The AT Outer Link and the Future of Intelligent Auckland Public Transport


 Formal Objection and Appeal to Central Government concerning AT Proposal to Sever the Outer Link

Working Paper in PDF

Facebook: Save the Outer Link <-> Petition


Please also make a submission to AT Have Your Say saying not to cancel the outer link, as this is the main way they will assess public response.
You are welcome to cite this website address in an electronic sumission as a reason for rejecting the changes.

The AT brochure pdf is here. You can attach a document to the brochure and post it, and are welcome to use the unsigned pdf above.

Feedback, ideas and crowd sourcing:Description: MacSSD:Users:chrisking:Desktop:Screen Shot 2019-11-23 at 4.43.15 PM.png


The Outer link is one of two critical circular routes that traverse all the major radial bus routes, thus providing Auckland with an integral cohesive transport network.


AT is proposing to sever the link by cancelling all buses between St. Lukes and Newmarket, cutting off the Eden Quarter from an essential East-West service, and compromising the network integrity of the entire system.


The outer link (yellow) provides a key critical component of Auckland’s integrated transport system, because it provides direct transfers to vastly more routes than any other, by virtue of both (a) its transversality, which intersects virtually every other central city route and (b) because its circularity enables efficient movement to a transfer link in either direction. It also serves one of the most accessible tourist introductions to Auckland. Severing it both removes the circularity utility entirely and reduces by half the transversality. The fact that AT would persistently seek to sever the outer link in the face of previous concerted public opposition indicates both ignorance of efficient system design and political arrogance.


They tried to do this in 2015, but met with public opposition and a Facebook campaign from Mt. Eden school children and agreed to retain it, but now for piecemeal reasons are trying insidiously again to finesse it through.


Left: The 2015 Save the Outer Link group[2] and a pic from Tom Day. Right: Frances Loo in Outer Link colours giving out protest information at the Mt. Eden AT event.
Top right: Patrick and Frances. Lower right: Pete Moth of AT and Mark Thomas.


We need a public action campaign to stop this deranged planning madness that degrades the entire system and by favouring some users against others, sets communities against one another.





This page and PDF document is to formally object to the proposal[1] to sever the outer link circular route between St. Lukes, Mt. Eden Quarter, the corner of Manukau Rd. and Balmoral adjacent to Greenlane Hospital and Newmarket and to request this specific proposal be withdrawn forthwith, for legal, system design and public interest reasons.


Because it has major future implications for the efficiency of Auckland's integrated bus and rail transport system in the age of climate crisis, to ensure Auckland can prosper as a 21st century city forming the hub of NZs commercial energy, it is also designed as a research study to inform central government of urgent needs to redesign transport strategies to ensure bus transport, which will remain central to Auckland's dynamism, fulfills the needs of a smart city economy, by comparison with international cities, from Singapore to San Francisco.


The current AT proposal to cut the Outer Link is falsely claimed to be a transport improvement on invalid grounds, when it is manifestly severing the continuity of the service, compromising the global integrity of the Auckland transport network, and providing no satisfactory alternative for residents in the Mount Eden quarter adjoining Valley Rd. It appears to be a disingenuous cosmetic revival of a previous proposal that was discontinued after it met significant protests from Mt. Eden residents and school children in 2015[2]. 


Description: MacSSD:Users:chrisking:Desktop:Screen Shot 2019-11-21 at 12.15.28 PM.pngContinuity of service depends on existing continuous services continuing to exist, which make the same quality of service possible, and not cutting off central regions from existing direct, continuous services at a time when increasing public transport needs are essential. The proposals are a manifest degradation of service to the cancelled sector and to the overall integrity of Auckland transport.


The two link routes provide transverse cross-town connectivity, complementing arterial radial routes.


The 2015 Review


The 2015 New Network for Central Suburbs Consultation Summary and Decisions Report[3] notes the following three paragraphs:

Some of the larger issues raised which led to the high opposition to the proposals such as keeping the Outer Link and keeping a direct connection from Orakei to the city have been resolved. As a result of these changes we would expect the numbers of people supporting the network to increase significantly if we were to ask the question again after announcing the final network.


During the consultation, a group of high school students put together a Facebook page to express their opposition to the proposed removal of the Outer Link service. Their key concern was the link between Mt Eden and Newmarket, with a number of concerns about losing the connection to St Lukes Mall. The Facebook page received 900 likes. AT staff met with the group to discuss their concerns and the group put in a submission to the consultation.


The Outer Link in the final network will travel the proposed route of the Crosstown 4 between Mt Albert into the city, and the city to Newmarket. It will then use Gillies Ave and Epsom Ave to travel to Mt Eden, before continuing to Valley Rd, Dominion Rd, and Balmoral Rd to St Lukes and Mt Albert. It will run in both directions, similar to the current Outer Link and will provide a direct and shorter service between Newmarket and Mt Eden.


It is therefore a matter of major concern that after this process only four years ago, when the process was stopped due to public objections, a belated attempt is being made to finesse a dismemberment of this pivotal component of transport infrastructure, on ad-hoc, piecemeal, socially divisive grounds.


Integrated Transport System Design Efficiency


There is an integrated logic to operating transverse circular routes, including the inner and outer links, to complement the radial emphasis of arterial bus routes and create an integrated transport system [14]. The proposal destroys this entire functionality. Effectively this reduces the outer link from a pivotal transverse route to a tangential radial route and would provide a markedly inferior service to all residents and users on the deleted section, severing direct transverse transport between St. Lukes and Newmarket through Mt. Eden.


The proposed changes fail to address the structural efficiency issues of maintaining an existing integrated transport design with intrinsic design integrity and instead directly compromise them. Integrated urban transport design efficiency depends on a global understanding of the entire network, optimizing for overall efficiency of transport fluidity as a complex system, measured either by mean number of transfers over all node pairs, or mean transit time.


Description: MacSSD:Users:chrisking:Desktop:Screen Shot 2019-11-22 at 9.20.09 AM.pngSmall world networks[4], based on the “six degrees of separation” concept of world networks, aim for efficiency by minimizing the overall distance between indirect neighbours. They are also believed to be optimally robust to fluctuations. In transport networks[5], which are plagued by traffic intensity fluctuations, particularly at rush hours, system efficiency depends on minimizing the overall number of transfers.


Relationship autocorrelation statistics (Moran's I and Getis's G) and probability p from Xu and Sui. A low value of Getis's G illustrates that the node has low connectivity is likely to be clustered, while a low value of Moran's I indicates that global correlation is small. This explains that small world networks lie optimally at the intersection point between I and G.


Xu and Sui[6] have demonstrated that small world transport networks lie at an optimum intersection point, between low-connectivity clustering and low global correlation, as shown in the above figure.


Wu et al.[7] likewise investigated the relationship between congestion and traffic volume in three types of network, small world, scale-free[8] (fractal networks which follow a nodal power law distribution), and random networks. The result illustrates that when traffic volume is large, random networks are more congested than scale-free and small world networks. Congestion is more likely to occur in a small world network than scale free, but under realistic parameters, the effect is small.


Some parts of such an integrated system must necessarily have differing levels of utilization to achieve the most efficient overall design. A critical functional connection may have significantly less utilization because its value lies in overall system fluidity, not route patronage alone. Therefore design changes based only on piecemeal efficiencies on individual routes will degrade integrated design efficiency, bringing it closer to a random network, driven by a Brownian motion “drunkard’s walk” of piecemeal planning decisions.


While radial services extend the network to new areas, enlarging the system, circular transverse routes improve small-world optimality, by reducing the required number of transfers in the serviced locations. Each trip between two inner city localities on the Outer Link, for example Mt. Eden to Newmarket, St. Lukes or Western Springs, would have to undertake one or two transfers to link to the other locality, which the link already provides directly. Cutting the link also removes two-way circular connectivity reducing fluidity and severs all the direct links in the deleted neighbourhoods. Furthermore, a person e.g. in Mt. Eden can often make a more direct journey to more outlying locations by using the link than transferring through the CBD.


Simulation of a transport network to account changes in network fluidity if a section of a circular link is severed in favour of A-to-B services, with an additional bridging line 7 to maintain services to all nodes. This shows (1) an 18% degradation of total network fluidity, expressed by the ratio of inverse average trip times and (2) significant node-specific degradation of services (f), most significantly on nodes remaining on line 5, particularly the directly affected sector but also on all outer radial routes, particularly the three lower right branches of lines 1, 2, and 3. Notice the increasingly red hue of (d) by comparison with (c) as subtracted in (e). The overlapping histogram (h) shows the distributions of travel times (blue & purple vs severed purple & red). The last 2 rows and columns of (d) are identical because all routes to and from 36 and 37 take the same time.


To analyse degradation of fluidity, on severing a circular link, above is a theoretical demonstration of the effect of severing an outer link (a) in a simplified bus transport network replacing it by A to B services (b) with an additional bridging link to maintain access to all nodes. To avoid exponential runaway caused by tracing all paths in a Travelling Salesman problem [22] , the network is first extended so both trips and transfers can be included as weighted journeys and the cubic Floyd-Warshall algorithm [23] of shortest paths through weighted graphs can be applied. The transfer stops are first indirected in the extended network (g), so that each 2-line transfer (e.g. 9-36) consists of a short journey between bus stops with a transfer wait and each 3-line transfer (e.g. 1-2-3) is likewise a triangular set of links. A weighting of 1 is applied to each single step trip between nodes and 2 for a transfer wait time corresponding to bus frequencies being the same as a two node journey and for a split journey on a radial outlier (upper-left).


The FW algorithm is then applied to the extended network to get minimum weighted distance arrays for both cases, as shown in (c) and (d). The algorithm also produces a second array which can also be used to discover the best path. The trips entering or leaving transfer points are then merged to deduct false transfer weightings at entry and departure from transfer stops, rather than a genuine transit between lines at the transfer point, by selecting the shortest path (e.g. of 9 and 36) for each journey. The array is then reduced to equate trips starting or ending at these nodes. The original and severed distance arrays are then subtracted (e) to give a deficiency array and the rows of this are used to determine net deficiency over all possible journeys from each node (f).


This enables one to calculate both a mean arbitrary trip length for each network and a fluidity score by taking the inverse. The individual node scores lower-right also enable analysis of precisely how all the nodes in the network are degraded individually. The mean arbitrary trip length for the complete network is 5.74, with fluidity 1.74, and that for the severed network is 6.99 with fluidity 1.43, an 18% degradation, showing just how sensitive the fluidity can be to fragmented changes, as pertains with the AT proposal. If the transfer weights are doubled to allow for significant bunching, the fluidities are now over 22% down at 1.35 and 1.07 with a relative degradation of 21%.


The reasons for the AT proposal compromise system fluidity precisely because they are driven by (a) piecemeal fragmented demands including a temporary bus bridge to bypass Mt. Eden station and limitations on the BalmoralGreenlane roue and (b) substandard network modeling which does not address and plan circular route timetabling efficiently. The former is counterproductive to professional transport design and the latter is an issue that other urban transport systems have already solved successfully, as the international examples above show.


Frances Loo notes in her report on discussions with AT staff that: The Outer Link is the 4th most popular bus route in Auckland (more popular than the 27 route). Even the less busy parts of the Outer Link are busier than some bus routes. However, the Outer Link does get a lot of complaints about its service regarding bus bunching and unreliability. Over the past two years, there has been a steady decline in the level of satisfaction expressed in the quarterly surveys of around 800 people. AT targets a satisfaction score of 6/10 or more from respondents to be 80%. The level of satisfaction for the Outer Link has fallen from 80% to under 70% – there has been a steady decline in the past two years and I believe this has been relative to other routes (although, as noted, AT also has an issue with the Inner Link).


The key problem driving the AT proposal is thus bunching, not lack of patronage. The core of the design fault is system-wide. It's manifested in the outer link more noticeably, but that's not a valid reason to sever the outer link, degrading overall system transfer efficiency and cutting the Eden Quarter off from its direct two way cross-town service to it's two key commercial areas, St. Lukes and Newmarket. The core of the problem is inadequate real time driver scheduling, which is draining transport efficiency on all routes. AT has an effective real time board system, informing passengers of actual arrival times and a hop card system enabling integrated transfers, but the scheduling feedback loop to the drivers is manifestly inadequate, so bunching is rife. Some buses are underpopulated because of following one or even two others and then there is a big frustrating gap for prospective passengers because the previous bus was ahead of schedule. The only rescheduling adjustments are made at very few ad-hoc fixed points like Victoria Park Market, resulting in passenger complaints.


A white shuttle car claimed to be used to take drivers from the yard to the Victoria Park Bus stop.
These cars are in short supply they used to be regularly available for the drivers to get between
those places. Drivers protest at their unavailability by arriving late.


The only rescheduling adjustments are made at very few ad-hoc fixed points like Victoria Park Market, resulting in passenger complaints. Eden local Brendan Waters provided the above photo and notes that a root cause of delay complaints here is that Auckland Transport has cut the frequency of the courtesy car that delivers the drivers to the bus stop at Victoria park, so the bus drivers protest by not caring about being late and neither the bus company nor AT  complain. Thus blaming the service for the public complaints appear to be intentional misinformation.


The real time board discrepancies between scheduled and actual arrival times show bunching is endemic to the current AT system design and this is what needs to be fixed, not cutting the link!


In response to complaints in 2017 [16] of buses suddenly disappearing from real time boards, leaving people's plans in "disarray", AT media relations manager Mark Hannan said:  the real time system was configured to display scheduled bus services on the electronic boards, based on timetable information. "Therefore, the information displayed is as accurate as the scheduled timetable for any given bus service," Hannan said. Some scheduled buses disappeared from real time boards because there were some issues with the current system, he said. If a particular service did not run, it had to be manually removed from the schedule to ensure it was not displayed on the real time boards, Hannan said. "If this does not occur, the specific service will continue to appear on the real time boards because the system still believes it is operating. It will "disappear" from the real time board at a particular bus stop once the scheduled arrival time passes."


This statement stands in frank contradiction to the 2018 decision to scrap timetables on all city link buses [19] in favour of a “regular service”.


What is needed is an accurate digital delivery to the bus driver of precisely where they are in relation to the other buses and the schedule, using their real time GPS position, coordinated by an operator-supervised network flow algorithm, with protocols for brief stops at several points en-route, to maintain reliable service intervals at all stops of the route, so passengers have a trustworthy, efficient service turnaround.


An outer link driver that I talked to about this said he was able to make radio contact if needed and showed me a digital interface that he said could compare his position with other scheduled buses that he said was not currently working. The reason he claimed was that it had been 'unreliable', noting that it was difficult to catch up if you were behind, so could only be used for slowing down if you were ahead.


The driver RT connection (centre) and apparently disused GPS console (below).


Eden local Dale Roberts claims that, due to cost constraints, AT switched off the GPS link between AT and the contracted companies Auckland Bus and Go Bus, that enabled drivers to know directly where they were in relation to other buses by GPS, and that AT removed the mandated stops on the bus driver schedule sheets given out at the start of a roster. Hence buses cannot be unbunched by mandated stops to distance buses back to the scheduled times. Once again, blaming the route for bunching appears to be an intentional misrepresentation and a cover-up for obvious networking insufficiency.


AT staff at the Mt. Eden consultation on Dec 7th claimed the GPS system is working to the extent that staff at Auckland Bus (not AT) have a GPS facility and can see where the buses are, but are only able to inform the drivers by RT, if they are seriously out of timing. AT staff at the consultation also tried to claim, in contradiction to the manifest evidence of bunching, that intermittent RT contact was sufficient to alert drivers and that no technology upgrade to provide a direct output would make any difference. This is obfuscation in defiance of technological revolution. Blame was then transferred to the government for the contractor relationship with bus companies. The claim was made that AT couldn’t request a system for drivers to use GPS information because that was an issue for Auckland Bus as contractor, despite this being an obviously negotiable position under any tender contract. Rather than accept that this was a technological advance that could solve bunching, I was advised to deal with Auckland Bus directly to convince myself how effective their system actually was, in defiance of manifest bunching on all routes.


The solution is for buses to set off slightly ahead of time and slow down and/or make graduated short pauses at several dedicated stops if they begin to run ahead of schedule, thus trading off a small increase in transport time against huge bunching delays. The interface needs to be a simple linear screen highlighting the relative positions of the bus and other buses on the same route, coloured by whether they are ahead, on or behind schedule. The reliability of the real time board system is also compromised by it not being fully automated, since the article notes buses have to be removed by hand10, so this needs improvement to make this reliable.


Glympse and Family Locator are free Android apps which already provide essentially the information needed to fix outer link bunching.


There is no defensible way AT can claim such a system is too costly. If AT claims the problem is specific to the outer link, there are only between 12 and 14 yellow buses on the circuit so we are simply talking about 14 satellite units and some software design work using 4G wireless communication to extend the real time board’s versatility. These satellite units could be designed around android cell phones or sim-capable tablets and easily developed by a software design team from a NZ university if they can’t be purchased internationally. There are already multiple android apps that provide exactly this kind of information to individual cell phone users. Simply installing a hands-free cell phone on each bus and setting up Family Locator with all the buses on it would immediately ensure bunching could be avoided. All that would be required to complete the design is an integration with the real time board to provide scheduling data as well – a straightforward software design project.


My Greenlane hospital appointment specifically asked patients to use public transport because of limited parking, so I was dropped off and took the outer link bus home, which was 2/3 full of passengers and school children, despite it not being rush hour. The driver categorically denied the claim that there were significantly fewer passengers on this section and said a lot of people used this section to cross town to get to Newmarket, contradicting the fallacious claims in the AT pamphlet. He confirmed that this proposal, with leaflets filling the bus racks, was simply a revival of the previous unsuccessful proposal, which was shelved due to public opposition.


Quality, Continuity and Directness of Service versus Multiple Transfers


Continuity and directness of service are critical to efficiency, as well as frequency and consistency. The quality of a public transport system depends on the degree to which it can successfully compete with diverse private transport.  The quality of service depends on being reliably able to transport people efficiently in terms of time and convenience from diverse origins to diverse destinations. Transfers between routes, as with HOP zones, aid fluidity without cost penalties, but integrated system design needs to minimize the projected number of transfers per trip, because transfers involve serial scheduling delays and other compounded inefficiencies, such as informational transport route opacity.


One cannot defensibly claim a transport proposal is an improvement if it exists only on paper, when the inefficiencies caused by introducing multiple additional transfers are ignored. A design needlessly causing more transfers while removing critical functional links intrinsically degrades overall system fluidity, causing transit times to increase without bound and eroding service quality and passenger confidence.


Claims that the current proposal is improving quality of service are manifestly false because:


(a) To achieve a connection across the severed section would require 2 transfers between St. Lukes and Valley Rd.  We have an adult grandson with Down syndrome who regularly needs to traverse this route between my daughter’s place in Grey Lynn the Disabled Citizens society workshop in Dominion Rd. and this will very likely cause him to get lost, possibly requiring a Police search.


(b) To achieve a connection from Valley Rd. to Greenlane hospital also would require a transfer at Dominion and Balmoral. I have a broken arm and have to make this journey today for an X-ray at Greenlane hospital and was appalled to receive this proposal in the mail.


(c) Two transfers would be required from Valley Rd. to the Epsom Library by this route - four transfers to get to the library and back, taking 90 mins return rather than 45 minutes at present The proposed 640 route would not serve it directly as the outer link does now, so both Epsom library and Greenlane hospital would no longer have a direct service.


Network efficiency reduction: 1000 sample Monte Carlo simulation of mean overall transit time: 22.5 min for a direct

15-min trip at fluctuating 15-min intervals (sd=2.5 min); 35.2 min for a 1-transfer trip of 2 10-min rides; 43.5 min for a 2-transfer trip of

3 7-min rides. Scheduling fluctuations are modeled as Gaussian with a standard deviation of 1/6 the 15-min service period.


To get an idea of the scale of network efficiency degradation the proposal invokes, the above histograms provide a detailed Monte Carlo simulation. If a person takes a direct route such as Outer link Valley to Greenlane requiring a 15 min journey operating say every 15 mins, with Gaussian scheduling fluctuations, their expected overall travel time is 22.5 mins (15 - 30 mins). To make a similar journey using 2 10-min sections with 1 transfer e.g. at Dominion and Balmoral, their expected travel time expands to 35 mins (20 - 50 mins) and for 2-transfers and 3 7-minute sections it is 43 mins (21 - 66 mins). Transit time uncertainty is also exacerbated because of multiple queueing stops at boarding and transfer. While the passenger has some control over the first boarding queue by using the AT real time board before going to the bus stop, they have no control over the transfer delays, and this cannot be programmed by the network efficiently because schedules which are good for one transfer direction are bad for the reverse direction.


Network efficiency reduction with real time board use but great enough variation in scheduling to cause bunching and schedule leapfrogging (sd=15 min). The one-transfer trip of 37 min is now more than twice as long as the direct 18 min one and the 2 transfer trip at 50 min is 3 times as long and can take up to 100 mins.


Now let's see how things go for a smart passenger who uses the real time board to arrive 3 mins before the first leg with little variation in the real time bus. The direct trip is now very efficient, with a mean time of only18 mins, but now with variation great enough to cause bunching or leapfrogging, when transfers are required, there is a significantly wider variation of arrival times at each transfer stop, due to bunching, so that buses arriving too soon cause a corresponding delay in the next available bus arriving. Now we have a more deleterious situation, with very wide spreads of travel times and longer delays for 1 and 2 transfer trips taking up to 75 and 100 mins.

Between 2001 and 2013, there has been almost no improvement in the use of public transport in Eden and nearby inner suburbs [17] , due to the view commonly held by car users [18] that public transport cannot compete for efficiency and convenience. Degrading the efficiency of existing connections can only exacerbate this at the wrong time.


It is essential that the thrust of AT network futures is targeted at improving travel time efficiency, rather than degrading services and network fluidity, because there is an increasing need for public transport to compete for efficiency with private vehicle use, where relative uptake of public transport has remained almost stagnant from 2001 to 2013, as shown above.


Claiming services that require multiple transfers are an improvement of service overall is clearly disingenuous because transfers require substantive rescheduling delays, and additional confusion and decision-making, e.g. for visitors, children, elderly and disabled, rendering the service opaque and degrading the utility of the service in terms of travel time and reliability, particularly for medical and other appointments, causing reversion to clogging the central city with cross-town vehicle traffic and parking overload. It is also frankly prejudicial for the physically disabled. We will simply have to revert to driving everywhere, which is counterproductive to NZ’s aims for carbon neutrality in a time of climate crisis and irresponsible planning on the part of AT.


Circular Transport Routes in Major World Cities


In its July 2018 announcement [20] , AT presented two vestigial network models for the central suburbs bus network, proposing to replace  the “direct service” option with the “connective network” option routing all bus routes radially through a small number of central nodes. The only nodes cited on their network map remotely in the central suburbs were Britomart, Newmarket, Remuera, Ellerslie and Avondale, all at train stations, providing no rationale concerning connectivity in the localities served by the outer link.


  The AT models are not remotely real world network designs of Auckland.


Neither in any way constitute valid bus network flow designs. No evidence of network analysis is provided to justify the models and the detailed proposed network map doesn't conform to either model, other than attempting to more prominently link train stations with bus routes.


Model 1 in no way resembles existing services, as it consists only of direct links between residential and commercial centres. It bears no correspondence to reality unless the North Shore were zoned exclusively residential and all suburbs to the south of the CBD relegated to factory complexes. It provides no direct bus connections at all between residential areas or between commercial areas indicating a profound lack of professionalism in the delivery.


Model 2 is simply a primitive radial one-node model with no hint of nodal organization. A connective network would only provide a remotely efficient service if there were near-zero transfer times between nodes by precisely timetabling every route to pause for immediate transfers at node centres, impossible unless bunching is addressed by precise driver-network coordination..


In a Model 2 regime, services would also be liable to suffer extreme overloads at nodes during rush hours by concentrating traffic through a small number of intense foci. Neither model in any way addresses flow efficiency, except by increasing inter-node service frequency, thus running buses further than necessary, causing more pollution and congestion.


No urban roading system is designed exclusively on radial connections between nodes, which is characteristic of rural inter-city highways. All urban centres are richly provided with transverse routes, without which transport would grind to a halt at bottlenecks. The same type of analysis needs to be applied to public road transport network design.


All of the real urban bus transport examples below give a more realistic assessable basis for a third model Model 3, consisting of radial routes from the CBD connecting with the rail network at key points, which also retains the inner and outer transverse circular routes to aid network fluidity, which is entirely consistent with current AT network design.


Description: MacSSD:Users:chrisking:Desktop:Screen Shot 2019-11-21 at 12.53.12 PM.png

Kyoto has 6 circular central city loops 203 – 208 [9] .


Major cities in other countries achieve circular route scheduling successfully in both train and bus networks. Kyoto is dominated by no less than 6 circular bus routes, all meticulously inter-scheduled, as shown above.


Description: MacSSD:Users:chrisking:Desktop:60_61_circular_route.jpg Description: MacSSD:Users:chrisking:Desktop:outer_circle_centro.jpg

Both Coventry [10] and Birmingham [11] have an extensive outer loop bus service.


The same applies for the extensive outer circular services in Coventry and Birmingham, illustrated above.


The Brazilian city of Curitiba (1.8m) has a circular link service 022-023 transversely connecting radial routes [13] .


The same for the outer circular link in Curitiba, Brazil, right.


Description: MacSSD:Users:chrisking:Desktop:image.jpg Go-KL consists entirely of circular routes [12] .


The same again for Go-KL in Kuala Lumpur, left, with four routes, all circular.


AT staff have tried to deflect these comparisons by claiming that some routes, e.g. Coventry and Birmingham may be less frequent, or that others run on dedicated bus lanes, but neither of these arguments hold validity, when the root problem is lack of efficient driver feedback with the GPS location of other buses on the route.


Claiming that circular routes have intractable timetabling problems fluctuations and en-route delays is thus manifestly incorrect. Timetabling on radial routes also suffers from bus density fluctuations and turn around delays, particularly at central city stops like C4. In effect, it is a confession of scheduling failure in the age of digital time-keeping. The solution to these issues is not achieved by severing the closed circle but better driver time-scheduling and central communication, as multiple international cases demonstrate.


The argument made in the AT proposal about timetabling being difficult, or impossible to achieve successfully for circular routes, because of scheduling re-alignment pauses, or so-called "bunching", is thus professionally indefensible.


Transparency versus Transport Disingenuity


The claim that "by focusing on the most used section of the outer link, we can improve the reliability of the service for the majority of users" is an actionable misrepresentation, which could lead directly to costly High Court action unless it is withdrawn forthwith, because it is achieved only at the cost of degradation of overall system integrity, and the complete disruption of continuity of service for users from St. Lukes to Newmarket, who may not actually be a significant enough minority or a minority at all, as claimed, denying the quality of a continuous transverse service from West to East entirely to the other users. Political divide and rule on the basis of tyranny of the majority has no place in any inclusive public transport operation.


If there is need to improve the services in Balmoral and Greenlane, address it constructively, not by cutting the outer link. The East-West corridor is a separate issue from the outer link and should not be used as a ploy to cover for a complete severing of a critical transverse circular route in the central city.


Neither does the proposed 640 route have any real bearing on justifying cutting the outer link, because it is a temporary fix to straddle Mt Eden station, which is being finessed as a solution for disadvantaged Eden Quarter residents, when it will later become superfluous, making them pawns in a cost cutting game involving the underground rail link, which should fund this additional service without destroying others.


It is a complete failure of integrated planning to severe the outer link between Manukau Rd. schools and Balmoral and Mt. Eden for general and school passengers and then have to introduce school bus services on the same route to compensate for this causing a school transport deficiency. This fragments a coherent service with integrity and passenger strategic choices into disconnected rump services with no real choices, denying all other users, including the elderly, a key transport connection between St. Lukes and Newmarket. What happens if a school child leaves home late, or is held back, or meets friends in the city after class?


The entire thrust of these ancillary proposals is to subdivide the public into subgroups, claim to provide better services to the west group, deny direct services to the severed south-west section passengers and provide an emergency backstop for affected school children, while justifying it using the temporary Mt. Eden station bridge service. This is inconsistent with public transport fairly serving everyone's needs without discrimination, by effectivelysetting passegner communities needs against one another.


The “have your say” questionnaire is also critically biased, forcing respondents to fill in an exhaustive unverifiable questionnaire about their current bus usage, and a detailed inquiry about just where one boards and departs, followed by why one likes the proposal, before an opportunity to say we don’t want it, relegating any ability to object to the very end.


(a)  This subjugates the validity of concerns people may have about their right to services being degraded, about family, friends or business clients needing to use the service rather than just the user themselves. It conveys the impression that AT seeks the approval of some regular personal bus users who desire a more frequent service along certain routes, at the expense of all other people with legitimate needs for public transport caught up in the cancellation of a critical link.


(b)  AT staff have also indicated varying strategies to inform people on different sectors of the outer link which will also add bias to the results, soliciting support from bus users in the west and comments from householders in Mount Eden. While this could be claimed to better address concerns on Mt. Eden about cancellation of services, it renders all claim of an unbiased election invalid.


(c)  Concern has also been acknowledged by the Albert-Eden Board that the roll out of notification has been deficient, making the responses inconclusive as a basis for justifying the proposal.


(d)  Of major concern is that AT have been advising people at the meetings that they could attach documents with detailed comments, but the internet interface, as far as I can see has no document upload facility, preventing any structured articulate objection being lodged. This raises issues about the legitimacy of the process, because it prevents interested parties such as the Mt. Eden shop owners from attending a hearing with AT to express their concerns. Democracy can survive only if it is well-informed. There absolutely needs to be a provision to lodge formal objections to significant development proposals as a brake on mismanagement. No direct e-mail address was provided until I was given   by Pete Moth of AT at the Dec 7th consultation. However, when I searched after the meeting, this address was neither on the AT Have your say page, nor the AT contact page.


Given the lack of any demonstrated supporting evidence about passenger numbers to justify such a divisive proposal, the lack of any form of defensible overall network analysis, and the fact that, given the evidence at hand, the entire bunching problem is being caused by AT not providing effective communication between the bus drivers and AT's own electronic scheduling, AT has no valid, lawful claim to proceed with the proposal to cut the outer link as a cover for its own failure to adopt 21st century digital driver scheduling, inconsistent with the significant funds spent to set up the front end of such a system in the real time boards at every bus stop. AT thus needs to abandon the proposal to cut the link forthwith and institute digital driver scheduling across the entire network to fix the bunching problems evident on all routes.




The Heart of the City [15] describes it thus: The OuterLink is the easiest way to get around the inner suburbs of Auckland City. It runs from Wellesley St, past the Universities, through Parnell, Newmarket, Epsom, Balmoral, Mt Eden, St Lukes, Mt Albert, out to the Meola Rd entrance to MOTAT, through Westmere, Herne Bay and then back to Wellesley St. The buses are bright amber and operate on a frequent basis, so no timetables are required. It's value both to working people and to international tourists is irreplaceable.


Eden Local Dale Roberts describes it thus: “If you use the Outer Link services at any time, please support the locals of Mt Eden, Sandringham, Epsom, Mt Eden Village, Green Lane / Manukau Rd , Ranfurly Rd Residents Village, the Disabled Citizens society workshop in Dominion Rd, the NZ Blind Foundation in Parnell and more that use the services”, to which we need to add Greenlane Hospital. These are thus major public health issues of disadvantaging the disadvantaged.


AT staff I raised the issue with of intellectually handicapped people such as my grandson now risking getting lost by having to use two transfers to get to the workshop tried to palm this issue off onto disability services. This is reprehensible and shows no understanding of the limitations on such facilities, the loss of autonomy for affected disabled people who can manage the direct connection, or the severity of disabilities required to justify individual transport arrangements.


To propose cutting these services to cover for a failure to invest in basic 21st century network scheduling hardware in the buses is indefensible and makes a mockery of major transport investments in rail and the proposed light rail to the airport, whose budgets are vastly greater. One doesn't improve an urban transport service on which lives depend by severing key linkages that provide an uninterrupted direct cross-town service to local users and comprehensive integration to the whole urban transport system, so the claims in the proposal are intentionally misleading and fallacious to achieve a stated end to make a cosmetic revival of a measure which has already proved demonstrably unpopular, driven by a failure to provide effective driver-network communication, causing the very bunching that is claimed to be the driving rationale.


Bus drivers protesting outsiide AT headquarters


The exacerbated strike action and lockout [21] involving involving AT, NZ Bus and the First and the Tramways Union indicates a toxic relationship of expediency at the source of these problems. According to Jarred Abbot, transport secretary for the First union in a TV1 interview [24] , the problem sources from 7-year contracts imposed when the previous government regulated, obliging Councils to take the lowest tenders, regardless of quality, resulting in up to a 30% loss in wages for drivers, who he says are earning less than a minimum wage, once mandated off times are included, leading to ongoing turnover, recruitment and public safety issues. This claim is consistent with Auckland mayor Phil Goff [25] saying he has some sympathy with suspended bus drivers, but won't be intervening to try to resolve the dispute, stating the council would leave the pay, conditions, and industry recruitment problems, to a working group set up by the government, of which Auckland Transport is a part.


If the central government wants to see public transport achieve any level of efficiency, to compete successfully with private transport, this needs to change immediately. Auckland's transport needs have been particularly badly served until comprehensive planning, including transverse routes occurred, while car traffic congestion has reached crisis proportions, resulting in urban pollution, needless efficiency costs and increased carbon emissions.


Withdraw the proposal to sever the outer loop and install digital scheduling on all routes to avoid bunching.  








[7] Wu, J. J., et al. (2006). Congestion in different topologies of traffic networks. EPL (Europhysics Letters), 74(3), 560.







[14] Farhani R et al. (2013) A review of urban transportation network design problems doi:10.1016/j.ejor.2013.01.001