I first heard about the bus route that has become the “Outer Link” quite a long time ago, late last year I think – as kindly some people at Auckland Transport sought my ideas on what they were proposing. My initial response was fairly mixed: positive about the fact that the route links together so many places, provides for so many trips – particularly trips that public transport doesn’t traditionally provide for, by offering decent frequencies seven days a week and through its “loop” structure. But at the same time I was also a bit sceptical about aspects of the proposal, in particular my concerns about whether it could keep to time or whether the buses would “bunch” together – like happened on the LINK bus route, though to an even greater extent because of the longer route.
This is what I said once the plan was released for consultation on that matter:
Essentially, I think that unless we have some well-targeted and necessary infrastructure upgrades along the route of the outer loop, it will not be possible for such a long, looping route to be reliable. What Auckland Transport really should be doing is driving a few buses around the route during peak times, identifying where the delay points are and investigating measures to deal with those delay points.
I repeated this concern about reliability and bunching in my submission – as I think a number of other people said. After all, these issues are inherent to loop routes (hence why the London Circle Tube Line was changed recently to more of a spiral route).
Well so far I haven’t had any problems with the reliability of the buses, or with the bunching of buses, I am somewhat concerned about the way in which Auckland Transport is avoiding bunching becoming a problem – by providing extremely generous timings for buses to reach certain “time points” around the Outer Link’s routes, then leaving the bus to sit at these various stops for what seems like an age before continuing their journey.
From what I’ve seen so far, on the quarter of the route that I’ve used (between town and Pt Chevalier), it seems that there are three time points even on that small part of the route: at Westmere shops, at Ponsonby and in town just outside the Civic theatre. In some respects the more time-points the better, as theoretically the wait at each time point should be minimised, but what I’ve seen so far is that the waits can be pretty long. Two examples include a 6 minute wait at Westmere shops for my outbound Outer Link on Sunday and a wait this evening at the Wellesley Street stop which seemed just about as long (it was long enough for the driver to have a cigarette while he waited to catch up to his scheduled slot).
Now I understand the point of the time-points, as without them you will undoubtedly end up with buses bunching and not being at all reliable, but I wonder if Auckland Transport has taken their response to that concern a bit too far. It’s incredibly frustrating sitting there waiting for the bus to simply catch up to its scheduled time while cars zip past you. So perhaps I’m starting to regret the huge deal I made out of wondering whether the Outer Link buses would bunch and be reliable – as we’ve now gone to the opposite extreme.
But what’s the solution? Auckland Transport might be wondering why I can’t just shut up as I’m never happy – and I can perhaps see the logic in that. Ultimately though, I think this was why I found myself so concerned about such a massive loop route, because either the buses would bunch and become unreliable or we would end up having to wait huge lengths of time all throughout the route, with the timetable reflecting a very conservative speed of the bus (I mean how on earth could the bus get six minutes ahead of its time between Ponsonby and Westmere on a Sunday for example). I think that the real solution will actually require a bit of work: eliminating the bottlenecks so that we don’t have to be so conservative with our timetable (through providing bus lanes at key points of the route), having traffic lights respond to the buses so that especially if they’re running at all behind times, the lights immediately go green enabling the bus to catch up. Once again that would ensure that the timetable doesn’t have to be so conservative (leading to huge waiting times).
Those are long term things though, in the short term I hope that Auckland Transport does keep a close eye on how long the buses are sitting at the various timepoints around the route, analyse whether it’s really necessary to have the timetable being so conservative (particular during off-peak times when travel times should be pretty standard) that the buses are constantly getting ahead of their schedule. In the meanwhile, I think I’ll probably avoid the Outer Link and try to get the 005 as much as possible – speed is important.
Processing...
As I have said before about the link, I think the answer is not to keep to a timetable, but instead to have each bus maintain an appropriate “leading/following distance”. This could be implemented by a linked gps based system, where each driver has a monitor showing how far ahead of the previous bus they are and how far behind the leading bus they are. The drivers could then just smoothly lower their speed without big stops. And if traffic is light, all the buses could go faster, increasing frequency
Yes, that’s exactly the right way to approach these issues. The point of a frequent service is indeed to do away for the need for timetables.
Not having a published timetable presents a huge opportunity.
Given that all the buses have GPS’s and data links to a central base for the real time screens real time software could be used in place of a fixed timetable.
Although I am a Mechanical Engineer (not Software or Engineering Science) I still think it would be quite fun to design the algorithm.
I would envisage the system working as follows:
– Data would be collected from the current buses to work out the average time taken to complete each (say 100m) block of the route. Depending on the time/skill of the programers it could vary dynamically based on time of day, day of the week, and even the time taken by earlier buses to predict traffic conditions.
– Real time data from each bus would be used to find if that bus is ahead or behind compared to its peers (weighted average favoring buses either side of that bus.
– A screen(or other alert) in the bus would display (or advise) desired speed. For example: 1. Max safe legal speed 2. Normal speed 3. Comfort/relaxed speed 4. hold 10 sec at every stop 5. hold 30 sec at every stop
This would allow buses to loop continuously at an appropriate speed for the day. The current situation seem to be timed for a heavy traffic day.
A real time system for maintaining bus spacing would be way better than the current “wait for timetable” system.
I like those ideas Scott. I suppose that’s my ultimate point – that we need to be smarter about this, not just make the timetable extremely slow then introduce a whole heap of really long hold points.
Might as well provide the best service possible with the available resources. Seems crazy to have $415000 buses, with drivers sitting waiting several times a lap just in case they run late at some point. Not having a published timetable gives huge operational flexibility.
Scott – I am (sorry was) a software engineer and everything you suggest is relatively easy to implement. In Australia some of the mine sites have driverless ore trucks fitted with GPS and an on-board computer controlled by a central computer. All the algorithms required for time to arrival, queuing etc are already in use. How much more freedom would we all have if we abandoned timetables and instead told the travelling public that, for example, on a particular route there will be ‘x’ number of buses spaced at specified time frquencies. The central computer can link to the internet providing real-time information as to when any bus was due at a certain stop. Again, have to agree with Admin – we need to get smarter and use technology to remove problems not just smooth out the edges.
‘keeping to the timetable’ seems like it will always be problematic, because traffic flows are so unreliable right? Perhaps if drivers are running a bit fast, they can pause for no more than a 2 mins to help regulate the service. 6 mins and a cigarette break is taking the piss I reckon.
An idea: Have a point on the route where buses pause to catch up with the timetable. Make the pause long enough so that there is some overlap, usually a second bus will catch up to the first bus. Allow passengers on the second bus a free transfer to the first bus, so that the buses wait, but the passengers don’t. The cost would be an extra bus, or a less frequent service. I would prefer a more reliable, lower frequency over unreliable slightly higher frequency.
We have the same problems with Sydney Trains. The newspapers continually winged about late running trains and so the timetables were given more “recovery time”. Your fears of busses bunching have been alleviated the same simple inefficient way. If none of the high tech methods above are implemented then the solution is to remind the drivers that they are paid to drive the timetable and not to get ahead of it. From a passenger point of view a six minute early bus is a disaster.
I reckon loose the timetable and get two more buses. At the end of each month the bus driver who did the quickest “lap” gets a bonus. (jokes)
One reason some people avoid public transport is exposure to tobacco smoke. In the light of that, why is the driver taking a cigarette break? That’s just wrong on so many levels.
You can’t possibly be suggesting that tobacco addicts be banned from jobs as bus drivers.
“One reason some people avoid public transport is exposure to tobacco smoke.” Really??? I have never heard that before. What is your source for this, or do you have any links to back up this claim?
I am genuinely curious as this makes no sense to me. Do these people refuse to leave their house/car/workplace in case they walk past a smoker? And if so, wouldn’t that be more hazardous to their health than a wiff of second hand smoke now and then?
Of course drivers shouldn’t be taking breaks while there are passangers on board a bus; but that goes for coffee breaks, lamington breaks, etc, as well as smoko breaks.
Well from personal experience I avoid going into the Wellington CBD due to the tobacco smoke. I also avoid transferring at Porirua station. It greatly annoys me when someone is smoking on a train station platform or at a bus shelter. If I do transfer from the train to a bus in Wellington I walk away from Platform D at the bus shelter and up Thorndon Quay.
I briefly lived in Melbourne. They ban smoking in train stations, on platforms and in bus shelters, but the rule is commonly broken. I could not get from Southern Cross station where the rule is broken to my job without walking past dozens of them so I chucked in the job and went to a lower paid job.
Don’t underestimate the nuisance factor of tobacco smoke. I will do just about anything to avoid it.
And yes I would hire non-smokers over smokers any day of the week (and if it is against the law to do so, I will only do it once for every butt I see on the ground, because that too is illegal and smokers are the biggest scoff-laws there are)
Again in Melbourne a tram driver stopped, got out, and had a cigarette filling the tram with smoke. I got him sacked.
People’s smoking affects my usage of public transport. Avoiding smokers on platforms (where it’s actually meant to be smoke free), not sitting by the doors where people often take the final drag and often shifting because those same people often sit by me. Also there was one occasion where there was someone smoking on a train
In Birmingham UK some years ago, customers complained to the local authority that empty buses were simply driving past them at bus stops without picking up passengers. The spokesman for the transport dept. told repoters who questioned him about this that “if the buses stopped to pick up passengers, they wouldn’t be able to keep to the timetable” – Seriously!
Wonderful, Bob…
Today I’ve noticed another method of keeping to the timetable – the InnerLink bus didn’t drive any faster then 30-40km/h (where there were at least 2 lanes, giving other vehicles opportunity to overtake). That reduced the dwell time significantly. It’s not ideal, but definitely less frustrating then spending 6mins at a bus stop.
Yes, the Links are perfecting the method of driving 10-20kms slower than all other cars and buses. This is fine, almost, until halfway through the journey they stop for approx 5 mins. This morning my driver pulled out her newspaper, which she was reading at each red light anyway, and just sat. My excitement for the Inner Link is completely gone. I’m going back to my 962 and 966 via Ponsonby to Newmarket. At least I know they’ll a) turn up, and b) get me there. Seems basic, doesn’t it?
Your concern was legitimate and it’s not your fault that AT decided to do this on the cheap.
hopefully the lack of a published timetable means that service is in ‘beta’ mode so operations can easily be altered.
Would be much better having 1 or 2 timing point, and these would be at the quietest points in the route. I imagine this could be at Newmarket? With 10min frequencies could have 1 transfer point, and if 2 buses met at this point, then people could transfer to the bus in-front.
A smarter system as outlined above would be much preferable though.
Combination of reduce the number of timing stops and drivers adjusting their speed via GPS. Also tighten up the timetable so there isn’t so much slack, but enough slack for a bus to gain 5 mins (say) in the space of 1/2 an hour. The risk is that buses will be behind schedule, but the schedule isn’t published anyway so I don’t actually think this is a big problem. Have slightly different timing schedules for peak traffic and off peak.
Also add another bus to the circuit so that frequency is *at least* 15 mins. Timing stops should be at the least busy points (Pt Chev terminus, for instance), not the middle of Ponsonby where you get to watch the 005 sail past you.
To add to Camerons comment “stops should be in the least busy points”
AT should reduce the number of time points as a function of the average number of boardings vs the average load of passengers.
It would be really easy for AT to ensure that there is an even headway over the section with a high number of boardings and remove unnessicary time points.
If it is unlikely that there will be pickups in the next section, there is no issue with the bus being ahead of schedule. Why delay 10 passengers on the bus when you are only going to pick up 2 more passenger?
Cheap and quick: perhaps the bus driver could tell the passengers something like “because we’re going a bit faster than the timetable I’m afraid we’ll have to wait here for x minutes”. Doesn’t change anything but maybe it’ll lessen the psychological blow to passengers by a) making them think they were going faster than usual to begin with, and b) knowing the wait will be a fixed and finite amount of time. From my experience with british trains, knowing that the delay will not be indefinite is always nice (if you trust anything they say).
The new Link services. So far there seems to be one major route objection and I have a grumble that the Wellesley St stops for the Outer Link don’t have any destination boards so you don’t know whether to keep walking or stop and wait for the bus.
The route objection is that the concentrated population of(hilly) Freeman’s Bay has lost a really useful evening and weekend bus now the 015/017 has gone. There is nothing now going up and down the K Rd/Howe St/Wellington St/Franklin Rd route. The advice on MAXX if you want to get from Howe St to Britomart is to walk up to Union St to catch the new 020 service. Or from Freeman’s Bay Primary you should walk down to Vic Park and take the Inner Link. This is a community that is used to public transport. They are not big car owners/users. Seems a real weakness in the new system.
The 020 runs right along Wellington St, just past Freemans Bay primary school.
In connection to the Freemans Bay bus services, I see the MAXX site in their FAQ section acknowledges that the old connections with K Rd have gone and may create an issue. We are talking a hilly suburb, here, with a lot of elderly and disabled residents who look to public transport to get about. But these will be issues for lobbying during the 6 month review period.
And sections of the MAXX website haven’t yet been updated to allow for the new central city services. The ‘How to travel’, Central Akld section still refers on the map of stops and routes to the 004, 005 and not to the new services. I have contacted them as we should be expecting heavy use of the site from tourists for the RWC.
The Outer Link is presumably trying to get a balance between the 15 minute frequency, the running time and the number of vehicles required. e.g. 6 buses would run 15 mins apart if the running time takes 90 mins. If a peak trip take 95 mins or 100 mins you would need an extra bus to keep the 15 min promise. The other option is to run the 95 min trip late if you are not too worried about 1 or 2 trips not meeting the customer promise.
If you popped in an extra bus in the 100 min example that results in 7 buses running at 15 mins apart = 105 mins. In this situation you either run them closer together 13-14mins apart, or have some slack at some of the timing points around the outside.
Over the peak the running times on the trips probably get longer and may vary which might explain why some trips have more wait times on the outer points. Depends a lot on the variability of running times and whether people are comfortable with having some trips run a little late, or whether you want them consistently 15 mins apart.
Will be interesting to see what happens if things get busier over the World Cup and when they defintiely get busier in the new year.