tip off
20

How can public transport work better in cities?

Bus network proposed for Toronto in 2009

Many people are surprised by the relatively small share of urban passenger travel made on public transport in Australia. Notwithstanding a slight narrowing of the gap in recent years, cars still dominate travel in capital cities, accounting for 85% of all motorised travel in Sydney, 89% in Melbourne and 90% in Brisbane. Their share in Perth and Adelaide is around 93%.

While there are a number of reasons for public transport’s low mode share, a key reason is it serves a small number of destinations – mainly the CBD – really well, but generally fails miserably for those travellers who want to go elsewhere, especially the great majority who want to travel between suburbs.

The ‘hub and spoke’ pattern was fine a hundred years ago, but now most destinations and origins are highly dispersed. For example, only around 10% of all jobs in Sydney and Melbourne are located in the CBD and a clear majority are more than 10 km from the centre. Also, fewer than 10% of residents live within 5 km of the centre.

As a consequence of this out-moded form, using public transport to travel across suburbs is often excruciatingly slow. In many cases services are infrequent, operate on restricted hours and require intolerable waits between connections. Where they exist, cross suburban services follow seemingly capricious timetables and tortuously circuitous routes.

Much needs to be done at the political and technical levels to address these problems. One important approach is the idea of a ‘synergistic’ public transport system (or ‘networked’, or ‘connected’, or ‘multi-directional). It conceives the entire public transport system as an entity – a network – rather than as a series of isolated ‘lines’.

In conceptual terms, the idea involves a backbone grid of very high quality public transport services operating across the metropolitan area of a city. Imagine a series of parallel north-south services overlain by a similar sequence of parallel east-west services, all operating at high frequencies until at least the early hours. Because Australia’s large cities have legacy radial train systems, the grid is likely to look more like a spider’s web than an orthogonal network – while that’s imperfect, the idea is essentially the same.

The grid must be dense enough that passengers can get to it easily from their origin or destination, ideally by walking. Once they’re on it, they can transfer between services at each node, giving them the ability to go “from anywhere to anywhere at any time” rather than just to the CBD.

The synergistic value of the grid lies in connections. But to be attractive, transfers must be as painless as possible. That means a single fare irrespective of the number of transfers; short waiting times in a pleasant and safe environment; and easy physical access between modes – the idea of crossing a major road to get from the station to the bus is anathema.

In particular, timetables either have to be coordinated (which is technically hard) or, preferably, services have to be so frequent that passengers don’t need to bother about timetables. Most practitioners think the minimum is an average wait of five minutes with a maximum of ten minutes (see exhibit).

I’ve described the high-level concept. In reality, cities differ – some have large water bodies, hills and higgledy-piggledy street layouts that mean only an approximation of a grid will be feasible (sounds like inner city Sydney). It’s likely the orbital services will mostly be provided by buses operating on existing roads rather than by light rail or bus rapid transit services with their own dedicated right-of-way.

It’s also probable the grid will have a hierarchy in service quality, with some services in areas of low patronage potential operating at lower frequencies. Moreover, a number of origins and destinations will test the idea of what’s a reasonable walking distance, with some patrons relying on feeder buses to get to their nearest station.

Nevertheless, this is a valuable aspirational model that sets a direction for long term planning. But it has considerable value now. For example, it suggests it’s worth asking if investing billions in new radial rail lines to carry small numbers of people from the suburbs to the CBD (e.g. proposals for new Rowville and Doncaster rail lines in Melbourne) is the wisest use of scarce transport dollars. Perhaps they’d be better spent on enhancing the connectivity of the public transport system so as to create a real, synergistic network.

Some cities have already moved part way toward establishing a synergistic transit system. For example, Melbourne has a single time-based fare structure and three suburban orbital SmartBus services operating from circa 4.30 am until after midnight, mostly at 15 minute frequencies.

There are many technical debates around the details and practicalities of this proposal. I’ve hopefully side-stepped most of them by looking at an idealised model.

Discussion of the networked approach is also frequently embedded in discussion of strategic issues – for example, whether or not it can work successfully unless it’s owned and managed by government; whether it should be approached as a replacement or competitor for cars, or as a complement; or whether it can work successfully in low density residential areas like the outer suburbs of Australia’s capital cities.

These are important and interesting issues but they’ll have to wait for another day – they’re far too complex to go into now. The key point is public transport in our cities should be planned as a network so it gives users access to all parts of the metropolitan area, not just the centre.

In the meantime, anyone interested in further exploration of the network approach could read Jarrett Walker’s new book, Human Transit, and Paul Mees Transport for Suburbia. Both are clearly written and explore interesting and challenging themes. There’s also this short paper by Publictransit.US and this longer one out of Griffith University by Jago Dodson et al.

20

Please login below to comment, OR simply register here :



  • 1
    RidesToWork
    Posted February 2, 2012 at 8:01 am | Permalink

    I like the idea of using bikes to extend public transport networks. In Holland and Germany, so many bikes parked at the train stations that you can hardly move!

    Now, with public bike schemes that could serve travellers at the other end, instead of having to walk or take the bike on the train, you could just ride to the station and use a public bike at the other end. This makes the public transport network more efficient – connections can be wider spaced and even a small increase in distances between stops could substantially reduce trip times.

    If only we could get rid of the dratted helmet laws and make the public bike scheme (and cycling in general) more popular, we might save millions on the density of a new public transport network, encourage people to get more exercise (reducing health costs) and improve cyclist safety by increasing safety in numbers.

  • 2
    bluepoppy
    Posted February 2, 2012 at 9:25 am | Permalink

    One of the other reasons which makes public transport planning a challenge is a change in lifestyle. Many families, now with both partners working, often have to factor in dropping off kids to childcare and picking up groceries on the way home. Life has become a matter of juggling many priorities that just does not fit easily with the limitations of public transport. It makes it much harder to have a one-size fits all approach.

    Also, many jobs include the need to visit clients meaning the capability of leaving the office at various intervals during the day to make those calls which generally necessitates the use of a car. Some cities do offer a continuous service within the CBD but in places like Canberra where businesses and departments are spread about (which is a good thing for congestion and parking) it means some of these journeys will be outside the CBD.

    We live about 15 minutes from the CBD and it still takes two buses to get into the city because of the interchange system. It is a catch-22 because of the first point made about lifestyles, it comes down to affordability and public transport participation rates. Certainly a more extensive network taking into account other metropolitan needs outside the CBD is a good first step.

  • 3
    Strewth
    Posted February 2, 2012 at 9:42 am | Permalink

    Good post, Alan. (But then I would say that as I agree with almost all of it!)

    Walker and Mees both emphasise the importance of a comprehensive network to serve dispersed travel patterns, and this is probably the most important take-home point as it’s a lesson that Australian transport authorities have refused to learn over the past century (with Perth a partial exception).

    All the same, it’s probably over-reaching to suggest that employment patterns are dispersed over the suburbs to quite the extent you’re suggesting. You mention that a majority of jobs in Melbourne are more than 10km from the GPO, for example, but you don’t have to go that much further out before the majority becomes a minority. According to the 2006 Census, two-thirds of employment locations in Melbourne are within 20km of the GPO. It happens that that 10-20km band includes a few large suburban agglomerations like the airport, Monash University, and the Chadstone and Doncaster shopping centres.

    The other thing to keep in mind is that the kind of medium-to-large businesses that generate significant commuter flows are found much more in inner Melbourne or Sydney than in the outer suburbs. So while there’s a lot of outer suburban employment, much of it is highly localised (and reflects the fact that the majority of travel to work is confined to one’s local area). Not to mention that since 2006 there’s been a surge in CBD employment in Melbourne, so even that minority of jobs in the actual CBD is on an upward trend.

    That’s why it remains important in the Melbourne context to complete the radial rail network to places like Rowville and Doncaster. As you say, a complete network requires both the high-capacity rail corridors and the supporting web of bus routes. It’s very difficult to build high patronage for public transport when there are large regions like Manningham and south Knox with no rail services at all. And in Melbourne, those rail services are still best oriented radially because that’s where the major travel flows are. Not to mention that they catch a couple of those big suburban employers like Monash Uni and Shoppingtown.

  • 4
    Steve777
    Posted February 2, 2012 at 9:50 am | Permalink

    Certainly in Sydney, unless you are travelling to the CBD or to somewhere on the direct route to the CBD, public transport is not an attractive or particularly viable option unless you can afford to spend half of your day travelling or waiting. If you are travelling across suburbs, then between uncoordinated connections and circuitous routes, it can be barely quicker than walking. And of course, ticketing is a problem, especially for occasional users of the system – where and when to buy tickets, what ticket you need, buying tickets at each change.

    It should be possible to establish some sort of grid, even in Sydney, using railway stations and major business / employment centres as nodes, even if meeting a standard of an average 5 minute wait would be challenging, especially further out. We should consider using use mini-buses for some routes, especially out of peak hours, to reduce the cost.

    I and many would-be users would also be happy to pay fares that are 50 – 100% above current levels if this would help get something like this up and running. It would still be cheap compared to the cost of parking anywhere near a major centre or getting a taxi (assuming you could find one). Subsidies could be provided for pensioners and students.

    The grid services should travel direct routes between the nodes rather than wandering up every byway. If the grid were dense enough, the majority should be within a reasonable walk of a bus stop or perhaps could be dropped off or park nearby. We would need to consider those with mobility problems, but it’s hard to see that they are well served now unless they are lucky enough to have a bus go past their place.

  • 5
    NeoTheFatCat
    Posted February 2, 2012 at 10:17 am | Permalink

    I think this is about more than just lifestyle but something inherent in the Australian culture. Other countries around the world face the same lifestyle issues.

    Firstly, Australians demand door-to-door travel. The idea of actually walking between any place and a bus stop is simple anathema.

    The second point flows from this – there is no way known we should be forced to change mode anywhere. We want one bus/train/tram to take us from our front door, all of the way to work.

    Next, our lifestyles are so busy that we just have to be able to travel when we want (need?) to. I can’t wait 20 mins for the next bus, I just have to make that journey now.

    Finally, why would we want to sit in a bus/carriage with all of those other people? They smell, make funny noises, play their music too loud, might do something strange or embarrassing!

    I live in Canberra. As a single car couple (why waste money on an asset that sits around 90% of the time?), I have used buses extensively as well as cycled to work a lot. I have also seen the comments from friends and colleagues about why they don’t use the bus system.

    I have also lived on and off in Geneva over the last 3 years and used their public transport system extensively. I think in Europe there is a culture of using public spaces, including transport – if you live in an apartment, then the public spaces are the playgrounds for kids and the social space for friends. In Australia, we use public spaces to get something done and then retreat back to our own (or our friends’) private spaces. People are used to being around other people, and joining the crush on public transport is not necessarily something strange. As a result, the transport has higher patronage which means they can run really good systems.

    Some of the things that I enjoyed about Swiss public transport:

    - connection between nodes. The connection from bus to tram to train doesn’t just happen at main interchanges. This means that cross-city trips can be planned without having to go into the main centre and back out again.

    - greater overlapping the closer you get to the city centre. This means that, although times of individual lines could be 20-30 minutes during the day, when you get to within 3-4 kms of the city then usually any service will arrive within about 5 minutes. This helps with increasing the use of transport for those short trips (eg. need to duck into the city during the lunch break).

    - great timetables. Transport arrives and departs according to the timetable. I have often been on buses where the driver hangs around for 30secs at a stop to make sure the bus is on time for the rest of the route. This makes it easier to plan journeys that span multiple times. It also means their web and mobile timetables are brilliant for connecting journeys.

    - integrated ticketing. One ticket and I can travel on bus, tram, train and even water taxi.

    - availability of tickets. Most buses and trams have onboard ticket machines, and about a third of the stops also have machines.

    - admittedly, the Swiss are very conformist and everyone buys a ticket, even though there is hardly any inspectors. This must make it easier to maintain their revenues

  • 6
    Alan Davies
    Posted February 2, 2012 at 10:23 am | Permalink

    Strewth: The “average job” in Melbourne was 15.7 km from the CBD in 2006. There is indeed some “bunching” around the 17-20 km radius but nevertheless the great bulk of suburban jobs aren’t in activity centres. Even with an undemanding definition of what constitutes a centre, 80% of suburban jobs are outside centres i.e they’re dispersed. You can see more in the Employment category in the side bar.

  • 7
    Ben Sandilands
    Posted February 2, 2012 at 10:31 am | Permalink

    On-line connectivity is I think coming to the rescue of bad public transport networks in this country. Once people realise their tablet works across most of the system they unlock travel time as useful time, something that cannot be achieved while driving without dying or going to jail.

    Yesterday for family reasons I was mostly in Sydney, but I had a signal at least 95% of the way from Moss Vale to the city including for the first time in the tunnels between Mittagong and Yerrinbool, a previous dead spot. I also saved at least $100 in parking fees and running costs, and recovered at least four hours of otherwise dead time sitting behind the car wheel.

    Connectivity doesn’t require any capital from the public transport providers. It is coming about because of demand on private telecommunications providers, and delivering a bonus to the public transport in the process.

  • 8
    bluepoppy
    Posted February 2, 2012 at 2:22 pm | Permalink

    I think it is more than just a difference in culture. Canberra for example, is very different from other cities. I have not lived here as long as some, but in other, admittedley larger cities, there is not the same wait for a bus nor the need to change buses for what are relatively short distances. The problem seems to be a greater spread of population whereas large cities have denser in-fill (and out-fill) which perhaps legitimise the cost. I am not sure what the answer is. My experience is that a mix of trams, trains and buses seems to fill out more of the gaps. In the US, many families went for live-in nannies, usually of international origin, to offset some of these childcare issues. My father migrated from Europe for many reasons including overcrowding but on my visits the public transport system is excellent.

  • 9
    Jim Wright
    Posted February 2, 2012 at 4:11 pm | Permalink

    During the 1970s, I worked for a well-known planning consultancy (sadly now gobbled up by a bigger entity). We played with several ideas for localising public transport.
    One such was to cut the number of stations along Melbourne railway routes, but to base minibus services at each station. The routes of the minibuses would be decided by computer. Each pickup/dropoff point had a pad where a customer could nominate a destination point. A centrally based computer would identify the nearest bus or the one providing the quickest route to the destination.
    Another was to build a monorail grid (similar to the Sydney one) across Melbourne. Customers would pick a pickup/dropoff point and a computer in the railcar would compute its most efficient route between origin and destination. A trial system was mooted for Adelaide, but unfortunately, Mr. Whitlam’s government was booted out and a day later, the Cities Commission (which had sponsored the proposal) was no more.

  • 10
    Patrick Bateman
    Posted February 2, 2012 at 4:56 pm | Permalink

    A huge issue I see is the lack of space for anything other than new bus lines. Australians will get all “The Castle” on anyone who dares to propose cutting swathes through suburban city areas for new train/tram/whatever lines.

    Another big issue I see is that Australian public transport operators are yet to work out how to quickly and easily explain to people which vehicle/train they need to be on to get from A to B. Compared to, say, the NY metro or London tube the maps etc available in Australia are a disgrace. In Adelaide I defy anyone to work out how to get between two suburban locations – or even from the city to a particular suburb – by public transport in less than 10 minutes.

  • 11
    Karey
    Posted February 3, 2012 at 2:21 pm | Permalink

    This idea is not new – contrary to the claim that the “‘hub and spoke’ pattern … [might have been] fine a hundred years ago”, older European cities like London and Lisbon adopted grid patterns for their metro lines years ago. See the London and Lisbon maps of their urban train lines. These both intersect with even more local scale bus networks.

    London: http://www.tfl.gov.uk/assets/downloads/standard-tube-map.pdf
    Lisbon: http://www.metrolisboa.pt/Default.aspx?tabid=271

  • 12
    Alan Davies
    Posted February 3, 2012 at 3:21 pm | Permalink

    Karey: Something grid-like in the London manner would no doubt have been better than hub and spoke back in 1912, but I daresay London had a considerably bigger population in those days than any of Australia’s capitals. My point is the centre of places like Sydney and Melbourne had a very large share of jobs (incl manufacturing), retail, services and a lot of the population at comparatively high densities. Many people relied on walking and the frontier of urban development wasn’t that far out anyway. The hub and spoke pattern, which was largely the sharp end of the country network, worked much better in those circumstances than it does now.

  • 13
    moneypenny
    Posted February 3, 2012 at 5:48 pm | Permalink

    Patrick Bateman – Adelaide Metro uses Google Transit for its journey planner. It should take you no more than 60 seconds to plug in your starting point and where you want to go. The results are pretty comprehensive.

  • 14
    alex.rosser
    Posted February 4, 2012 at 6:12 pm | Permalink

    Alan Davies writes “Because Australia’s large cities have legacy radial train systems, the grid is likely to look more like a spider’s web than an orthogonal network — while that’s imperfect, the idea is essentially the same.”
    I’m not sure that a spiders web is imperfect. A typical trip would consist of a radial leg and a circumferential leg – similar to an orthogonal network requiring an NS leg and an EW leg.
    In most cities the population density increases towards the hub (CBD) so the spiders web, hub-and-spoke plus circumferential links, is more likely to match existing population distribution, and be better able to take advantage of the existing radial links, than trying to overlay an orthogonal grid.

  • 15
    Paul Wallbank
    Posted February 5, 2012 at 3:34 pm | Permalink

    Alan and Karey, I’d be a little bit wary of citing the London Underground as an example of a “grid” metro system based on Harry Beck’s schematic.

    If we look at a geographically accurate map of the London Underground you’ll find the network is much more hub and spoke than grid and its development is more about historical accidents than any city planning.

    The site below has links that contrast the schematic versus the real map.

    http://diagrams.org/fig-pages/f00022.html

    In an Australian context, we have to remember our transport systems have developed as a result of political imperatives. It therefore shouldn’t be surprising that developments during the post WWII period have reflected the perceived needs of car drivers rather than train, bus, ferry or tram users.

    The interesting thing is how our transport systems and suburbs are going to develop over the next fifty years if we assume that energy is going to be increasingly expensive and credit more constrained than it was over the past quarter century.

  • 16
    Steve777
    Posted February 6, 2012 at 9:56 am | Permalink

    Another force at work besides escalating energy prices is increasing congestion, which has been prising commuters in our larger cities out of their cars, ever so reluctantly, for the past few decades.

    Commuting by car to the Sydney CBD has been impractical for most people since at least the 1970′s. Commuting by car is increasingly unviable in major suburban centres such as Chatswood and Parramatta. Local councils are aiding and abetting this process by means of road closures and traffic calming, forcing more traffic onto clogged main roads, and also through parking restrictions, so that you have a long walk to arrive at your final destination or have to pay through the nose. It would be good if these sorts of measures could be coordinated with public transport planning rather then being done piecemeal to garner local votes.

    Be that as it may, long journey times and lack of parking would make any half decent public transport option attractive. I think ‘build it and they will come’ would work for many suburban centres. I think we should have a trial of a relatively cheap option such as a series of direct bus route across suburbs between major centres. I’m sure that these would attract a lot of commuters.

  • 17
    Tea Chloe
    Posted February 12, 2012 at 12:59 pm | Permalink

    Great piece I especially like the use of the Toronto system at the top. I lived in Toronto and the difference in transport is quite astonishing. I don’t know why so many people seem to think there’s some sort of “Australian culture” the makes pt unviable or how anyone can think that Australians have much more time pressures than people anywhere else in the world. Do we really think we’re busier and more rushed than new yorkers?! I think not. In other cities people have children and do groceries and go to meetings and the pt system helps rather than hinders people’s lives. It’s actually much easier to get around if the train runs every 2-4 minutes. Sure you might have to walk a few blocks but suddenly you don’t mind because your commute is faster and you don’t have to find parking. It feels much more door to door when the bus or train takes you directly in to the shopping centre rather than trying to find parking somewhere at the back of a vast parking lot… I found it much less of a hassle over all. Now I live here and drive and it makes life much harder.

  • 18
    Supsup
    Posted February 15, 2012 at 1:04 am | Permalink

    I think give this writer a big gold star!!! Seriously, it seems soooooooooooo ok freakin obvious but still nothing has been done … It’s great to see it so well set out here in this way.

  • 19
    IkaInk
    Posted May 2, 2012 at 12:14 am | Permalink

    I missed this post. I must say Alan, I thoroughly approve.

  • 20
    Drbuc
    Posted June 11, 2012 at 11:25 am | Permalink

    An easy way to do this in Melbourne, would be to follow the way Berlin builds its U-Bahn.
    For its new line under the main street, Unter-den-Linden, the new line runs directly under the road. For a grid train system in Melbourne, this would be the way to do it. Pick major roads that carry a lot of traffic, and then dig a tunnel directly beneith them that can carry trains. To follow the Berlin example the best idea would be to build tram sized trains, and shuttle them a long the line at a high frequency.
    For instance it would be possible to do this along Bell St, so that it connects up with the Craigieburn, Upfield, Epping, and Heidelburg? lines. If continued it could also cross the Yarra and connect up with Tempelstowe and Doncaster.

    Another location for this idea would be North Rd in Brighton, which would connect up the Sandringham, Frankston, and Dandenong lines, as well as giving Monash Uni Clayton a train connection. This line could keep running to Springvale Rd where another line could be run underneith it.

    Many other main roads could also be considered. And if they have 3 lanes or more each way it would be easy to remove a car lane each direction and replace it with rail, which would be far cheaper than digging underneith.

Please login below to comment, OR simply register here :



Womens Agenda

loading...

Smart Company

loading...

StartupSmart

loading...

Property Observer

loading...