Friday, December 30, 2005

Service frequencies, trip combinations and triangular numbers

Consider a village (A) that has two round bus services a day to the next town (B).

The timetable might look like this:

Depart A 7:00am
Arrive B 7:30am
Depart B 7:30am
Arrive A 8:00am

Depart A 7:00pm
Arrive B 7:30pm
Depart B 7:30pm
Arrive A 8:00pm

Assuming that no overnight stays are desired, the number of trip choices from either place to the other is one. Thus the service is inflexible and would only cater for the minority of travellers for whom the timetable was suitable, or those who had to make the trip and had no other alternative.

Now supposing an extra trip was added so it departed A at 1:00pm, called at B at 1:30pm and returned to A at 2:00pm.

The level of service has increased by 50% (3 services instead of 2) but what has happened to the travel choices possible?

The answer is that instead of one choice only, there are now three combinations available for visitors to B from A (leave 7:00am, return 2:00pm, leave 7:00am, return 8:00pm, or leave 1:00pm, return 8:00pm).

Adding a fourth trip (at say 9:00am) increases choice dramatically. Double in fact, so that instead of three possible combinations there are now six.

Similarly five trips allows ten choices and six trips gives fifteen choices.

Maths boffins will see a pattern in this. Each extra service increases the number of combinations added by one. Hence the pattern (1), (1+2 = 3), (1+2+3 = 6), (1+2+3+4 = 10) and so on. As a point of trivia note that all answers are 'triangular numbers' due to the pattern formed (below).

O Total = 1
OO Total = 3
OOO Total = 6
OOOO Total = 10
OOOOO Total = 15

At this level a relatively small service increase (say two extra services) dramatically increases travel choices and the practicality of using public transport for many trips. This is something to bear in mind when planning services, particularly in areas that may currently get only one or two a day.

Thursday, December 22, 2005

Merry Christmas, and good luck catching the bus!

If you need to catch public transport in Perth on a public holiday or over the December - January period, it's simple. With few exceptions trains and buses run Sunday timetables on public holidays, so services connect as they would on a normal Sunday.

Transperth has consistent service arrangements between trains and buses, and across all bus companies. This is a key advantage of having a consciously-planned integrated network, with co-ordinated routes and timetables.

Melbourne's system has a degree of integration through thanks to the unified Metcard ticketing and (now) Metlink signage, advertising and websites. However beneath the gloss, it is still correct that is 'the system' is still best described as a loose grouping of seperate self-contained fiefdoms, each of which do their own thing.

Since the departure of National Express, the biggest fiefdom is the partially reintegrated train network. This is followed by our world-reknowned tram system. Then there are the bus companies, large and small.

A consequence of this seperation is that timetables generally make sense from an individual operator perspective but not a network point of view. Hence it is not uncommon for buses every 20 minutes 'connecting' with trains every 15 minutes. Any connection achieved owes more to good luck than good planning. At other times duplication may cost-effectiveness and service quality may both suffer.

Never is this mess more obvious on public holidays and the Christmas-New Year period, where all combinations of service arrangements coexist.

There is a measure of conistency with on public holiday timetables for trains and buses as Saturday schedules are usually used. However differences reappear with the reduced summer weekday timetables, with trams resuming full services a week before the trains.

Buses are an even bigger dogs' breakfast. For instance, some bus routes operate on Sundays and not public holidays. Yet others run public holidays and not Sundays. To make it even more complex, some bus operators are like the trains and trams and run public holiday services to a more frequent Saturday timetable. Reduced summer timetables may apply on some routes, but again the dates for these vary. The consequence of this is that journey times may increase by up to 30 minutes or more, due to a lack of co-ordination. In extreme cases some areas may have be left with no public transport at all for several days.

The variations are so numerous that transport enthusiasts have made a sport of keeping up with it as this Bus Australia thread attests. This is full for those who enjoy that sort of thing, but is likely to be confusing for the ordinary person who wishes to get between A and B.

Given the complexity of modern travel travels and that many trips involve at least one and sometimes two connections, it makes more sense to keep the number of variations to a minimum. For instance there may only be weekday, Saturday and Sunday/Public holiday timetables. One of those timetables should apply on public holidays, with conistency across the network. This can be further simplified by having common 7-day timetables as is largely the case for trains in Perth and Melbourne's western suburbs.

Keeping variations down requires reaching a common agreement on public holiday timetables amongst all operators, be it a Saturday or Sunday service. It is also important to harmonise the dates starting and finishing date for any reduced summer timetable, so that this is uniform for all modes.

Exceptions may still be needed, on some services, but should be route rather than operator-based. A sound reason for an exception may be if a specific route has a particular travel needs. For instance routes that serve a university might operate weekday service on a public holiday if it is still a teaching day. The use of a harmonised network-wide service frequency hierachy (eg 15/30/60 minutes for all routes) can actually increase flexibility here as individual routes can be varied without throwing out major connections while minimising travel time increases.

Standardising holiday services is a small but important step towards establishing a connected transport network. To succeed, this needs to be done well in advance for staffing and leave planning reasons. Thus it is desirable that the Department and operators agree that unity is desirable and start planning now to harmonise schedules for 2006-7 and beyond.

Merry Christmas everyone, and good luck catching the bus/train/tram this festive season!

Saturday, December 17, 2005

Good news on V/Line-Metlink fare integration

The changes, to come into effect next year, are described at Viclink.

Fare integration means an end to the extra cost and inconvenience of buying multiple tickets for the one journey.

It will almost eliminate the problems that visiting country bumpkins have with ticket machines on trams, as they won't need to use them for 99% of trips. Similarly local regional city buses are a mystery to nearly all tourists. Now there's incentive to use them and patronage should increase greatly if propery co-ordinated and advertised.

In any system there will be anomalies and unfinished business. For example it is desirable that private bus routes like 788 (which are integrated for 60+ but not for everyone else) are brought into line given their proximity to Melbourne and the commuter/tourist potential. Ditto for Skybus getting full Met fares (even if Zone 3).

But overall it's a major advance that will make transport cheaper and more usable to more destinations. In fact you'd need to go back to the 1980s to find a fare reform that was more significant.

Monday, December 12, 2005

Webcam of Flinders Street Station

Located at Eureka Tower, Southbank.

view images

Sunday, December 04, 2005

Supply vs Demand-led service provision

The public transport version of the chicken and the egg dilemma seems to be the debate about which comes first; supply or demand.

Transport operators and governments may take the view that as existing services are not overcrowded at the moment, providing more services would produce few returns for the cost. Cost-effective patronage growth would have to come from better marketing, such as 'Travelsmart' programs or increasing charges for motorists.

Consumer and passenger groups counter by saying that people cannot use services that do not exist. Hence service increases must precede patronage growth; marketing alone isn't enough and could be counter-productive.

In anticipating future demand, others such as shopping centre builders and road engineers are also big fans of supply-side planning. Just the act of initial supply helps shape consumption. Provided the product isn't completely to the customer's disliking, the total market for it will probably be bigger than if a more timid demand-side scarcity-based approach was taken. The slogan 'build it and they will come' explains the supply-side planning assumption well.

That these debates apply in areas other than passenger transport was brought home when listening to today's Hindsight program on Radio National (downloads available). The program profiled Gordon Barton, who got his start by challenging (then) government-controlled railway monopolies on freight by forming a trucking company in the 1950s. IPEC grew and became a dominant company in its field.

In 1979 he moved to Europe and successfully established IPEC there. The Hindsight report mentioned that established transport companies waited they could fill a truck before despatching it. Hence it took many weeks to send items a few hundred kilometres. Though not pointed out, this is a good example of a demand-driven approach, ie not running a truck unless the load was sufficient.

Barton's genius was that he sent off trucks on a regular schedule, regardless of whether they were full or empty. Regular departure times allowed connections with other trucks, forming a co-ordinated network. In the early years running nearly-empty trucks around the countryside was expensive. However IPEC Europe provided a superior service, and as the loads increased, it thrived. In this example, Barton showed that his agressive supply-driven approach not only worked but took business from his more complacent rivals.

Friday, December 02, 2005

Public transport more important than coffee bars: 'Creative Class'

I'm currently reading The Flight of the Creative Class by Richard Florida. This book argues that human creativity has become a key driver of economic growth and cities that attract people of the increasingly-mobile 'creative class' will prosper in the future. Though the United States has historically been a magnet for talented migrants, Florida argues that socially tolerant cities in Canada, Australia, Asia and northern Europe are of increasing appeal to the 'creative class', with implications for the US's economic competitiveness.

One finding was that public transport, particularly fixed rail, is very important to the 'Creative Class'.

A common feature of leading creative centers around the world is efficient and heavily-trafficked subway and light-rail systems. The availability of subway and rail transportation was a key factor cited by creative people in the interviews and focus groups for The Rise of the Creative Class, trumping amenities like bike trails, coffee bars, and music venues. (page 201)
Florida mentions that though public transport exists in the US's larger and more creative cities it is sorely lacking in too many others. He goes on to say that:

On this critical dimension (public transport), large cities and regions outside the United States have another powerful advantage in the growing global competition for global talent. (ibid)

In other words, if cities are to attract world-class talent, they also need world-class public transport.

All this is consistent with research by Peter Newman & Jeff Kenworthy (Cities and Automobile Dependence, etc) which highlights the key role of public transport in enhancing a city's liveablity and prosperity.

Saturday, November 26, 2005

Games on public transport

A grown-up version of tag has 'detectives' chasing the 'suspect' around the city on public transport and using mobile phones to give clues. Developed by Toronto graphic designer Joel Friesen, it's called Live Action Scotland Yard (L.A.S.Y.).

L.A.S.Y. has reached high geek status, as testified by mention in Slashdot. Participants have also been warned to prepare for a season of Live Action Super Scotland Yard in 2006.

Closer to home, finding places rather than people is the object of the Metlink Challenge. Sponsored by Environment Victoria and Metlink, it teaches secondary students how to reach places by public transport.

Then there's those who aren't interested in chasing anyone or going anywhere in particular. I'm referring to transport enthusiasts, sometimes known as 'gunzels'. A favourite sport has been to travel a city's entire suburban rail network in a day. In Perth or Adelaide this is simple but it's harder on the larger systems of Sydney, Melbourne and Brisbane. Detailed timetable planning, knowledge of buses, on-time running and some lucky connections are all required. Melbourne Scribe did it several months back and an account may still be on his blog.

Thursday, November 24, 2005

Inquiry into Managing Transport Congestion

The Victorian Competition and Efficiency Commission is conducting an enquiry into managing transport congestion.

Suggested ways to manage congestion are many and varied. The road lobby wants more roads, while opponents say that this is like curing obesity by loosening one's belt. Another, promoted in Melbourne 2030 (though not backed by substantive projects) is to shift more car trips to walking, cycling and public transport. We can also design our towns and cities better to reduce the need for long trips, encourage more to work from home and consider congestion levies, tolling roads or paying more for parking.

Then there is debate over whether congestion is as big a cost as claimed; or maybe it's cited as a reason to fund the pet projects of vested interests. Contrary to a vew of there being a 'car culture', I believe people are fairly pragmatic and will use whatever transport mode that works best for them. If we think the modal split is too lopsided, then making the forms of transport we want to encourage more attractive might be a solution.

More information is here, including an issues paper and submissions already received.

The closing date for submissions is December 2.

Friday, November 18, 2005

Service boost for regional Victoria

The Victorian Government's recent Moving Forward in Regional Victoria statement has some welcome service increases for bus services within and between country towns and cities.

Age article here

For those downloading the strategy document, Action 12 lists the towns that will be getting the new or improved services.

Sunday, November 13, 2005

Housekeeping (updated 2019)
1. A slight change of name; from to . Please update any links and bookmarks.
2. Significant posts from have been moved to . Unfortunately the comments haven't been able to be moved across. This affects about 6 months of posts, from April 2005 (when we started) to October 2005.
3. has been closed down. 

Sunday, October 16, 2005

Timetable tidbits

1. Back in August Adelaide Metro upgraded many of its bus services. Highlights include six extra 'Go Zones', doubled Sunday service on many routes, a 'Mega Go Zone' along the O-bahn and a frequent metro service to Adelaide Airport.

The outer north-eastern suburbs past the end of the O-Bahn seem to be the big winners. Several now have services every 15 minutes. Operating hours have also been extended, with some, like the route 544 series operating until after midnight, even on Sundays and public holidays.

The changes are also covered in the AATTC's Table Talk.

2. Here in Melbourne, outer suburbs are being given some much-needed extra bus services, with $44 million to be spent over 4 years on 50 new or upgraded routes. These services will be introduced to growth areas in Melbourne's west, north, east and south. Yarra Ranges routes to be improved are listed here. Industry and passenger groups say that about five times the amount offered was required to provide sufficient service to make bus travel more convenient.

3. Craig has been busy examining the quality of bus-train connections at important interchanges. This is very important; who wants to be stuck for 30 minutes at a bus stop on a cold, rainly night? Craig's findings are presented here. Particularly note the results for the Smartbus routes (700, 703, 888/9), the design objectives of which included enhanced connections with trains.

Thursday, October 06, 2005

Learning from Perth

It's hard not to be impressed with the public transport improvements that have been made in Perth over the last 15 years or so.

From planning to close its rail system in the 1970s (Fremantle line services ceased in 1979) to restoring, electrifying, modernising and expanding the network from the 1980s, Perth public transport has come a long way. It is a credit to those involved that this was achieved despite the prevailing anti-rail mindset amongst transport planners at the time.

Where major decisions were required, the right choices were always made. The work was completed on time and on budget. Since the early 1990s, patronage in Perth has grown faster than in any other Australian city.

To tell how Perth did it was Prof Peter Newman from Murdoch Uni, who spoke last night at a well attended meeting convened by Environment Victoria.

Unlike in other states, WA has just one minister and one department responsible for planning and infrastructure. One advantage of this is that road and public transport proposals can be evaluated against one another and fairly compete for funding. The effect of this in WA has been a shift of resources from roads to public transport since 2001.

A casual observer might see the Perth rail improvements as being part of a coherent transport plan. After all it appears logical; restoring a defunct service, modernising the whole system, a new line, increasing services, some more extensions, etc. However in practice, the Friends of the Railways (in which Prof Newman was involved) vigorously lobbied for one project at a time, and had no big future vision.

Politically, Labor had been pro-rail, with one of its first acts being to restore the Fremantle service in 1983 after the FOR rallies revealed the strong public support. However since then it has been the Liberals that have moved furthest, from closing the Fremantle line in 1979 to promising a southern line in 2001. Peter also cited a change of mind by former Liberal leader Bill Hassell, who recognised that closing the Fremantle line was his party's biggest mistake and has supported rail in recent newspaper columns. Although there remain party differences (the Liberals are more pro-road and advocated a less direct southern line option), having a measure of bipartisanship in favour of public transport has been a real achievement.

Although state issues such as education, health and police are always important, rail extension was seen as 'building for the long-term' and thus had high public support. Prof Newman also said that the 2004 state election illustrated that it was possible to be pro-public transport and anti-freeway and win politically. Though to be fair, with the southern rail project decided and under construction, there were other issues that dominated debate, such as a proposed water pipeline and economic management.

The view that (at least for longish radial trips) rail was more popular than the bus substitute was substantiated with a 30% patronage loss when the Fremantle line was replaced by buses. This is despite a somewhat more frequent bus service. This loss was recouped when the rail service was restored. Prof Newman also cites the financial viability of the Northern suburbs line, with it breaking even on running cost (it doesn't pay enough to return capital costs, but neither do most road projects).

Perth has also been a pioneer with TravelSmart (direct marketing of public transport to households) and now Transit Oriented Design of suburbs. The 1997 Liveable Neighbourhoods strategy is supported by both sides of politics. It involves new developments around existing stations (Subiaco is the most high-profile success which has doubled station boarding) and also proposed stations (eg Wellard). TOD is also proposed when established tired residential suburbs are being redeveloped.

Peter Newman said that this acceptance may have drawn from the tradition of strong urban planning, dating back to Hepburn-Stephenson in the 1950s. This plan made provision for rail, but this was removed by subsequent road planners. What we have now, with a northern suburbs rail line, is more in accord with the original plan.

Another issue is that Perth did do planning well, but much of it was car-oriented, so its street layout in many suburbs is transit and pedestrian-hostile. Because Perth is such a modern city, more of its suburbs are transit-hostile than (say) Melbourne or Sydney. However it looks as if the will to reverse this is greater in Perth than in other cities. Thus planning by itself is not necessarily a good thing; the question really should be what sort of planning, as bad planning may sometimes be better than none.

Prof Newman strongly emphasises the role of private developers in developing transit oriented developments. Although he didn't mention Joondalup (which is regarded by some as a failed TOD as it didn't pedestrianise enough), he saw great possibilities for the Murdoch precinct and Northbridge (when the line is sunk). None of these would have been possible without the southern railway (direct alignment) providing a catalyst.

Here in Melbourne it is nothing for local government to be pro-public transport and advocate better services to their area. However Peter Newman noted the opposite for Perth, with little support to preserve, restore or build services from local government. There was also no help from the Federal government or transport planners. The general view was that FOR were fighting a losing battle, despite the fuel price hike that year.

The helpful factors were that there was a bipartisan city planning culture in Perth, and that original plans had a place for rail. Research was key, as was community and media support. The bureaucracy were anti-rail before 1983, but came around when government policy changed. Hence the provision of leadership from the Minister and Cabinet was essential to achieving change.

In this vein Prof Newman warned that 'If you reject the Minister and Cabinet they will reject you'. For this reason it was important to tell a story, give credit to strenghten their hand, since politics is a numbers game. Thus transport supporters should also make use of networks in business, bureaucracy and community and build supportive coalitions.

In a statement that might not go down well with local Save our Suburbs groups, Peter saw potential for co-operation between transport planners and developers, including the recently in vogue public-private partnerships. 'They do it with freeways - why not public transport?' suggested Peter. However every Melbourne commuter knows that too much private control can emasculate transport interchanges, as occurred in the 1980s with Box Hill and in 2003 with the botched Melbourne Central redevelopment.

In response to questions, Prof Newman cited the tax system's bias towards cars and driving as an area of federal policy to change. He strongly praised the recent 'Sustainable Cities' parliamentary commitee report and mentioned Malcolm Turnbull as an ally. AUSLINK has the possibility of funding rail transport, but this was so far confined to freight.

Commenting on Melbourne, Prof Newman found that people here were cynical about public transport and the possiblity of improvement. He described our rail system as good but a trifle slow. TODs on railway stations were also needed; he criticised 'some film star' (Geoffrey Rush) for opposing a four storey development proposed for Camberwell Station. When asked about service frequencies, he said this should be sufficiently frequent throughout the day not to require a timetable; a basic 15 (maybe 10) minute frequency was suggested, with 5 minute services in peak periods.

Peter ended on an optimistic note, saying now was a good time to get improved public transport on the agenda. Opportunities to be seized included local government/community interest (Peter is talking transport at the Municipal Association of Victoria conference), the recent higher oil prices and the completion of Regional Fast Rail as presenting valuable opportunities to win ministerial and government support for improved public transport in Melbourne.

Sunday, October 02, 2005

Designing the Metlink style

Sometimes one hears about aspects of public transport in the most unexpected of places.

An example was last night's Right, Left or just Centred debate, held as part of Character, a series of public discussions, displays and tours about graphic design.

The discussion was about the role of the graphic designer; are they artisans who are primarily accountable to their customers (normally fee-paying corporate clients), or do they also have responsibilities as citizens to be mindful of the political consequences of their work?

One of the speakers was Steven Cornwell from Cornwell Design which did the Metlink rebranding.

By any standards the Metlink rebranding is a massive undertaking, ultimately extending to thousands of railway stations, tram routes and bus stops across Melbourne. No two locations are identical, and the number of different signs required must be into the tens of thousands.

The organisational aspects are similarly complex; rather than just one client, there were six (Connex, M> Train, Yarra Trams, M> Tram, DOI and the bus industry). The project will cost $20 million, but uniquely, the operators will be spending it doing up infrastructure they don't actually own.

Stephen Banham from Letterbox described the Helvetica font used in Metlink signage as 'incredibly predictable' and 'banal'. In the world of graphic design, each typeface carries a certain voice. For instance, you won't see 'Wide Latin' lettering on skim milk cartons. Helvetica is seen as being 'corporate', 'homogenising' and 'responsible for blanding the visual environment' and has even spawned a 'Death to Helvetica' movement of which Stephen is a part.

Of course clients will not necessarily embrace the graphic designer's idea of cool, and since the former hold the purse strings, they get the final say. This was true with Metlink whose owners used consumer testing to filter possible designs. Even though this might cramp the creativity of designers, this is not such a bad thing; signage for essential services like transport is better readable and boring than unreadable and idiosyncratic.

Further reading:

Age article

Award for Metlink rebranding

Sunday, September 25, 2005

The patronage clock

Before 7am

The first two hours of services see use by early (mainly blue collar) workers. They may also be used by young late-night revellers who missed the last train the previous evening.

7am - 9am

Peak hour trains and trams are dominated by working age people, with a high student presence as well. The extent of the latter depends on schools and universities en-route, with the major private schools and universities generating the most trips.

Buses remain popular with students travelling locally. At one time they were also heavily used by commuters as rail feeders, but this fell after the 1991 bus service cuts made walking or driving to the station more attractive. In middle and outer suburbs where the rail network is well beyond walking distance (such as Doncaster), buses have a greater role, though public transport modal share is still only around half the metropolitan average.

9am - 3pm

Older people tend to avoid 'crush-hour' trains and trams, even without special off-peak pricing. However after the morning peak, from around 9:30am, the retired form a large component of train and tram patronage and dominate bus patronage. Tertiary students and part-time workers also travel during these hours.

The 10 am to 3pm period is also popular for families with pre-school children to visit city attractions, especially during the warmer months. Highest patronage is during school holidays and/or when there are major events such as the Royal Melbourne Show. A similar pattern exists on weekends, except school-age children are also present.

3pm - 7pm

School children kick off the afternoon peak, around 3pm. Unlike the morning peak (where almost everyone is travelling at the same time) the childrens' afternoon peak subsides just as the workers' peak is building up. Different groups of workers finish at different times, with early-starting blue-collar workers finishing from 4pm, administrators from around 5pm, retail 5:30 - 6pm and others even later. Hence the afternoon peak is longer and flatter than the morning peak and subsides slower, with high outbound patronage continuing throughout the evening. Buses have a very sharp after school peak, but for reasons discussed have a smaller workers peak until most services cease around 7pm.

Though outbound services are consistently busy, there are huge variations in the patronage of in-bound services. For instance weekday trains to the city from Frankston and Sandringham around 6pm tend to be lightly loaded, whereas similar services from Dandenong and Ringwood are heavily used. I put this down to the existence of strong intermediate trip generators at Glenferrie, Box Hill and Clayton/Huntingdale on these lines. Brighton is unlikely to ever become a major trip generator, but possibilities exist for the Frankston line given a station at Southland.

After 7pm

Night services are most popular with young people, students and those returning late from work. Unless there are special events, older people and families tend to be under-represented at these times.

As discussed above, outbound trains are more popular than city-bound trains most nights of the week. The main exception is the 'mini peak' on in-bound services on Friday and Saturday nights. Passengers on these trains are almost all teenagers and young adults.


Weekend patterns are mostly similar to off-peak weekday characteristics. Popular leisure outings are to CBD-based sport, arts and entertainment events, the beach and shopping.

To this should be added a minor peak period of retail workers who need to arrive at work before approx 9am and leave after approx 5pm. Catering workers should also be included but their busier times are different, with lunches and evenings being most significant. This movement is much less than the weekday peak due to (i) the narrower range of occupational groups involved and (ii) the attractiveness of driving due to the limited weekend bus service.

Saturday, September 17, 2005

Government reports on public transport

Two reports on aspects of public transport came out this week.

* Victorian Auditor General's Office
Franchising Melbourne's Train and Tram System

The main theme was that the original train and tram franchise agreements entered into by the Kennett Government were unworkable, and that the current government had no choice but to renegotiate them.

The original contracts were unsustainable because they assumed rapidly growing revenue and shrinking costs. This revenue increase was to be obtained through massive patronage growth and reduced fare evasion. Without service improvements and co-ordination the projected patronage growth was never going to happen. And fare evasion was never going to fall due to our 'open system' and the fault-ridden Metcard automated ticketing system. Metcard's limitations also made it impossible to correctly apportion revenues between operators, as required in the contract.

On the expenditure side, the contracts proposed a shrinking government contribution. Thus the operators would have to meet more of their costs through fare revenue. And unless there were massive efficiencies obtainable from reduced costs, then before too long the operators stood to lose more and more each year.

This could not be sustained so the biggest franchisee (National Express) took their bat and ball and went home. After a period where the government ran the surrendered services, the State Government negotiated new contracts with the remaining operators (trading as Yarra Trams and Connex) to take over the entire system. These new contracts provided a more sustainable revenue base for the operators, more evenly distributed risk and established a central marketing body (Metlink).

These contracts were negotiated by the DOI and were praised in the audit report. It looks as if they will bring organisational stability to the system and we will be freed from the incessant name changes (changing each year in some areas) which confused the public.

The signing and later failure of the initial franchise agreements is a story of greed on both sides. On the one hand the government wished to screw the operators down on cost. The government also wanted the operators to bear much of the risk and lumbered them with a dud ticketing system that hurt their revenue.

Winning the right to run a large part of Melbourne train or tram system would have seen as a filip to the winning company. Operators might even have been willing to use Melbourne as a 'loss leader', in the hope that other, more lucrative, contracts will be won in other cities. Perhaps for this reason, and a belief that gaining market share from rivals was more important than profitability, the bidders were silly enough to sign up and promise huge patronage increases.

In retrospect, the patronage increase promise seemed particularly rash. By agreeing to it, over-optimistic operators violated the business principle that they should only agree to be held accountable to something within their control.

It is that some aspects of attracting and retaining passengers (such as servive reliability) are (mostly) within the operator's control. However many of the factors critical to growing patronage are beyond an individual operators' control.

One example would be that individual rail operators have no control over bus routes and timetables, even though co-ordinated services would increase the number of people with access to a station and thus train patronage. Thus rail patronage is largely limited to those within walking distance of a railway station.

Then there is the case where advertising activity by one enterprising operator benefits the other operator almost as much. This is because passengers do not (and can not) differentiate between tram operator. Thus the operator not spending the money gets the benefit for no expenditure, so is actually financially better off than the operator who spent some money to promote the service. Another form of single-operator marketing might simply shift patronage between public transport modes, and do nothing for overall modal share and thus revenue.

The new arrangments tidy up marketing with Metlink, but do not address patronage-affecting matters such as co-ordinated bus services. It is hoped that these will be the subject of revised bus operating contracts, which are currently being discussed.

* House of Representatives Standing Committee on Environment and Heritage Inquiry into Sustainable Cities .

A significant part of this report deals with transport. Recommendations include improved public transport (including federal funding), removal of FBT tax concessions that reward heavy car use and increasing tariffs on 4WD vehicles not used for primary production (currently they attract a concessional rate compared to other cars, dating from before they became fashionable in the cities).

Friday, August 19, 2005

A legitimate role for car sharing schemes

There has been some recent debate about whether car-sharing schemes should be included on the proposed public transport smartcard. I won't discuss this here, but will instead go into the merits and limitations of car sharing and its likely effect on public transport.

The space for car sharing exists because there is currently a large gap between people's transport choices. For most people, and excluding taxis and hire cars, these boil down to:

(i) No car, use of public transport, walking and cycling only

(ii) full-time car ownership only, or

(iii) Full-time car ownership plus some public transport use

Because of the high fixed costs of car ownership, (iii) is little cheaper than (ii), at least unless petrol prices double or treble. (i) is an option most wouldn't accept.

Car sharing adds another choice to the mix, and unlike full-time car ownership, I believe it is likely to be favourable to public transport. It also makes a useful distinction between car ownership and use and encourages participants to use the right travel mode for the job rather than automatically reaching for the car keys.

Car sharing is a niche market ideal for singles, couples (particuarly childless) and the retired. It is most suitable in suburbs where public transport is good and access to a shared car is a reasonable alternative to universal car ownership.

What sort of trips would car sharers use vehicles for? I would anticipate these would be specialised trips that public transport does not efficiently serve and is unlikely to do so, even with substantial service improvements. These include some forms of bulky goods shopping, one-off home renovation work, many country trips and possibly late night travel.

Though those 'difficult' trips could only be 20% of a person's total number of trips (the rest being commuting or shopping where public transport is convenient), use of car alternatives is so impractical that the need to make such trips is a key inducement to car purchase. Once the decision to bbuy a car has been made (including acceptance of the fixed financial costs) there may be a temptation to use it more, even for some trips that were previously done by public transport.

In some circles there may be concern that car sharing may reduce public transport patronage and thus should be discouraged. In third world countries this is a high risk as car owneship is low. However where car ownership is already high, such an effect will be marginal as few of the abovementioned 'difficult' trips are made on public transport anyway.

In any Australian city, the most likely outcome is a small reduction in single-owner car trips and a slight shift to public transport, walking and cycling. This is because although people have access to a car when needed, it's not quite as easy as grabbing the keys and walking 20 steps to the garage. Chances are the local tram stop will be as closer than the carshare depot and the former is the mode that will be chosen for more trips.

Environmentally, an average of 0.25 (or even 0.5) cars per adult (participating in a car sharing scheme) is an improvement over one car each. Walking/cycling/public transport supplemented by easy access to a shared car could lower transport energy use per capita compared to currently available choices. Car sharing is certainly superior to the so-called 'green' car loans, which are more spin than substance as they do nothing for car-dependence, resource economy, urban amenity, congestion, parking or indeed financial wellbeing.

Shared car ownership challenges many existing attitudes to various transport modes by providing a 'middle way' that may be attractive to many. It breaks down the current tight link between car use and car access and the traffic engineers assumption of universal (full-time) car ownership. Also, because participants of car sharing schemes are likely to use public transport for many trips, it breaks the stereotype of the latter as being a residual welfare services and makes it more a mode of choice.

In local planning car sharing could dismantle the doctrine that houses (and shops) should have a minimum number of parking spots. This has good planning and amenity implications and breaks down the 'concrete jungle' as the need to provide for parking and car traffic is the number one cause of urban blight and pedestrian and transit-hostile local centres.

Though not a substitute for good public transport and unlikely to appeal to a large market, car sharing has a niche role that should be recognised and encouraged.

Wednesday, August 03, 2005

Passenger information old, new and improvised

1. Something old

The side of a tram shelter showing a clear tram map. These maps are examples of detailed yet easy to read passenger information.

2. Something improvised (but better than nothing)

A timetable at a Glenhuntly Road bus stop.

3. Something new

Through the local papers Metlink has started advertising that pocket travel guides will shortly appear in letter boxes.

Saturday, July 30, 2005

A timetable feast

With the opening of Perth's new Thornlie line and service improvements on some of the others, a host of new train and bus timetables have appeared on the Transperth site.

You have until August 8 to view both old and new timetables.

The timetables show four major improvements to Perth train services:

1. 15 minute (or better) clockface service at nearly all stations 7 days a week.

The only stations with lesser service are Loch St, Grant St and Thornlie (on Sundays). Some south-eastern stations have four trains an hour but these are not clockface on Sundays. Before this change only two lines and a small minority of stations received a 15 min service on Sundays.

2. More peak express services on all lines.

Instead of an ABAB alternating pattern of peak and stopping services, there are now more expresses between 'all stations' services. At major stations service headways can be down to 4 minutes but as high as 20 minutes for smaller stations. Perth is a two-track system so expresses cannot overtake stopping trains as occurs on the longer Melbourne lines.

3. Travel times on the Armadale line reduced by up to 5 minutes, with express services running 7 days a week.

This is achieved by stopping Thornlie services at all inner stations to permit express running on Armadale trains.

4. The opening of the Thornlie line. Services are every 15 minutes Monday - Saturday and 30 minutes on Sundays and evenings.

The '15 min x 7day' running is nationally singificant for several reasons. (i) Perth now has consistently higher service levels than other Australian cities, (ii) Apart from some exceptions on the Armadale line, Perth effectively has a '7 day timetable' with consistent service frequencies and depature times and (iii) these improvements were achieved in a city that bans much Sunday trading (reaffirmed by referendum earlier this year).

Transperth is to be congratulated for this substantial improvement to train services which has grown the network while making rail services faster and more frequent.

Now onto buses. Many bus timetables have changed. Thornlie Station is now shown, with one route (228) now starting there. Apart from that there was no wholesale truncation of routes at Thornlie (instead of Cannington), probably due to the desirability of direct services to Cannington via Carousel.

Perth makes extensive use of circular route feeder buses serving stations along the Armadale line. 231/232 was the best of these, with what is effectively a 15 min service through the staggering of two 30 min routes. This achieves the desirable aim of having every train met by a bus.

(this message appeared earlier but was revised after the new Armadale timetable came out)

Sunday, July 24, 2005

Bus info via SMS

Via Matt Cook, an article on a UK service where passengers can get text timetables recently appeared in Australiasian Bus News.

To quote from the story:

To use the service, bus users simply enter the seven-letter code from any bus stop into their mobile telephone. Within seconds they will receive a reply detailing the next departures from that stop, together with the number of the bus service, the time of the message and the name of the selected stop.

What I wasn't sure about is if it provided real-time information or not.

If not, such a service may have some use if people know their bus stop number and can store it in their phone. But for people already at the stop people can look at a paper timetable for free (if there's one there), so what's the point?

Real-time information is more useful but more expensive. To do it accurately would require fitting all buses with GPS receivers and automatic position reporting system transmitters.

A less high tech approach could be to provide a (non real-time) timetable if the bus is running on time, but allow the bus depot's radio operator (who is in contact with drivers) to alter SMS times if services are delayed. Passengers would still be informed of delays but information may not be completely accurate. An example could be '8:45am 627 bus from Stop 45678 delayed 10 min'.

Locally Connex has an SMS Updates service to inform passengers of late and cancelled trains. Passengers can also get train timetables by SMS.

Railway stations normally have 'green button' next service information and PA announcements if services are disrupted. Trams are mostly fairly frequent, so unless there is a major road blockage waits seldom exceed 20 min.

So buses probably need real-time information more than trams or trains. The Smartbus project attempted to redress this by providing real time 'next bus' displays at selected stops. However due to their high failure rate these cannot be considered a success. Also their expense makes it impractical to provide them at all bus stops. Vandalism is a constant risk, and maintenance costs are high.

In contrast, the cost of numbering all bus stops so that they could potentially be made part of a future SMS alert system is negligible. This is especially if it is made part of the current Metlink signage project.

Saturday, July 16, 2005

We're getting smartcards

Firstly the official announcements from the DOI's Transport Ticketing Authority.

The members of the winning project team are Keane Corporation, Ascom, ERG Group and Giesedke & Devrient.

Some background

Melbourne has an excellent integrated multimode zonal fare system that other cities (eg Sydney) would do well to copy. However it's in the implementation where Melbourne has come unstuck, not once but twice.

This history explains why public transport ticketing, ticket purchase and validation is much more contentious in Melbourne than almost anywhere else, where it attracts a much lower profile.

The ticketing problems started with the 'scratch ticket' debacle of the early 1990s. Scratch tickets proved unworkable, fare evasion thrived and the system was abandoned soon afterwards (though not before enough tickets to last for many years were printed!).

Next was the automated Metcard system introduced in 1996. Taxpayers paid dearly for cost and time over-runs. An early assumption that most passengers would buy tickets at retail outlets (as is common in other cities), a design unsuited to passenger needs and frequent machine breakdowns led to public hostility and widespread fare evasion, particularly in the first five years.

More recently, anti-vandalism modifications, better maintenance and allowing tram machines to sell daily tickets eventually made Metcard workable. However, maintenance is not cheap, and even if only a few days are missed, functionality quickly declines. In its FAQ the DOI recognises that Metcard is merely 'tolerable', not 'world class', which it hopes Smartcard will be.

The Smartcard challenge

Provided above is the historical context in which Smartcard finds itself. Its success depends on some as yet unanswered questions answers to which will eventually become known.

These include both machine and human elements, such as the extent to which passengers will adapt to the 'tag on tag off' regime (especially in crowded buses and trams), the design of vending machines for the disposable tickets (given that current Metcard machines do not accept common note denominations for commonly purchased tickets) and handling of breakdowns (given the extra passenger faith required).


Another opinion.

Transport smartcards around the world.

Proposed transport smartcards for Australia.

Friday, July 15, 2005

Is this Victoria's cosiest bus shelter?

With today being almost exactly midwinter and temperatures plunging, passengers are probably thinking about comfortable bus shelters. Here's one found outside Ashleigh House in Bergen Crescent, Sale.

Shelter and sign. There's no timetable but Viclink
shows the stop is on Sale's town route and gets three buses a day.

Closer up. All glass is intact and generous seating is provided.

Inside. Spotless!

Friday, July 08, 2005

London transport blasts

A most horrifying and shocking terrorist attack that has taken the lives of ordinary people going about their business. My sympathies are for the bereaved.

Sunday, June 19, 2005

Composite timetables: Where they're needed

In a previous entry promoting composite timetables, I mentioned that their presence has the potential to unlock the potential of many existing underused bus services.

In contrast, their absence makes bus services appear poorer and more confusing than they actually are. As a result, patronage and revenue are both lower than they should be for the service level offered.

Multi-route composite timetables are most useful for major suburban shopping centres and universities. Typically these destinations are located slightly beyond walking distance of trains and trams (typically 1 to 4km distant), so public transport inevitably means buses.

On their own most bus routes are not particularly frequent. However several may overlap, providing a worthwhile (but not always widely known) feeder service along their common portion. Particularly where the overlap includes a direct link between a major trip generator and railway station, there is great potential for better information (such as composite timetables) to boost patronage.

Practical applications

Examples of major links that would benefit from composite timetables in the Oakleigh-Clayton-Chadstone area include:

* Oakleigh - Chadstone

* Oakleigh - Monash Uni Clayton

* Chadstone - Oakleigh

* Chadstone - Monash Uni Clayton

* Chadstone - Holmesglen

* Chadstone - Glen Waverley

* Chadstone - Box Hill

Other combinations of major trip generators and railway stations amenable to composite timetables (sometimes already provided) include:

* Cheltenham/Southland/Moorabbin

* Footscray/VUT/Highpoint

* Preston/Northland/LaTrobe Uni/Reservoir

* Frankston/Karingal/Monash Uni/Mt Eliza

* Dandenong/Doveton/Dandenong North

* Narre Warren/Fountain Gate

Oakleigh - Chadstone demonstration timetable

Yesterday I produced a demonstration composite Saturday timtable for the busy Oakleigh to Chadstone link. My observations of it were as follows:

* From Oakleigh to Chadstone there are 88 Saturday services, running from approx 5:30am to 11:30pm (note that late evening services are excluded from entering the shopping centre).

* There is an average of one bus every 12 minutes over the 18 hour service span. Service between 7am and 7pm is approx one bus every 10 minutes.

* Because of varying headways on different routes, intervals between buses can vary from 0 to 20 minutes during the day, and are typically 30 min at night.

* The maximum wait anyone would have during the day would be 20 min, but in the majority of cases is less than 10 minutes. Had the passenger relied solely on the timetable for the best-known Oakleigh-Chadstone route (Smartbus 700) they would have had waits of up to 30 minutes on Saturday mornings.

* Because of the high combined service already provided (at least during the day), the likely low level of passenger awareness due to limited passenger information including absence of composite timetables), and the fact that both Oakleigh and Chadstone are in shared fare zones, the opportunity to secure increased public transport patronage in the Oakleigh-Chadstone corridor is exceptional.

* A composite Sunday timetable was also produced, but was of limited value as only two of the ten Saturday routes run and do not share a common route.

Chadstone departures demonstration timetable

A more ambitious project, to produce a timetable for all Saturday services leaving Chadstone, was then completed. I found that:

* Chadstone has approximately 224 bus departures on Saturdays, with service from approx 5:30 am - 12:30am (note previous comments on late night services).

* This is an average of 11.8 buses/hr over those 19 hours. That's one departure every 5 minutes. Between 7am and 7pm, 197 buses run, giving over 16 buses per hour, or better than one every 4 minutes.

* Because of varying headways of the different bus routes, waits can be higher than indicated by an average bus/hour figure.

* Buses from Chadstone pass at least 13 railway stations, with Oakleigh, Holmesglen, Box Hill, Carnegie and Glen Waverley most visited, probably in that order.

* Frequent connections are provided to both the Cranbourne/Pakenham and Glen Waverley lines, though not necessarily at the same stations. Provided people travel between 8am - 6pm waits of 10 minutes or more for a bus travelling to these lines are rare.

* Some instances where bus companies had co-ordinated routes were noted. For instance 800 and 804 both run hourly but provide a combined 30 minute service to some destinations.

* Despite Saturday afternoons being peak shopping time, some timetables reduce frequencies after noon.

* The 224 services were divided into northbound and southbound. 58% were outhbound and 42% northbound. (note that some are both north and south, eg a service that runs to Holmesglen before finishing at Oakleigh, while others go north east or north west so these percentages aren't quite accurate)

* At least during the day on Saturdays current service levels are already adequate to justify expenditure on innovations such as passenger information, composite timetables and marketing, at least for passengers near stations on Cranbourne/ Pakenham and Glen Waverley lines. Sunday services remain subject to previous comments on co-ordination with trains and clockface scheduling.


These examples demonstrate the value of composite timetables in improving passenger information and boosting patronge.

Wednesday, June 15, 2005

Integrated passenger information at Ormond

A photo tour of Ormond Station

1. Station welcome board

This is the first sign seen by people entering the station. It is at the base of the ramp leading up to the station's main platforms. Because of exposure to passers-by, it is ideal for public transport promotional material.

Components of this display include:

a. Welcome to Ormond Station sign

b. Ormond Local Transport Guide (with bus info)

c. Travel times to major destinations

d. Local area street map with bus routes identifed

e. Metlink URL and phone number

f. Rail system map

g. Fares information

h. Promotional material promoting off-peak use

2. Station foyer

3. Fare zone information on ticket vending machines

Unlike the large map and sign away from the TVM, this small sign is within the customer's line of vision and provides location specific zone information. Major destinations are identified by name.

4. Nearest Metcard outlet

Provides name and directions to nearest ticket outlet, right at the TVM to which passengers naturally gravitate. These directions are reinforced on local transport guide maps.

5. Local taxi company numbers

Provided at payphones.

6. Local transport guide


a. location of bus stops

b. available bus routes and destinations

c. approximate bus/train running times and frequencies

d. Metcard outlet and Metlink phone number

The guide pictured is located in the foyer near TVMs, but there are others at the entrance to the station and adjacent to the bus timetable display.

7. Marketing material

Marketing material to encourage patronage by publicising off-peak and non-work trips.

8. Timetables

Bus maps, timetables and local transport services guide provided on Platform 2. Although passed by many people alighting the train (including the peak afternoon traffic), the station foyer might have been a better location.

9. Directions to buses

Bus direction sign in underpass. Also includes Metlink phone and web info.

10. Passenger information at bus stops

A local travel guide installed at a bus shelter. Where stops have no shelter, permission needs to be obtained to place a guide in a nearby shop window.

About the Ormond Project

Project Aims

a. to increase public transport patronage amongst both existing and new passengers

b, to improve connectivity between transport modes

c. more effectively market public transport

d. improve customer service through better passenger information

e. improve fare compliance

Careful design and a staged approach

Passenger information needs vary according to the stage of their journey and their level of experience. Information provided must be specific to what the passenger wants to know at the time and not confuse them with either too much or too little.

Passengers to Ormond receive information as a progressive series of stages as they approach the station or bus stop. Each sequence (or 'thread') has been designed for the following types of passengers:

a. Passers-by (potential passengers who may be attracted to public transport)

b. Passengers catching a train at Ormond Station

c. Passengers leaving Ormond Station to catch a bus or call a taxi

d. Passengers leaving a bus to catch a train

These 'threads' are separate but intersect to form a seamless whole. Not all passengers need follow each stage in each thread. For example, passengers with pre-purchased tickets have no need for zone information on TVMs. However, for those who need it, such information is provided at the exact place where it is required (in this case on the TVM itself). Experience has shown that this is more effective than providing the information on large posters somewhat remote from where it is needed.

Information and reassurance

To provide continual (but unobtrusive) re-assurance, a successful action is followed by its verification or reinforcement. This builds confidence. The passenger then moves onto further steps until their journey is complete.

An example of the information/verification process is the on-platform signage. Passengers see arrows giving directions to each platform and their destinations. One of the platforms (3) is only used part-time and is accessible via the underpass. When they see the arrow for Platform 3 (information), the viewer naturally looks across in that direction. The first thing they see is the large '3' (verification). They see that they need to cross the tracks to reach this platform (information). When they look back to the first sign they saw, they see a note that Platform 3 is accessible under the tracks via the underpass (reinforcement). This gives passengers enough information to reach Platform 3 as they know where to go and how to get there.


Cost of materials has been kept down to approximately three to four dollars per station. Improvements such as larger poster sizes, use of colour, better printing and enhanced protection against the elements would increase this figure and are desirable in the long-term. The project receives no official funding. If considered worthwhile, the project's co-ordinator would welcome opportunities for it to be integrated into other passenger information projects currenly underway.

Saturday, June 11, 2005

Congestion charging

Ari is advocating London-style congestion charges for our major cities. I'm a bit more cautious, and gave some alternatives in my reply.

My fear with CBD-focused congestion charges is that they imply that rising car traffic is mainly a CBD problem and that congestion is the main problem that arises from a transport system that is too car-oriented.

On its own, without better public transport services and capacity, an entry charge would influence peoples decision making. This is fine if is encouraging a shift from car to public transport, but not if it encourages many (particularly from suburbs with limited public transport) to avoid the CBD entirely.

Even worse would be if such a charge was to distort business location decisions away from the CBD and towards American-style 'edge cities' or business parks. These are typically located off freeway off-ramps, remote from public transport infrastructure and are not built for pedestrians. Hence they are less accessible by public transport than inner-city areas, and so encourage people to drive.

An example is when the Coles-Myer headquarters moved from the city to a less accessible suburban location. Before the move most employees took public transport. After the relocation most drove, because public transport was so much less competitive.

Hence the planning and transport outcomes of an agressive CBD entry tax could be precisely the opposite of what we want. Transport is a metropolitan-wide, and not just inner-city issue, and a CBD congestion levy in itself isn't enough.

Instead I'd prefer a package of measures, applicable across the metropolitan area, including:

1. A metropolitan parking levy on parking spots (including those at suburban shopping centres)

2. Abolition of planning requirements requiring developers to provide a minimum number of off-road parking spots

3. In return for 2 above, developers to comply with 'walkablility' principles and contribute towards public transport

4. Changes to salary packaging rules, to include public transport passes and removing perverse incentives to drive company cars.

5. Petrol taxes that decline in proportion to distance from the CBD

6. Caps on the number of parking spots

7. Improvements to public transport to boost services and capacity.

Harmonised and clockface timetables

Once a bus route has been designed the next task is to design a timetable, taking into account things like the route's importance, its likely patronage, connections with trains, user-friendliness, service reliability and resources (buses and drivers) available.

Harmonised headways

Except in peak times, suburban trains and trams run to a regular headway or frequency. On busy bus routes that provide an important link between railway stations and major trip generators that are off the rail network, buses should run at the same intervals as trains. This provides reliable connections.

* A hypothetical harmonised timtable

A bus timetable that has been co-ordinated with train arrival times looks like this:

Train arrives A - Bus departs A

10:05am ----------- 10:10am

10:20am ----------- 10:25am

10:35am ----------- 10:40am

10:50am ----------- 10:55am

Notice that it has constant waiting times, so that passenger can jump on any train and know that there will be a good bus connection. And there'll be no need to rush to a particular train as all will provide a good connection. If a train is narrowly missed, the next one will arrive in 14 minutes.

Total journey time is a constant 30 minutes, assuming a 15 minutes on the train, 10 minutes on the bus and 5 minutes transfer time. Passengers get a reliable fast trip (proportion of total travel time in motion 83%/waiting 17%) and are saved from juggling (or even carrying) timetables.

* A hypothetical unharmonised timetable

An unharmonised bus timetable that has not been co-ordinated with train arrival times looks like this:

Train arrives A - Bus departs A

10:05am ----------- 10:10am

10:20am ----------- 10:30am

10:35am ----------- -

10:50am ----------- 10:50am

The bus connection from the 10:05 train is good, but passengers on the next two trains need to wait 10 and 15 minutes respectively. It's even worse for passengers arriving at 10:50, who might see their bus depart (with almost no passengers on board) as they leave the train. They must then wait 19 minutes for the next bus at 11:10am.

For what should be a reliable 30 minute trip, travel time can be as high as 45 minutes. In this case the time in motion is a low 55% and the waiting time a high 45% of the total trip time.

Though the train service in both examples runs every 15 minutes, the effective service frequency for people making this rail-bus journey is reduced in this last example. Of the four trains an hour, only two offer a satisfactory connection.

If the train arriving at 10:30am is missed, then a 44 minute wait until the next satisfactory connection (11:05am arrival) is required. This wait is more than triple the 14 minute maximum wait required in the first example.

* A hypothetical harmonised timetable for a local route

Train arrives A - Bus departs A

10:05am ----------- 10:10am


10:35am ----------- 10:40am


This example would be suitable for a quiet local route where buses are timed to meet every second train. Though inferior to the 15 minute service, passengers still know that good connections are ensured from every second train.

Even though there are fewer buses than in the above (unharmonised) example, the maximum waiting time for trains that provide a good connection is 29 (not 44 minutes) and their arrivals are at even clockface intervals. As with the non-harmonised example above, the maximum wait for a bus is 20 minutes.

Comparing these two examples show that where trains run every 15 minutes, there is comparatively little benefit to be gained by increasing bus services from a harmonised 30 minute to a non-harmonised 20 minute headway. In contrast, going from 20 to 15 minutes is a massive improvement that should see patronage soar.

This also means that if two nearby bus routes run every 20 minutes, upgrading one to every 15 minutes and downgrading the other to every 30 minutes would vastly improve one while only slightly degrading the other, leading to a net benefit.

* A composite harmonised timetable for two local routes that serve a common road

Two less frequent routes that follow the same road for much of their trip can be timed to provide a frequent and connecting service for people living along that road. The presentation of passenger information on such routes is discussed in the section on composite timetables.

Train arrives A - Bus departs A

10:05am ----------- 10:10am (Route 200)

10:20am ----------- 10:25am (Route 201)

10:35am ----------- 10:40am (Route 200)

10:50am ----------- 10:55am (Route 201)

* Rules and limitations of harmonised headways

The main rule when harmonising headways is that the service frequencies must be multiples of one another. Hence a bus service running every 15 minutes can provide good connections with a train system where the basic service frequency is also every 15 minutes. Buses running every 30 minutes can also provide predictable connections for quieter routes. However feeder buses running every 20 minutes trying to connect with trains running ever 15 minutes (or vice versa) are inefficient and are unlikely to attract maximum patronage.

Harmonised headways are a necessary but not sufficient condition for good connections. At their worst they can mean that bus running every 15 minutes consistently misses a bus also running every 15 minutes.

However improving a connection when headways are harmonised normally involves only a minor tweaking of the timetable (starting all services 5 minutes later or earlier) rather than wholesale changes.

Harmonised headways are of utmost importance when buses are infrequent and/or fulfil a feeder role for trains. As demonstrated here, moving from a non-harmonised to a harmonised headway system can dramatically cut travel times and make the service more attractive.

The main limitation with harmonised headways is that especially where a long bus route serves multiple railway stations, it may not be possible to obtain good connections with each service. Nevertheless it provides a predictabilty of travel that is absent when trying to connect 15 minute trains with 20 minute buses.

Also timetable harmonisation becomes less important when services are less than 10 minutes apart, such as with peak-hour trains and trams. In these circumstances, active transfer management (eg alerting bus drivers of details of arriving trains) may become more desirable to provide efficient transfers.

Clockface (or memory) timetables

Ideally passengers want to be able to travel without a timetable. This means frequent or at least regular timetables so that departures are the same minutes past the hour at all times.

The following service frequencies are clockface: 5, 6, 7.5, 10, 12, 15, 20, 30, 60 minutes. Services will be at the same minutes past the hour so people will be able to memorise them.

The following service frequencies are not clockface: 17, 25, 35, 40, 45, 50 minutes. Passengers are unlikely to be able to memorise the times for these services.

Perth train timetables represent one of the the purest examples of clockface timetabling. Sunday and evening services run every 30 minutes, whereas weekdays are every 15 minutes. It appears as if the planners started with a base 30 min Sunday timetable and added services in between to provide a 15 minute service six days a week. All passengers do is to remember a certain number of minutes past the hour and (peak times excepted) will then know all times for their station without carrying a timetable.

When planning services, clockface timetables should be encouraged and non-clockface timetables are discouraged. The only exception where non-clockface times may be desirable is to provide connections with other services. The most frequent example is when trains are every 20 minutes and a service meeting every train cannot be provided. In this case a non-clockface 40 minute service frequency may be required to provide connections, and is arguably preferable to a 60 minute (clockface) service.

* A harmonised and clockface timetable: Perth's Circle Route

Day/TimeTrain98/99 busHarmonised?Clockface?
M-F 6am - 9am5 - 10 min15 minn/a*yes
M-F 9am - 4pm15 min15 minyesyes
M-F 4pm - 6pm5 - 10 min15 minn/a*yes
M-F 6pm - 7pm15 - 30 min30 minyesyes
M-F 7pm - 9pm30 min30 minyesyes
M-F 9pm - 12am30 minno service--
Sat 6am - 7am15 min30 minyesyes
Sat 7am - 5pm15 min15 or 30 min**yesyes
Sat 5pm - 7pm15 min30 minyesyes
Sat 7pm - 12am30 minno service--
Sun 8am - 7pm15 or 30 min**15 or 30 min**yesyes
Sun 7pm - 12am30 minno service--

(*) In these cases exact timetable co-ordination is not possible, but service frequencies are mostly sufficient to keep waiting times down.

(**) At these times some rail lines and a section of the Circle Route operate every 15 minutes.

98/99 is an orbital route that provides inter-suburban and rail-feeder services. Apart from the somewhat limited operating hours, the timetable is very user-friendly, with a great effort made to co-ordinate services, harmonise headways and use memory (clockface) timetables.

* Headway harmonisation and clockface timetabling of Melbourne's proposed Route 700 Smartbus timetable

Day/TimeTrain700 busHarmonised?Clockface?
M-F 5am - 9am3 - 10 min6 - 15 minn/a*no
M-F 9am - 4pm15 min12-19 minpartial**no
M-F 4pm - 6pm3 - 10 min8 - 18 minn/a*no
M-F 6pm - 8pm10 - 30 min13 - 18 minn/a***no
M-F 8pm - 12am30 min20 - 30 minmostlymostly
Sat 6 - 11am20 min30 minnoyes
Sat 11am - noon20 min15 minnoyes
Sat noon - 6pm20 min20 minyesyes
Sat 6pm - 12am30 min30 minyesyes
Sun 7am - 11am30 min31 - 35 minnono
Sun 11am - 7pm20 min31 - 35 minnono
Sun 7pm - 9pm30 or 40 min30 - 37 minnono
Sun 9pm - 12am30 or 40 minno service--

(*) In these cases exact timetable co-ordination is not possible, but service frequencies are mostly sufficient to keep waiting times down.

(**) The average service frequency (every 15 minutes) matches the trains, but the non-constant headways do not.

(***) Due to variations in train headways during these shoulder periods, headway harmonisation would not be possible. Active transfer management is suggested instead, especially for trains from the city.

In contrast to Perth's Circle Route, which is harmonised and clockface at all times, Saturday afternoons and evenings are the only times where Route 700 services are consistently harmonised and clockface. On weekdays the average headway is appropriate, but is too irregular to be a genuine 'memory' timetable. On Saturdays, both the 30 minute morning and the 15 minute pre-noon services are wasted because though both are clockface, neither consistently connect with trains. Sundays see neither clockface nor harmonised service.

This exercise demonstrates that though proposed service levels are mostly sufficient, they are not necessarily arranged to provide the best and most reliable connections. The use of harmonised and clockface timetables as recommended here has the potential to make Route 700 to become a genuinely smart service that people will wish to use.


Harmonised and clockface timetabling has a big role to play in making public transport more reliable, faster and easier to use. A Perth-style review and revision of bus schedules to improve co-ordination is recommended, starting with the premium Smartbus services.

Friday, June 10, 2005

How to transfer to a SmartBus

Arrive at Chelsea Railway Station, Platform 2.

1. See direction sign for buses.

Note that Route 706 hasn't existed for at least seven years, and that directions to Route 857 (which also serves Chelsea) are missing!

2. View bus timetable at station.

Having timetables at stations is commendable, but unfortunately this one is very dated.

3. Note the Rail Substitute Bus stop, conveniently outside the station.

Our bus stop is now visible on the other side of Chelsea Rd. We'll walk to it in the 'approved' fashion, using the pedestrian lights provided.

4. Cross Station Street (a - b on map below).

5. Cross Chelsea Road (b - c on map below).

6. Cross Station Street again (c - d on map below).

Our bus stop, with a nice new-looking shelter is visible on the right.

7. Arrive at our bus stop.

8. View timetable to look up bus.

Again an old timetable, but a bit newer than the one at the station as it shows Sunday services!

9. Wait (for an unknown time due to the old timetable) and board bus.

Other points:

a. Instead of waiting for three sets of pedestrian crossing lights, here is the unofficial, unsignalised direct path that most transfering passengers take (a - d on map below).

Note that the concrete strips in the road are for signs and are not pedestrian refuges. Although the width may look daunting for some, and could be unsuitable for people in wheelchairs, this crossing point is safer than it looks as car traffic is not high.

b. The community information guide is provided outside the council building midway between two pedestrian lights on Station Street.

A location nearer the railway station and pedestrian signals may have been handier for alighting passengers and pedestrians.

Map of area


Smartbus aims include better passenger information, better connections and greater ease of transfer between services. Though Chelsea is a major point on the 888/889 Smartbus route, it does not appear that Smartbus principles have yet been extended to this end of the route.

It would seem that the Rail Substitute Stop is ideally located relative to both the railway station and the Chelsea shopping strip, and that this location should become the main stopping point for all northbound bus services.

Thursday, June 09, 2005

The ins and outs of frequency guides

It isn't very often that even a local newspaper devotes half a page to bus timetables, or, rather the lack of them.

However the June 7 issue of the Caulfield Glen-Eira Leader did so after a local bus company replaced timetables along Route 246 with service frequency guides as a trial.

With a frequency guide you'd be able to see that buses came every 10 minutes in the middle of the day and every 40 minutes on Sunday evenings, but wouldn't know whether it's 1 or 39 minutes until the next one.

Following an outcry from travellers and MPs, the operator has discontinued the trial and said that full timetables would return.

For individual stops, this is a good outcome. However, there are still cases where frequency guides are the most concise and efficient way to convey travel information. These include:

* Individual stops where he service runs every 5 minutes or less at all times

* City-wide travel guides or websites that include a brief summary of services

* Local area transport guides or summaries

* Tourist and other brochures that include a section on 'getting there by public transport'

Unless services are very frequent, frequency guides should always refer users to full timetables and provide a contact number for additional information.

Tuesday, June 07, 2005

Freeways and induced traffic

Though freeways are often seen as a cure for traffic congestion, the opposite can be more true. For a short time after they are built, freeways can indeed reliveve traffic on parallel roads. However their existence encourages more and longer car trips, so congestion eventually gets as bad as it was before the freeway was built.

High Riser discusses a local Melbourne example, where he compares traffic in Waverley Rd before and after the South-Eastern freeway was built.

More detailed discussions of induced traffic appear here and here and here.

Saturday, May 28, 2005

Making public transport faster

One performance measure of public transport effectiveness is its travel speed versus alternatives such as driving, cycling and walking.

For a fair comparison, total origin to destination travel times should be used; not just in-vehicle times. Thus minutes spent waiting, transferring and finding a parking spot should all be counted.

The ratio between public transport and car travel time is particularly important. During peak times, where origins and destinations are near the same rail line and services are frequent, public transport can exceed the speed of car travel.

At other times public transport is generally slower. If the PT:driving time difference ratio isn't excessive (say less than 2:1) people may still accept this if public transport offers other advantages (eg cheaper, no parking hassles, able to read when travelling).

The real worry is where the ratio is much higher. This is true for many local trips, where a 15 minute drive translates to a 60 minute public transport trip, ie a ratio of 4:1. With this sort of difference, only those who have no choice will persist in using public transport.

To make public transport faster, we need to seek time savings in each component of the public transport trip. These components are:

a. Walking to and from bus/tram stop or station

b. Waiting for service to arrive

c. Ticket purchase and validation

d. In-vehicle travel time

e. Interchange time (includes walking + waiting)

As a proportion of total journey time, a, b, d and e are the most significant. When expressed as perceived time (which seems longer than actual time when one is not moving), b and e are most critical.

Possible improvements for each of these components are as follows:

a. (i) bus/tram stops located at intersections (ii) fast, direct and safe pedestrian access to stops from the surrounding area (iii) pedestrian and transit-friendly urban design {notably permeable street layouts}

b. (i) more frequent services (ii) clockface/memory timetables (iii) timetables at stops (iv) real-time passenger information {so passenger can consider alternatives}

c. (i) pre-purchased tickets

d. (i) tram and bus priority over cars (ii) more direct routes

e. (i) timetable co-ordination {including harmonised headways} (ii) more frequent services (iii) good transfer point design, including fast, direct and safe pedestrian access between all transfer points and to surrounding streets (iv) timetables at stops (v) real-time passenger information {as passenger can consider alternatives}

Friday, May 27, 2005

Transfer points - more examples

The last posting showed Elsernwick as an example of a good transport interchange where passengers could easily transfer between train, tram and bus. South Yarra is another. Although both locations feature a lowered railway and lack level crossings, grade seperation is not a prerequisite to construct an effective transfer point.

A transport interchange typically comprises several bus stops often located outside railway stations. They are provided at major railway stations, suburban shopping centres and universities. Examples of interchanges include Ringwood, Dandenong, Oakleigh, Chadstone Shopping Centre and Monash University Clayton.

Interchanges may be under cover and include a walkway or escalator linking it to the nearest shopping centre or railway station. Through car traffic is either excluded or calmed, so passengers can easily transfer between services.

Buses must generally deviate from their route to pull in to the interchange. This generally improves pedestrian access from major trip generators, but at the expense of operational efficiency and travel times, especially for through passengers. Where traffic light priority is missing or ineffective (such as outside the Monash Clayton campus), a 100 metre deviation into an interchange can add as much as five minutes to bus running times.

Transfer points encompass interchanges but are used here to mean are any location where two routes intersect. In their simplest form they comprise two pairs of stops with easy pedestrian access between them and any nearby railway station. Examples range from many suburban railway stations, major intersections to CBD railway stations such as Melbourne Central. Buses and trams normally 'pass through' rather than 'enter' a transfer point, so they do not slow through passengers.

Transfer points are essential to designing an accessible transport network, but for too long have been overlooked in favour of large, higher-profile interchanges. Rising traffic combined with poor pedestrian access have eroded the usabilty of many. Passengers who arrive in time for a bus or tram connection may find they cannot board due to poor pedestrian links between stops. Such delays commonly extend journey times by 12 to 30 minutes and mean that many short-distance local trips involving a transfer can take up to an hour to make.

Station redevelopments, urban design and local pedestrian and traffic management all make or break transfer point effectiveness. As demonstrated in the Melbourne Central case (discussed later) issues of management and control are also critical.

The following are brief summaries on various transfer points and interchanges around Melbourne.


Broadmeadows is a busy suburban rail terminus and the site of a major shopping centre.

The bus interchange is located immediately south of the shopping centre, but is some distance from the railway station. Passengers transferring between train and bus must cross the busy Pascoe Vale Road at either pedestrian-actuated lights or via a large footbridge. Passengers should allow between 5 and 10 minutes to make this transfer.

The overall urban environment is extremely pedestrian-hostile and no one with a choice would make a transfer here.


Glenroy comprises a railway station and bus interchange with shops to the west. Metlink signage has recently been added. Access between station, buses and local shops is good.

Box Hill

A major suburban centre, with offices, civic centre, TAFE and hospital close at hand. Box Hill railway station is one of the busiest outside the CBD.

Box Hill's concept of forcing passengers to go through a shopping centre to move between train (basement) and bus (top level) is not conducive to passenger convenience or travel speed. The new tram extension terminates some distance from the centre, rather than turning into the mall to be nearer the station.


Glenhuntly is at the south-eastern edge of the extensive network of reasonably frequent train, tram and bus services across Melbourne's inner-southern suburbs. It contains one of Melbourne's four train/tram level crossings.

The ease of transfer between train and the Glenhuntly Rd tram is mixed.

Alighting train passengers wishing to travel towards Elsternwick can board at the stop east of Royal Ave or west of the railway line. Royal Ave is sufficiently quiet not to present a significant barrier to pedestrians. Because there is no pedestrian underpass, and alighting passengers must cross the tracks through gates, the choice of stops is useful when one side is closed due to a passing train. Overall ease of transfer is good, though passenger information is currently limited.

Transfer to eastbound trams is less straightforward. The stop nearest the station is adjacent to the railway line across Glenhuntly Rd to the station. No official crossing is provided at this point, but access is quite easy during quiet times (eg off-peak weekdays and evenings).

The main exception to this is just after a train has been through, which of course is the exact time that passengers will wish to transfer to the tram. When the boomgates are down traffic is banked up. As soon as the boomgates lift there is an uninterrupted stream of traffic, which often contains the tram one was intending to catch.

There are pedestrian crossing lights west along Glenhuntly Rd towards James Street. This requires a walk away from the tram stop and then towards it again. Also, the crossing is not particularly pedestrian-responsive as it is on an excessively long cycle.

This poor responsiveness combined with the distance between it and the stop (approx 100m) means that total waiting and walking time has caused the tram to be missed. Yet a better located crossing and quicker-acting lights could have allowed a connection to have been made, without time penalty for the tram.

Because this poor design increases average transfer times, it increases the likelihood that passengers find walking quicker than waiting for the next tram, particularly at night.

Melbourne Central

This is the CBD's second-busiest railway station. Prior to 2004 it provided direct and convenient access to trams on Elizabeth and Swanston Streets. Redevelopment in 2004 curtailed this easy access to Swanston Street trams. Signage inside and outside the station is almost non-existent, and local knowledge is required to successfully use it as a transfer point.

Melbourne Central is a cautionary tale of how a previously good transfer point can be destroyed by a botched redevelopment that was approved by an incompetent (and subsequently sacked) planning minister. Although the general principle of integrating retail and transport facilities is good, such redevelopments must not harm passengers' ability to transfer. Crucially, control of railway stations and bus interchanges must not be surrendered to non-transport interests who have motives other than passenger amenity.