updates and refreshes of technology and despite that it has still cost only a fraction of the some $40m spent on the Auckland system. Its still working fine despite their earthquakes.
Its a travesty that we have a reliable homegrown system that is a living testament to the best of NZ’s No 8 wire tradition that was overlooked because some bureaucrat chose an overseas solution in the interests of their own career prospects. Somehow the supplier with the longest track record of success got overlooked.
BTW I see that their system not only delivers information to public signs but also to smartphones and computers across the internet. They have done for years. Why can’t we just hand the job to someone who knows what they’re doing?
So does Auckland’s system. There’s a terrible website and even worse mobile app. But there’s also a public API, and while AT just choose not to tell anyone it exists or how to use it, there are a bunch of third-party apps that can also show the real-time info (I use https://play.google.com/store/apps/details?id=com.dyancat.bus)