The high cost of bad design: Glen Waverley Station and buses

A trip last night on the Springvale Road portion of the Route 902 orbital provided some lessons on good and bad transport interchange design.

The good example is Nunawading, in Melbourne’s east. The rail was sunk (to remove a level crossing) and a new station built a couple of years ago.

The bad example is Glen Waverley, which was last redeveloped in the 1960s. This was lauded as a good example of a shopping centre / railway station development in the Victorian Railways staff magazine of the time. However, looking back, it serves as an example of what not to do.

Nunawading

What’s good about Nunawading? The station has exits directly on Sprinvale Road. The bus stop for the station is just another stop. The bus does not leave Springvale Rd. Passengers on it are not delayed. Hence it efficiently serves both passengers transferring from train and those making short local through trips, such as from Forest Hill to Doncaster Rd. Bus operating times are also reduced, making it possible to provide a direct, fast and frequent service for a given number of buses.

The ability to efficiently serve multiple trip types increases patronage and contributes to the success of the SmartBus orbitals. It also makes for a more versatile public transport network, better able to cater for the majority of trips that don’t involve CBD travel.

The old Nunawading

View Larger Map

Glen Waverley

About 30 minutes south of Nunawading is Glen Waverley. Unlike Nunawading, this is a rail terminus (though some would like it extended to Rowville, possibly via a tunnel). Most important for this discussion is the distance of Glen Waverley station from Springvale Road.

The 1960s redevelopment placed a large car park between Springvale Road and the station, with the intention that this be used by commuters and shoppers. However in doing so it severed the station from its main potential north-south catchment for buses, bicycles and pedestrians – Springvale Road.

This short-sighted design meant that buses serving Glen Waverley Station must divert off Springvale Road, pull in to the bus interchange, and then rejoin Springvale Road. The requires passage through several sets of traffic lights, lengthening the journey.

Glen Waverley

View Larger Map

Comparison

How much is Glen Waverley’s delay compared to Nunawading? Comparing Route 902 travel times between stops immediately before and after each station provides an indication.

Southbound trips between the stop at Tunstall Av (before Nunawading Station) and West St (after Nunwading station) take about 3 minutes. This is about 1km distance and includes crossing the very busy Whitehorse Road.

Southbound trips between Landridge St (before Glen Waverley Station) and Ingram Av (after Glen Waverley Station) take about 8 minutes. Again this is about 1km with a busy road crossing, so is comparable to the Nunawading example.

For a 1km trip, the five minute time difference is quite stark. Having to leave and rejoin Springvale Rd slows the bus almost down to walking pace for the better part of ten minutes around Glen Waverley. This makes it unattractive for local trips not involving the station. Whereas Nunawading’s arrangement imposes no such delay.

Cost implications

As well as wasting passengers’ time, the operational cost imposed by inefficient off-road interchanges such as Glen Waverley’s is considerable.

Back-of-the-envelope calculations indicate substantial increases to bus running costs, which would be better spent on improving service. That’s even with a five minute delay, which would likely be higher during peak times. And the better the bus service, the higher the cost.

Route 902 has about 70 trips each weekday in each direction via Glen Waverley. A five minute penalty for each trip costs 700 bus minutes or over 11 bus hours per weekday. If a bus costs $100 per hour to run, the daily cost of the delay (not counting foregone ticket revenue from lost patronage) would be over $1000 per day, about $7000 per week or over $365 000 per year.

When multipled by the longevity of the project, the overall cost reaches into the millions. This does not include the effect on other bus routes. Nor the amount required to remedy, for instance to construct a redesigned railhead featuring a station fronting Springvale Rd along which buses would remain.

I think the main lesson from Glen Waverley is that some projects can seem a good idea at the time but impose high future costs and hobble future network development for generations.

Taj Mahals or stopping points – the role of suburban stations

Though victimisation rates indicate otherwise, there is a widespread perception that people feel most safe in their own homes, somewhat safer on the street and less safe at railway stations (especially at night). If the latter holds back patronage, it can become a self-fulfilling prophesy.

Passengers frequently call for staffing, toilets and better waiting areas at stations.

And train operators face multi-million dollar annual bills to fix vandalism and clean graffiti at stations.

A way to approach these issues is to reappraise the relationship between stations and their surrounds. Relationships can be human (eg through station staff and/or ‘friends of’ station groups) or physical. Today I will discuss only the latter, since most Melbourne stations are neither staffed nor adopted by a friends group.

Go back 60 to 90 years and stations were hives of local activity. Every significant Victorian town or suburb had one. Visitors, goods and news often arrived there. Trains’ modal share was higher then than now. And there were more staff – including signallers, guards, maintenance, porters, clerks etc. Not suprisingly station facilities had to be large enough to accommodate all this activity.

While lengthening commutes and more recent patronage rises has grown the railway’s absolute contribution to the transport effort (as measured in passenger kilometres) their role has narrowed relative to that of cars and trucks. Railways in Victoria are now almost exclusively passenger concerns. And in Melbourne this is heavily skewed towards CBD area commuting, which while substantial, accounts for a minority of work trips. (Whereas trams tend to be used for diverse purposes throughout the day and local buses have a large ‘captive ridership’ role).

The only interaction that many who drive to work in the suburbs have with the railways is waiting at boom gates or hearing news reports about rail crime. The latter may give rise to perceptions that stations are unfamiliar, hostile, and unsafe, unused by ‘people like us’. This is reflected in personal safety concerns on trains and at stations, particularly at night.

How can one ‘lift the veil’ and improve perceived station safety to be no worse than any other public place? And what about other passenger concerns like toilets, staffing, information or nicer waiting areas?

Some of the best bus and tram stops comprise a simple seat under a shop verandah. They are of the street, not separate to it. Access time to local facilities (including retail ticket outlets) is measured in seconds, reducing travel times and the chance of getting rained on. Public toilets may be nearby. And no one complains about their lack of staffing.

In contrast some other types of stops, like mid-road tram safety zones, have no shelter and require passengers to cross a road. Bus interchanges may be off the main street and have limited facilities. Railway stations, especially if ringed by parking, billboards or (now) over-sized buildings may be similarly cut off.

Could the railways’ quieter stations take their cue from bus and light rail (eg Route 109’s Port Melbourne terminus)? Is there scope to give up the concept of ‘station as place’ in return for more open platforms integrated with surrounding preferably active streets (which may be seen as safer than an unattended station)?

Station - streetscape integration may require knocking down walls, removing unused buildings, taking down dividing billboards and access that puts passengers before cars. More open layouts make stations less of a mystery to non (but potential) users, and less forbidding at night. But it’s not one size fits all as vacant station buildings could be offered to community groups instead (as sometimes already done). In both cases, the community, accustomed to seeing stations as eyesores or magnets to crime, might then start to take a more charitable view. Even at the same station the differences can be marked; Mentone’s Platform 1 integrates well with the surrounding area while the Platform 2 side is shielded by billboards and parking.

If moved from the station’s fare paid area to the street outside, facilities like station toilets could serve both. Facing an active street rather than a railway could improve passive surveillance. It might be possible to involve the local community more in their siting and management, with the proviso that any relocation remain convenient to train passengers. In quieter locations one toilet (open for more of the day) could replace two and any savings used to increase the number of stations with toilets nearby.

Information is another area where transit and community needs can be brought together. Precinct maps and wayfinding signage can promote local shops and attractions as well as directing arriving passengers to buses and surrounding streets. Urban design and arts projects can strengthen these ties, with the ideal being a natural intuitive flow with signage merely consulted for confirmation and cross roads only minimally impeding access.

Some of the above is more relevant to smaller stations, preferably with edge rather than island platforms. Busier stations with more lines will always remain places in themelves and justify their own facilities and staffing. However their interface with the surrounds remains extremely important to their success.

The photos interspersed above are two stations (Grange in Adelaide, Kellerberrin in regional WA). Though minimal they appear to serve the area’s needs and interface reasonably well with surrounds.

Why even culs-de-sac with walkways are bad

There's been some debate on culs-de sacs. We all know they're terrible for pedestrian permeability.

But is providing a walkway at the end of them a panacea? Some have suggested that this provides the best of both worlds - safety for children due to the removal of through car traffic - and permeability for pedestrians.

While better than culs-de-sac without end walkways, they present some problems. Overall I think they're inferior to a grid streets (with narrowing and traffic calming where necessary).

I give five reasons why in my response to the previous link:

1. Opportunities for graffiti/vandalism. Cul de sacs create blank side fences. These are a canvas to vandals. Urban design that minimises these is good (for the same reason there should always be verges, roads and then houses facing railways – never back fences).

2. Poor passive surveillance compared to a continuous street with house frontages. Increase opportunities for vandalism and assaults against pedestrians. Criminals think they can ‘get away’ with more if there are no sightlines.

3. Low information and legibility. Walkways are less prominent in the street directory and only locals may know about them (compared to continuous streets). Trip planner type mobile apps that only consider streets that cars run on might not have detailed pedestrian access way data.

4. Impermanence. It’s not only in posh areas that cul-de-sacs can be closed. To take a random example, when it was built c1979, the end of Wimmera Court (Werribee) was open.

Now it’s closed and part of private property. Since then a shopping centre opened. Had the end remained open it would only be 10 – 12 min walk from it. Now it’s nearer to 15 min, with the perceived time longer due to less directness. I contend that a legible 10 min walk vs a less legible 15 min walk is a huge difference in walking’s attractiveness and thus its share.

Police and residents may push for closure due to apparent crime problems (see 1 above). But especially if there’s a wider pedestrian access issue that affects others from outside the street, then such calls should be resisted due to its effect on the pedestrian network. But it would have been better not to build culs de sac in the first place.

And I don’t think we should just concentrate on access between schools/shops and houses, even though this carries the higher volume – we should also consider anywhere to anywhere trips – eg kids visiting friends houses.

5. I don’t know if it’s my imagination, but in poorer areas the most derelict houses with the worst gardens, the most number of cars on blocks and sheets for curtains are gathered at the ends of courts. If that’s not factually the case it looks worse around the bowl of a court. Check out Studley Court, Laverton on google. Also front fences on houses at the ends of culs-de-sac can be casualties of wayward cars as well. Though not a believer I do think Feng shui has some good design principles, and I that a house on a cul-de-sac bowl draws the shortest straw.

Bringing Stations to Life: Cafe 3162 at Caulfield

Caulfield Station's Cafe 3162 opened yesterday. It's a project of Bringing Stations to Life, a Metro Trains project intended to generate activity around railway stations.

Cafe 3162 serves Asian cusine, in keeping with the international student demographics of the area. It was very quiet earlier this evening with more staff than patrons. Probably understandable given its newness and the university holidays.

The cafe is built on railway land. It abuts but is remote from Platform 4; the opportunity was not taken to add a serving window from the fare paid area (unlike McDonalds at Box Hill). For in-train cleanliness though this is probably a good thing!

Given that 'Bringing Stations to Life' is about the interaction between the station, its urban environment and community, it's worth a few comments on this aspect.

The position is across the road from Monash University and the Route 900 SmartBus stop. While just metres from these facilities, lack of zebra crossings, poor visibility (especially if there is a bus waiting) and fast traffic make direct pedestrian access poor or dangerous. If one was to use the 'approved' method to reach the cafe, ie via the zebra crossing near 7-11, the comparative advantages of the cafe's proximity disappear.

From inside the building one can see Platform 4. The view is through narrow windows covered in mesh (possibly to retard glare from the afternoon sun). The view is not reciprocated; very little of the cafe's insides can be seen from Platform 4 (at least during the day). It presents as a rusted metal edifice and passengers would have little clue that it was a cafe since its back is turned to the tracks.

Although only a day old, the building is already tatty. Graffiti (on the street side) appears easy to do and hard to remove due to its rusted surface. Maybe in time this will give what some may call 'character'.

What would I have done instead at Platform 4 Caulfield? Firstly I'd be inclined to demolish all billboards, walls and existing buildings that present a barrier between the station and the shops on Sir John Monash Drive. There would be ticket machines, timetables, toilets and perhaps a remodelled waiting room, but little else.

Secondly I'd have a second station entrance, located on the down end of the platform. Positioned at the current site of Cafe 3162 this would allow more direct access to Monash University and the Route 900 SmartBus stop. This would be aligned with a second zebra crossing and traffic calming on Sir John Monash Drive, slashing train to bus access times from about 2 or 3 minutes to about 30 seconds.

Thirdly, as part of a broader plan for the area, I would remodel access to Caulfield Plaza, providing a proper footpath between it and the station precinct.

So what's the verdict? I don't think implentation has been ideal at Caulfield due to reasons and missed opportunities outlined above. Nevertheless there is no doubt that the concept of 'bringing stations to life' is a very good one and rail operators have a constructive role to play.

The concept encompasses matters such as surrounding land uses, precinct design, passive surveillance and pedestrian access. Most of all it aims to make station precincts versatile multi-purpose spaces that nevertheless remain efficient as transport facilities.

I sometimes think that in the past stations have sometimes (unwittingly) erected barriers around them, and this gives rise to some of the negative personal safety perceptions.

In some local station cases a simple un (or modestly) fenced platform and shelter, accessible from every direction and surrounded by supportive land uses offering passive surveillance may be the way forward (similar to the Port Melbourne tram terminus of Route 109). Yes ticket checking will be harder and there may be a risk of people jumping fences. But bringing platforms closer to people have urban amenity benefits as well.

In contrast heavy concrete structures like at Moorabbin, Roxburgh Park or Sunshine distance a station from life. These rank amongst Melbourne's least attractive stations to wait for a train, as well as being a blank canvas for vandals. Grade seperations as seen at Huntingdale, Oakleigh or Boronia, though marvellous from a transport efficiency viewpoint, further isolate station platforms from main streets and thus urban life. And multiple levels create shadows which lessen passive surveillance and require more intensive policing than a single level designs with no walls and only a high canopy roof.

There are sometimes tensions between efficiency, safety and design aesthetics. Nevertheless even events such as Sunday trash and Treasure Sales (eg opposite Bentleigh Station) demonstrate how simple things bring station areas to life. It folows that a key aspect of station design is the provision of adequate space near stations (preferably open access and multi-purpose) to allow such activities to thrive.

Urban plans, density and the middle class

Noel Pearson's 2010 John Button Oration (highly recommended - listen here) got me thinking about the relationship between urban plans (such as Melbourne 2030) and the relationship between it and various economic classes (especially the middle class).

To summarise, the intellectual middle class says what ought to be built and the capitalist landowning class goes off and builds, taking a risk in return for a possible profit. While planning regulations (largely drafted and interpreted by the middle-class) have some influence, the capitalist can apply the ultimate sanction, ie choosing not to build if they see no profit. This puts a limit on the strictness (or 'unworkability') of planning regulations, otherwise undersupply will be the result.

Urban planning can be seen as a tussle between a portion of the middle class (either as bureaucrats, academics or activist residents) and the capitalist upper class. The working class has a role as consumers through housing and shopping preferences. There is thus a consumer-producer relationship between the working class and the capitalist (if buying a project home for example), but the relationship between the working class and middle class is very limited.

I previously described Melbourne 2030 as a plan that favours the grouping of certain land uses (cafes, galleries, educational establishments and higher density housing) around transport hubs.

Favoured land uses represent the interests of the humanities-educated portion of the middle class plus some perceived working-class wants (or should haves). The latter being desirable to foster 'diversity' and 'inclusion'. Certain other land uses, such as bulky goods retail, fast-food and (especially) light industrial, are not part of the academic's world so tend not to feature prominently in their plans. Ignoring them does not make these uses go away, instead they go to the periphery without public transport or pedestrian access (Tullamarine and Laverton North spring to mind).

Lacking capital, the interested middle classes (especially academics, bureaucrats and council planners) use plans to advance what they see as the public (or is it their?) interest against the capitalist landowning class. To the extent that they are implemented they advance the former group's quality of life and aesthetic sensibilities. A Marxist analysis would also see city planning as part of a broader contest between intellectual and capital, with each group wanting their strength to prevail.

The middle class, even its humanity educated portion, is not homogeneous. As often played out in The Age, it is deeply split on the question of housing density.

There is what I would call the 'technocratic' or 'urbanist' middle class which favours higher densities for a more compact, less sprawling city better able to be served by public transport and other services. A substantial proportion of this group would have lived in or visited overseas cities with higher densities than ours. Some may work in planning. Their arguments are both environmental and economic, with the economic one being an echo of Le Corbusier's vision of cities as being 'machines for living in'. While it is another debate, this group may be quite relaxed about the city's capacity to house a higher population (though being quite well paid they will always be able to afford to live in the city's 'better' areas).

There is common ground between this group and developers (whose profit increases with density, since fixed costs can be spread across more units). The developers need this group to argue their case, while the urbanists need the developers to allow their preferred city form to be built.

Opposed to densification (at least in their backyard) is another faction of the middle class. 'Suburban protectionists' are less enamoured with high-rise and go by names such as 'Save our Suburbs'. This group professes a wish to preserve their neighbourhood's quiet streets, leafy verges and an always-free parking spot in front of the house. High-density is also thought by residents to be associated with crime, especially if 'outsiders not like us' (eg public housing tenants) move in. Hence NIMBYs ('Not in My Back Yard') are vulnurable to accusations of xenophobia or selfishness. While another (though related) debate, protectionists often question the benefits of continued population growth in our big cities, and may propose decentralisation (for others of course) to lessen pressures on our bigger cities.

As homeowners,suburban protectionists' key economic interest is their house value.

This is especially so for the middle-class.

The biggest determinant of wealth in Australia is whether you live in your own home or rent (those who both rent and invest are a small, savvy minority). Poor people are by and large not homeowners, so values are irrelevant to them. Then there's the rich. While they may live in nice homes the genuinely financially independent hold most of their wealth (80 - 99%) in other assets, such as their own businesses, rental properties, shares and the like. In contrast the middle-class person's home is their biggest asset, with most of their other wealth (superannuation) being untouchable until retirement. Hence concern that a development may lower house values is quintessentially a middle-class concern.

The relationship that high density leads to better public transport is also disputed. 2004 - 2008 for instance saw a train patronage boom but no commensurate service increase. And if higher traffic slows buses and trams, their efficiency will fall without priority. Academic support (in the sense that successful public transport is more a function of service planning than density) is offered by Paul Mees of RMIT.

To conclude, planning and the argy-bargy that goes with it, is a game between a protectionist middle class and a capitalist class (legitimised by a technocratic middle class). The middle class has brains but no capital. The capitalists have money and can co-opt as many middle-class minds as necessary. The middle-class itself is divided. Hence we generally get a largely market-oriented plan with some urbanist features such as denser activity centres and protectionist features such a (movable) urban growth boundary.

Pedestrian Access Strategy launched

Read it here

Station Park & Rides - is there a higher and better use?

A term popular in property valuation and development circles is 'highest and best' use. In commercial real estate the value of land depends on the use to which it can be put. If it has high rental, development or subdivision potential it is worth more than land that does not possess these attributes.

Failure to develop land for its 'highest and best use' represents an 'opportunity cost' for the owner. The owner may still profit from its existing use, but the return will be lower. But if the return the owner is getting is at a lower risk, or if it is taken in a non-financial form such as a better standard of living derived from living on a large block, then they may be quite content not to develop further.

But where publicly owned land is used for marginal, low-return uses, it is only prudent that one asks if it has a higher and better use. In this case the utility may be measured in financial terms, social terms or a mixture.

Parking space at stations is a case in point. Its provision may attract passengers who find driving to the station preferable to walking, cycling or taking a bus. The land may be useful to keep in reserve for future infrastructure, such as extra tracks, stabling, passing loops or bus interchanges.

On the other hand, because parking is uncharged, it returns no direct income to the railways and so may incur opportunity costs such as foregone rental income (if the land was developed), lower bus patronage or reduced walkability.

To try to quantify opportunity costs, one needs to assess both capital and income gains possible from a 'higher and better use' for park and ride land. This needs to be compared to the maintenance costs of park and ride, and if a park and ride is being built or expanded, capital costs as well.

These factors vary by area; a park and ride in a high-land value inner or bayside suburb imposes a much higher opportunity costs than using a scrap of otherwise unusable land near an outer or ex-urban station. It's also worth mentioning that even with different types of parking there are differences; substituting high-turnover short-term shopper parking for low-turnover long-term commuter parking benefits local retailers so probably represents a higher and better use, at least for the local area.

While coarse, the following back of the envelope calculations may assist in estimating opportunity costs.

The first thing is to look at the cost of park & ride. The Victorian Transport Plan has a $60 million program. 1700 new spots are mentioned, but these are only part of the program, so to claim $35 000 per spot sounds excessive. However a figure less than half that is mentioned here. The cost in Perth appears to be a bit over $10 000 per bay. Passengers contribute up to a maximum of around $2000 in fares annually, which buys unlimited travel in Zones 1 and 2.

The other input is the land itself. Wikipedia reckons you should allow 30 square metres per average parking spot, including circulation areas and landscaping. In other words, 100 parking spots equates to about 3000 square metres, or 0.3 hectare.

How much is the land worth? This varies greatly depending on the locality, use and zoning. But going on residential values, $400-500k will buy a house on a 600 square metre block in a median suburb. If we take the land component as being $300k per block, then that's a value of $500 per square metre. Land near a station may be regarded as undesirable due to noise and traffic despite its convenience. So it could be worth less - maybe $300 - 400 per square metre. That is unless its development potential makes it more rather than less attractive. Anyway, let's say the 0.3 ha is worth $1 million, which is very conservative.

How does this relate to the 100 parkers and riders? They're getting exclusive use of a million dollar asset for 10 hours per day. Based on a land value of $10k per spot and an imputed commercial rental return of 10% pa, they are getting a 'free lunch'. 10% of $10k is $1000 per year, or over 50% of their fare (which ought to be going onto running the service). And this does not include the capital costs required to build the car park.

Having done the costs, let's get to the opportunities. What can you do with 0.3 hectare, or 3/4 acre in the old money? It's good for five houses on standard 600m2 blocks, fifteen townhouses or villas each with generous courtyards. Stack said units to make a 3-storey block and we're up to 40+ flats.

Assuming $400k per house, $300k per unit, or $250k per flat, that's a developed site value of $2 million with houses, $4.5 million with units or $10 million with flats.

With a 5% rental yield on the above prices, annual rental revenue would be about $100k for the houses, $225k for the units or $500k for the flats. While we have over 200 suburban stations, if just 20 could be developed, the potential rental income would be in the low millions, or the financial equivalent of hundreds if not thousands more commuters buying Zone 1+2 full fare yearly tickets.

Good design could incorporate improved amenity, for instance passive surveillance of the station precinct. Less NIMBY-minded councils may even provide a waiver on car space requirements in return for a portion of the development being for social housing or (say) a womens refuge.

While there would be a loss of parking spots, this would have to be balanced against the increased population with the station on their doorstep who'd likely on average use the train more, and not just for work trips and the gains possible with improved feeder buses and bicycle facilities.

All things considered, it appears that park and ride may represent close to the 'lowest and worst' land use at many station locations and incur opportunity costs that are not always recognised.

RACV calls for bus stop access improvements

The RACV, has been doing some useful work on Outer Melbourne Footpaths and their connectivity with bus stops.

Their investigation got significant coverage in the local papers, such as a story in this week's Mordialloc-Chelsea Independent.

The RACV's decision to lobby on footpath access to bus stops is appealing for several reasons:

The first is that they have found a gap in bus service provision. The buses may run, stops may be put in, signage may be erected, but the paths to the stops is some other department's problem so do not always get built.

The second is that the costs of remedies are low. Hence (unlike say level crossing seperations) there should be no government budgetary barriers to extend footpaths to bus stops.

The third requires one to imagine oneself as a motoring lobby. You may suppport a bus improvement but as it will be paid for by others it's a cost-free way of being seen to be green. Your members would not begrudge this since almost as many as the general population take transit, walk or cycle as well as drive. Plus the choice of footpaths running parallel to roads has no possibility of delaying cars, as could be the case with alternative projects like pedestrian crossings, roundabout removals or roadspace conversions to bus lanes. All things considered it is very good politics.

The only possible criticism of Outer Melbourne Footpaths is its limited scope. Footpaths to bus stops are good and needed. But full access to stops also needs fast, direct and safe pedestrian access across all roads at all times of day. These require crossings, traffic light cycles and urban design that do not always put the motorist first, and accordingly may not always be favoured by their lobby groups.

Despite the study's limitations the RACV has successfully brought an important public transport issue to general attention. For this it deserves passengers' thanks.

Sydney versus Melbourne

A trip to Sydney last weekend allowed me to compare public transport and land use between it and Melbourne. Sydney does better than us in some things and worse than us in others. Below are the top fourteen differences found between the two cities' transport.

Seven things Sydney does well

1. Co-locating major shopping centres near railway stations

Major examples visited include Blacktown, Chatswood, Hornsby, Macarthur, Macquarie Centre and Parramatta. These centres are either incorporated in the station complex (Blacktown, Parramatta) or accessible via a short, wide mall (Chatswood, Hornsby).

Melbourne's best example of a station integrated shopping centre is Box Hill. Greensborough, Frankston, Ringwood (Eastland) and Sydenham (Watergardens) are adjacent to but are less integrated with the station. More common though are our larger centres, such as Chadstone, Craigieburn Town Centre (proposed), Cranbourne, Forest Hill Chase, Northland, Southland and Werribee Plaza which are all away from stations.

2. Density of residential and commercial development around some suburban stations

Much higher suburban densities apply in Sydney than Melbourne, with multi-storey buildings seen at stations 20km or more from the CBD. Parramatta is a key example, having evolved as a second CBD, being handier to the western suburbs than Sydney CBD.

The pictures show development around Chatswood and Milsons Point respectively.

It is worth mentioning though that some centres are highly unbalanced and contain high density residential but not shopping or offices of a commensurate standard. An example is Epping, which should have better shops for its housing density. In this case the development of the car-based Macquarie Centre (now linked by rail) might have affected Epping's viability.

3. Wide operating spans

A plus for Sydney, with the difference most apparent on Sunday mornings for trains. This is because Sydney operates a passenger-friendly uniform weekend timetable, with no variations between Saturday and Sunday timetables. The effect of this is that first trains have arrived in the Sydney CBD by 6am versus after 8am for Melbourne.

The other main examples of good spans are (i) All-night NightRide buses that operates all nights of the week, mirroring the suburban train network, (ii) 24 hour service on some busy State Transit routes and (iii) 24 hour service on the Light Rail.

Countering this are spans for some outer suburban bus routes, which can be more restrictive than Melbourne since we introduced 7 days/9pm service to many of ours.

4. Wayfinding signage, maps and bus route directories at stations

This is done effectively, as shown in the examples below (Blacktown and Macquarie University Stations).

5. Maps in printed timetables

These are of high quality, at least in STA printed timetables. Metrolink bus stops have similar maps.

6. Physical interchange between train and bus

Often you don't even need to cross roads, and, as this example from Lindfield shows, there is sometimes even direct access from the platform.

7. A more versatile and legible City Loop

Trains operate in both directions and there is no midday reversal or different weekend operating patterns. This makes it better for within-CBD trips than ours; a major advantage for them given their lack of our tram grid network.

Seven things Sydney does less well

1. Fares and ticketing

A lack of integrated fares and confusing ticket types was the most important failing observed during the trip. While it maintains a bureaucracy to write 200-page reports about the most trivial of fare rises, the NSW Government lacks the will and ability to fix a problem that other cities like Adelaide, Perth and Melbourne solved more than 25 years ago. Apart from inconveniencing passengers, the absence of a proper city-wide fare system has also led to the abandonment of the partially-installed 'T-card' smart card project (Card reader at Lindfield Station below).

If you transfer between train and bus you will likely need to pay another fare. If you need to break your journey you'll also pay as tickets are good for a single ride and not a time period, as in Melbourne. Certain railway stations, being privately built, attract special surcharges that range from the merely annoying to the rather steep.

Sydney does not have a fare system as such. Rather there are multiple systems for train, airport train, State Transit bus, private bus, T-way, light rail, ferry, monorail etc. There are many different ticket types but each one doesn't do very much. It probably doesn't matter if you make the same trip each way and do not change modes or break your journey, but otherwise travel can be very expensive indeed. Tourists get even more confused, and have to work out the best deal from the pile of fares brochures pictured below.

In contrast, because of its integrated fare system, Melbourne provides all its fare information in a single simple brochure and the best choice for tourists (Daily ticket, usually Zone 1 only) is (a) clear, (b) versatile, (c) widely obtainable, and (d) good value for money.

2. Base frequency of trains

Off-peak Cityrail services commonly operate every 30 minutes (though exceptions exist). Because most major stations get trains on two or more lines, this generally means they receive four trains per hour. However many of the smaller stations on one line or bypassed by expresses only get two trains per hour.

This service level is similar to Adelaide or Brisbane (weekday off-peaks), somewhat inferior to most of Melbourne 3 or 4 trains per hour, and greatly inferior to Perth (4 trains per hour). The capital (and labour) intensive Cityrail network cannot be automatically be associated with a 'turn up and go' frequent service (like a European subway, or even Perth trains during the day) and passengers are more reliant on timetables and journey planning than they should be.

While the flip side of infrequency can be faster service at major stations due to off-peak express running, my own view is that this is only justified once service frequencies are already high (eg 10 - 15 minutes). If this cannot be achieved then my preference is for the Melbourne pattern, ie few off-peak expresses but giving all stations a somewhat higher base frequency.

3. Frequency and spans of some bus services (particularly private)

To be fair limited frequency and span is not unique to Sydney and can be found in suburbs of any Australian city. Also noted was a marked difference between the service on the Parramatta - Liverpool T-Way (wide span and reasonable frequencies) to the lesser offerings available on the T-Way to Rouse Hill. The example below is better than many, but still includes some non-clockface running that may not consistently connect with trains.

4. Information at many bus stops, especially a lack of maps

Even some heavily patronised State Transit stops (eg Sydney Airport) have extremely rudimentary information. Sometimes the only information provided is route number, final destination and times, with details of intermediate stops and maps missing. This contrasts with the maps contained in printed timetables which were praised above.

5. Lack of differentiation of premium and regular bus routes

Again not unique to Sydney, with Adelaide (Go-Zones) and Brisbane (BUZ) showing the way forward. And to be fair it should be pointed out that more Sydney Bus routes offer high service than usual in other cities, so the State Transit name might already be associated with good hours and frequency (much like trams are in Melbourne).

The new Route 10 'Metrobus' service between Kingsford Smith and Leichardt (via the CBD) was particularly puzzling. It is marketed as a frequent premium service for which a timetable is not required (and is not provided at stops or in leaflet form). Service frequency is 10 minutes peak, 15 minutes off-peak and 20 minutes weekend. Buses are air-conditioned and low-floor and all stops have route maps.

This could almost qualify as a premium service except for two things. Firstly the 9pm finishing time is earlier than all the 'regular' routes around Leichardt, which combine to provide a very high quality service until midnight or later. Secondly the 20 minute weekend service frequency does not really qualify as 'turn up and go' and unless services are very frequent a full timetable should have been provided. As it is the combined timetables at stops include all routes except 10. There is also a consistency issue; other regular State Transit routes offer more frequent service but these (correctly) still have printed timetables.

6. Legibilty of bus network in the CBD

Admittedly this suffers when compared to Melbourne trams, which together with their frequent service and the grid street layout, provide the 'gold standard' unlikely to be equalled elsewhere.

Central Sydney is less planned and has a larger number of distinct bus routes versus Melbourne's fewer, and apparently straighter trams. There dominant feature appears to be a 'spine' up George Street with a number of interchanges around the city (eg Railway Square & Circular Quay). Some overall 'network' information at stops would be nice, although to relieve crowding it might be thought better if people making short CBD trips walked instead.

7. Vandalised trains

No pictures will be published, but there was more evidence of external tagging (and even murals) than in Melbourne. Train insides were also dirtier and likely to be graffitied. The ability to see out of train windows is limited, though scratching is less than in Perth. Limited visibility within double-decks, with their split-level design and poor communication between carriages (compared to our Siemens) might increase vandalism opportunities and lower perceived safety. On the credit side though, Melbourne is the nation's capital for lineside graffiti, and Sydney was cleaner in this regard.

Other observations

Noticed in Sydney was much higher (over?)staffing on the rail system. For example most stations were staffed and there were more in each station. Trains still have guards. Transit officers were sometimes seen at stations but never on trains travelled on. Neither were tickets checked.

DDA access in Sydney has yet to catch up to Melbourne, where all but one stations are accessible. However where access was provided this was commonly through lifts (which has staffing implications). Also, unlike Melbourne, many suburban stations have been rebuilt and modernised. This indicates that like Perth, Sydney prefers to knock things down, whereas Melbourne will mostly retain. Similar observations apply to the CBD, where Melbourne retained more of its built heritage than Sydney, and espedially Perth, where any building over 40 years old is a rarity.

The other consequence of this cultural difference is whereas Melbourne will leave a rail based shopping strip alone to lose status and gracefully decline (as shoppers flock to the greenfields car-based mall two kilometres away), Sydney is more inclined to demolish an existing rail-based strip and build a Westfield right there. The Melbourne approach works where there is a robust strip shopping tradition and to preserve heritage, but risks building 'two cities' and entrenching car-dependence as the mismatch between the location of major nodes and the rail system widens. In contrast, Sydney's 'knock down and co-locate' recipe, though it lacks the 'character' of our successful strips like Chapel Street, Puckle Street or Glenferrie Road, represents better transport/land use planning and is preferred in outer areas where existing shopping strips are weak, in decline and can't be saved.

While the same buckled signal cable housings seen in Melbourne was also seen in Sydney, the track in Sydney was of better quality, due to their higher use of concrete sleepers. The times when track work is done is no mystery either, with many line occupations on weekends. My first train trip in Melbourne after returning felt like flying with mild turbulence. Thanks to their long-standing program to eliminate them, level crossings are rare in Sydney but remain widespread in Melbourne.

The trip was a success and many things were seen that are not described above. Thanks to Scott & Damo for their ideas and guidance before and during it.

Melbourne 2030 - is it working?

Any plan to put more people onto public transport has two complementary prongs to bring the two closer together. These are as follows:

* Bring good transport to more people. This may require line or route extensions, longer-running and more frequent services, and bus route reform with key corridors served by Adelaide-style Go-zones. The general aim here is a convenient network of frequent services extensive enough to be within walking distance of most homes and jobs. These matters are within the province of the various transport plans, such as 'Meeting our Transport Challenges' and the 'Victorian Transport Plan'.

* Bring more people to good transport. Amongst other things this requires clustering denser housing, jobs and shops around railway stations, discouraging large-scale developments remote from transit and requiring that all neighbourhoods have walkable, crossable and bus-friendly local streets. These topics, but particularly the activity centres and an urban growth boundary are part of the 'Melbourne 2030' plan.

The two are not independent; achieving the former is easier and cheaper if the latter is done as less route mileage is required to achieve equivalent coverage and resources can be put to better span and frequency instead.

It is important for the long-term relevance of public transport for a plan something like 'Melbourne 2030' to exist and be implemented. Those who think otherwise take Melbourne's rail-based mid-sized suburban centres too much for granted and haven't lived in Perth where the newer stations, mostly located mid-freeway, scarcely have even a milk bar nearby. Hence public transport there is effective for CBD trips to work, but more complex trips (eg shopping after work) will probably requires extra bus trips instead of a simple dash across the road, so are less attractive.

So this brings us to the question - is Melbourne 2030 working?

Bob Birrell's mob at Monash Uni doesn't think so.

His team found that new housing simply wasn't being built around activity centres. Instead it was all happening away from railway lines and transport hubs, contrary to the Melbourne 2030 aims.

To be fair, I should mention that Birrell has always been a critic of denser living, urban growth boundaries and transit-oriented development. Hence the report is a polite 'I told you so'. In this they are like Wendell Cox/Demographia. But unlike developer-funded Wendell, Monash do have academic standing, independence and rigour so they are worth taking notice of.

Even better though is to make one's own observations, and these are documented below.

Exhibit A is Carnegie. This is a Melbourne 2030 major activity centre between two major university campuses, not far from Chadstone Shopping Centre and only a few stations along from affluent suburbs like Toorak and Malvern.

This 5-part video shot last year, shows the suburb's retail and residential building boom. The shopping centre pictured as being under construction is now open. A major new library opened in 2007. Transit improvements that year included extended hours for local bus routes 623 and 627 and the new Route 900 SmartBus along Dandenong Road. There's both development and transit, so count this as a success for Melbourne 2030.

Exhibit B is Frankston, a principal activity centre. Under major redevelopment, this centre is still in transition with many empty shops. Unlike Carnegie it has lost major stores and its retail diversity has declined. For example, Dimmeys, vacated its large premises and is now a small shop in an arcade. And Spotlight has abandoned Frankston CBD altogether, joining a major new auto-based shopping area near Bunnings on Moorooduc Highway. Buses though have gained, with the new 901 SmartBus and longer hours on most local routes.

If 2030 works, Frankston might be a better place to watch a movie or grab a coffee, but as a diverse retail centre it has lost ground. We are not seeking much new residential above CBD shops, and the station's pedshed is poor as there is no exit onto Beach Road from the northern end of the platform.

If we apply Monash's test of whether there is more development further from the station than nearer to it, the shift of activity to the bulky-goods, auto and hardware precinct on the highway do not demonstrate success.

Exhibit C is the Hoppers Crossing area, particularly the shopping area around Hoppers Crossing Station and the Point Cook Town Centre.

The Hoppers Crossing station and retail precinct is pedestrian-hostile and has not thrived. Though designated as a Melbourne 2030 centre, there have been no attempts at denser development, despite its potential, given the station, shops, hospital and nearby university campus. Buses though have benefited; two Hoppers Crossing routes were upgraded to 'minimum standards' (ie 9pm finish time 7 days a week) in 2007.

Point Cook and its shops have grown rapidly in the last couple of years as a heavily car-based development. Its gridlike street patterns are relatively pedestrian and bus friendly. Bus services though limited; after upgrades elsewhere Point Cook is probably the largest remaining area in Melbourne without Sunday buses.

Just like Central Plaza (Altona Meadows) took over from the shops around Laverton Station (now largely abandoned), there is a risk that Point Cook and an ever-expanding Werribee Plaza could take over from the shops around Hoppers Crossing Station. The risk is particularly high due to Hopper's poor walkability and lack of surrounding housing within a 10 minute walk. Much more than Frankston, the Hoppers Crossing station area can be considered to be a missed opportunity at best and a failure at worst.

These examples demonstrate a great deal of truth in the Monash study; despite its intentions, Melbourne 2030 has not materially changed the basic direction of suburban development of new centres. Transit centres that have successfully developed appear to be in high-value inner areas in which housing demand is (by definition) already highest.

Otherwise the economics of higher housing densities (as advocated in 2030) are poor, both from a homebuyer and developer point of view. The further out you go the worse it gets, which explains why the Mitcham Towers development did not proceeed.

From a buyer point of view, one can buy a basic 3 bedroom house on a full sized block in Werribee for as little as $200 000; even less if you look around. This is below replacement cost so represents good buying, especially if you find somewhere so cheap it's only slightly dearer than the land alone. The cheapest established unit in the area (which may or may not be near the station) would sell for (say) $180 000. Hence houses are much better value than units.

To build a new new version of same, the land would cost around $100 000 and construction cost of around $150 000 per unit. A development comprising two units on the block would cost $200 000 per unit. The economics get better when you're building three or more units per block, but even so it will still not be possible to sell a new unit for less than the cost of an established house on a full block, so purchaser value will still be poor. And for the private investor, buying two established $200k houses is likely to return more with less hassle than buying only one house, demolishing then building two units.

To profit from developing you need either high sale prices or low costs/cheap land. Cheap can mean nasty so often new units will be in inconvenient and inaccessible areas such as Tarneit (instead of central Werribee or Hoppers Crossing). While some buyers are willing to sacrifice backyard space for closeness to facilities, in somewhere like Tarneit they get the worst of both worlds. Nevertheless enough people like the sniff of new carpet and paint so these denser places on cheap land still get built, making a net negative contribution to Melbourne 2030 aims.

The classic retiree downsizer will find that selling their $300 000 house and moving into a $250 000 unit will not boost their nest egg after sale costs and stamp duty are deducted. Besides the house would be better for grandchildren to stay and likely appreciate faster in value. So provided they can still look after the house they're better off staying put.

For first homebuyers, outer area houses are hardly dearer than units, so there is no financial imperative to buy the latter. Also from the developers point of view, the low prices on established houses puts a ceiling on what can be charged for new units (it can be a bit more but not $100 000 more). This imposes a tight lid on costs so development is only profitable if land costs are low and big sites are available.

Hence for both buyers and developers the economics of denser housing in outer areas is quite poor, and it's only likely to work in unserved, cheap land areas away from transport. The problem then is that people pay in other ways, through increased car dependence and the atrophy of shopping centres near railway stations. Plus of course the high cost of providing extra kilometres of infrequent, indirect and poorly patronised bus routes.

That is not to say that the principles of Melbourne 2030 activity centres are not worthy, even in outer areas. Hoppers Crossing Station is surrounded by poorly used land that could become improved housing and retail, though probably at the cost of significant land resumption. Bus stop and station pedsheds could be extended through better pedestrian facilities and resuming land to open up culs-de-sacs for better bus routes. And Melbourne 2030 needs to say something about designing bulky goods and industrial areas to be suitable for efficient public transport.

Unfortunately 2030 overlooks these major retail and employment landuses, so betrays its bias towards the cafe-sitting inner-suburban unmanual white collar type (who may guiltily drive to the Nordic IKEA more often than the out-of-mind Bunnings or Spotlight).

The situation is much brighter in the high amenity inner suburbs. A basic 1-bedroom unit might sell for $200 000 to 250 000, while houses might be $500 000 to $800 000 or more. People on average incomes can't generally afford the latter but the area is super-handy so they'll put up with a unit. And it doesn't hurt that many residents are students or professionals who don't care for gardening much either.

Building costs for units are pretty much the same everywhere, innner or outer suburbs. Anything with more than one level (eg 2-storey townhouses with garaging underneath) will cost more. But a new unit or townhouse for $500 000 doesn't look so bad in value to buyers if the house next to it is worth $700 000 (as it could be) rather than $300 000, as in an outer suburb.

And, though the land value is higher, the resale value for the developer is higher still, permitting a better profit to be made. Plus new townhouses will often have garaging, open-plan and two bathrooms that the Edwardian next door won't. Whereas in the outer suburbs most new houses are the now-standard 4 x 2, despite shrinking household sizes. This again means that new units and townhouses in outer suburbs are less advantageous in term of features or better value relative to full-sized houses.

But the situation is more favourable in inner suburbs, and this is why these are redeveloping as someone has identified sufficient buyer appeal to make a profit and hence worth doing.

To summarise, the relative prices of homes and the economics of developing are not helpful to Melbourne 2030 succeeding around outer suburban nodes, which in many cases are only a few kilometres from cheap greenfields land accessible by freeway. In such areas individuals are no better off buying a unit as they're hardly cheaper than houses and are probably not even near the train.

Turning this around may require an unfashionable level of housing-commission style government intervention to resume and develop sites near stations. A problem here is that developers can sometimes be wary of bureaucracy or getting involved in such projects. Plus governments can change or projects can be shelved, so relying too much on them presents a risk bad for business.

Melbourne 2030 also failed because it was not associated with a major capital works program of walking and cycling projects in established suburbs. The only thing it is associated in the public mind is increased density at little direct benefit to existing residents.

Some apparently simple and cheap transport projects like opening culs-de-sacs can cause huge controversy, yet may be essential for efficient buses. Nevertheless this should not be a reason for inaction; bottom-up community consultation clearly supports better urban design and better transit access, and there are many allies in seniors, disability, school and local government that could implement pedestrian amenity projects and assure community support.

Disability access is particularly useful for pedestrians since two birds can be killed with one stone. It will be increasingly realised that although the bus buying program is ensuring that all new buses will be low floor, stops will remain inaccessible. This is because increasing road traffic and a fetish for roundabouts is causing more roads to be uncrossable on foot, let alone wheelchair.

Then there are market-based pricing or tax mechanisms based in incentives rather than direct interventions. I regard vacant land and buildings, particularly near stations, as an opportunity use, crying out for a higher and better use. Sitting on it is not in the public interest; there could even be an argument for vacant land near stations to be taxed more highly than developed land. This could encourage owners to 'develop or sell' rather than hold for speculative purposes.

Density and distribution of jobs is another factor, with a worry being offices in light industrial or bulky goods areas. Offices don't even need the sort of heavy truck access that industrial areas need, and low-density 'business parks' are unnecessary. The high costs of 'free parking', and the opportunity cost this presents (from public transport access to being able to walk somewhere for lunch) needs to be realised.

Clustering around transport nodes is desirable; from a public transport perspective it's easier to move people from many suburbs to one centre than to move from many to many. And with the daily commute the largest distance travel most people often make, it makes sense to aim to increase efficiencies here by worrying as much if not more about 'job densities' than 'housing densities'.

Stimulating public transport

This week the Prime Minister announced billions of dollars in cash handouts and funding for various projects to 'stimulate the economy'. The object appears to be, in Keynes style, to get people spending and thus retain jobs. There are some things like home insulation subsidies, road blackspot fixes and community facilities that might benefit. However the spirit of the personal handouts appears to be that it matters less where the money goes than if it ends up being spent. Besides most of those working aren't too badly off at the moment; lower interest rates and lower petrol prices are bringing relief to the key outer suburban 'mortgage belt' electorates.

But what if you had a stimulus package to spend on public transport? What sort of things would be in it, and what would be missed? Given the emphasis is on short-term spending on quick projects, very large rail and road projects would not be included. And though highly desirable, and necessary to maximise use of public transport infrastructure, off-peak and shoulder-peak service frequency increases involves some recurring funding so might not be sustainable if funded by a one-off 'stimulus measure'. Hence while important, they won't qualify for money from this particular bucket.

So having excluded them, here's my list of top five transport projects that would deliver lasting benefits from stimulus money.

1. Small scale pedestrian access improvements. Each metropolitan municipality might have an average of 50-100 such projects, which might include zebra crossings, underpasses, traffic islands, roadworks, roundabout conversion to traffic lights, cycle modification, etc. These would be focused around shopping centres, bus stops and railway stations, and also larger roads where there are long intervals between safe crossing points.

2. 1000 kilometres of new cycleway to form a new city-wide network, including grade-seperation with roads on major routes. This rate of construction is several times that which currently occurs.

3. Modification of suburban streets in transit-hostile neighbourhoods. This might include widenings (to permit bus movements), modified roundabouts (to permit buses), opening of culs-de-sac (to permit more direct bus, cycle and pedestrian movement) and new pedestrian and bicycle bridges over rivers and creeks between suburbs. The latter would expand the pedsheds of bus stops and railway stations and reduce backtracking. On larger roads, minor works to assist bus and tram movement could also be done, perhaps complementing the worthwhile 'blackspot' program.

4. An expansion of passenger information, with bus information at railway stations and general information in 'main streets' off the system. Travelsmart-style maps at interchanges and in public places would also identify key walking and cycling routes.

5. A program of works to improve train reliability and make the currently fragile network more robust. This might include projects such as accelerated concrete sleepering, better security for often vandalised trains, lightning protection of vulnurable signal boxes, track duplications, communication improvements and more.

A veritable shopping list for sure, but doing them would kill two birds with one stone; 'stimulate the economy' and deliver improvements of enduring value. At least the higher taxes we'll have to pay in the future to support increased public debt will have had something to show for it in this period of building.

Must there always be a 'bad side'?

Go to almost any rail-based retail strip in Melbourne and you'll find a 'good side' and a 'blighted side'.

The good side has the busy shops; the bad side has the empty shops, derelict buildings, tattoo parlours, pawnbrokers, sex shops, secondhand dealers and other low-rent and marginal uses.

The reason is visibility and pedestrian access. A thriving retail shopping strip caters to many needs and tastes. You won't want to go into every shop, but if there's one across the road that grabs you, one expects to be able to walk directly to it and be inside within 20 or 30 seconds. Poor visibility, waiting or backtracking kills the shopping experience and reduces customer count and sales. This is because the number of shops one can impulsively entered in a 30 minute visit to the strip is reduced. If people can't get adequate and accessible variety they they'll go elsewhere (eg a large shopping centre where the traffic problems are left outside in the car park).

This goes some way to explain why shopping strips based around or near traffic sewers (North Road Ormond or South Road Moorabbin) are failures, while those based around smaller streets (Puckle Street Moonee Ponds or Koornang Road Carnegie) are successes. And along the same street, eg High Street Preston, those sections with closer intervals between pedestrian signals win more shoppers than those that don't.

Nevertheless traffic along a shopping street (or more precisely long gaps between pedestrian crossings and long cycles for those that exist) isn't the only enemy of easy access to shops. Railway lines or major highways, running across a shopping street can divide an area, reducing visibility and/or access. Shops require much closer links between them than do houses; customers often go from shop to shop, but people seldom go from house to house in a street. Not only does visibility have to be good, but retail shops need to be almost adjoining and on the same level to maintain interest and encourage the shopper to 'drop in'.

Anything like a major highway or railway line violates the need for good sight and quick access and condemns the less favoured half of the shopping strip to decline (*). An underpass helps, but not if it is indirect and results in extra walking (as many low-grade DDA-compliant ones can be). Overpasses have similar problems. Even the 'cage' appearance of at-grade gates or booms reduces sightlines, even though in 80% of cases these won't be closed and access is faster than the perception of same.

What about grade seperation of road and rail? Local retailers appear to like this and think it will fix all their problems. While there may be other benefits of grade seperation, fixing a divided retail strip is not necessarily one of them. Indeed it could make matters worse; making a two dimensional problem into a three dimensional problem, further 'killing off' the less favoured side.

Oakleigh, Huntingdale, Sunshine and Richmond station grade seperations are all failures from an urban amenity standpoint. All have caused or encouraged significant urban blight in the overpasses' shadow. Even Elsternwick, which I have praised before, has it's 'bad side' of the line, as Reuben points out (and it was that post which encouraged this one).

If grade seperations aren't the cure for divided shopping strips, what is? I don't think there is one given the clear sightlines and continuous access required within a shopping strip for it to function as a whole. In this case it might be better to admit defeat and rezone the 'blighted sides' of shopping strips from 'retail' to 'residential - high density' with mixed uses also permitted.

This has a number of benefits. For a start it increases the pedshed of the railway station and local shopping strip, generating custom for both. While residents of this area will need to cross the railway or road to reach the shops this is acceptable since once over the line access between all shops in the strip will be good. Urban blight will be minimised since the retail strip will no longer be divided and there will be increased pedestrian traffic from local residents. And if the strip needs to be expanded, some houses could be rezoned mixed use or commercial towards the unblocked end of the strip.

(*) Reservoir seems to be one exception - with two active shopping strips on both sides of the railway.

Do we need a public transport architect?

Sometimes later add-ons to a railway station, bus shelter or other infrastructure can detract from its function, amenity or aesthetics. Other times buildings may go unused (eg old goods sidings or kiosks) when they could do something useful, such as bicycle parking or a business. Whether it's an inconvenient entrance, an insensitively located ticket machine, a mid-platform obstacle or an errant billboard, expediency can sometimes prevail over passenger needs or built heritage.

Example: A tale of two platforms

Mentone Station was identified here as a 'high amenity perception' station. It's easy to see why from the photo below. The things that make Platform 1 special include a pleasant treed outlook, visibility of people and shops. A low-profile fence, a gap in the shelter and easy access all combine to integrate the station with the surrounding community. Rather than the platform being seen as a 'fortress', 'isolated' or 'unsafe' the perception conveyed is safe and welcoming with good passive surveillance.

The next photos are also of Mentone, but Platform 2. The difference with Platform 1 couldn't be more stark. Outside is a carpark that is full of commuters during the day and empty at night so it is less active than the shops overlooking Platform 1.

The main damage however is caused by the long line of billboards along Platform 2. This visually cuts the car park from the platform and shops and the platform from the people and traffic outside.

While Mentone is a very safe suburb, the poor nighttime safety perception of an isolated carpark and unstaffed station platform still lingers in many minds and the blocked visibility makes it appreciably worse. The fix in this case is simple; remove the billboards to improve amenity and perceived safety. A previously blocked vista to shops and buses would also be opened.

The example shows that good design can make a difference and that a generally well designed station can be spoiled by later add-ons. There is also the matter of competing criteria; in this case the dominant (only?) design factor in erecting the billboards appears to have been exposing waiting passengers on Platform 1 to the most advertising possible.

With booming patronage, proposed 'tag-off' ticketing requirements, and the need for existing stations, platforms and entrances to carry 2 or 3 times their current passenger throughputs, we will need to rediscover the need for good station design.

To cope with increasing demands, some of the expediencies put in place during the 1960s - 1990s patronage slump/stagnation may need to be reversed. A public transport architect conversant with rail operations and passenger traffic may well be a sensible investment to ensure that stations, shelters and stops are reviewed with throughput capacity and amenity maximised.

More from the Pedshed Series: Pedestrian and transit access around intersections

The areas around intersections are hotly contested on the road network. Frequently public transport comes off second best, with bus stops and pedestrian crossings planned or moved mid-block to add more lanes or room for turning cars.

The video below looks at the corner of Springvale Rd and Wells Rd in Chelsea Heights. This intersection is 30 kilometres south of Melbourne at the start of one of Melbourne's busiest roads. Topics covered include provision of footpaths, traffic lights versus roundabouts, location of bus stops and access to major trip and job generators.

Transport: Visions for a Sustainable Future

Held in the ironically-named 'BMW Edge' at Federation Square earlier tonight this gathering attended by about 400 provided a stage for various academics and lobbyists to say their bit on transport. Thanks to Transport Textbook for the notice.

Here's a quick summary from each of the speakers:

William Mitchell (MIT Design Lab) advocated electric micro-cars aimed to satisfy last-mile 'mobility on demand' that existing bus services generally do not.

Like Flexicar they turn car ownership on its head, with 'swipe and drive' one-way rental replacing full-time ownership. They provide an intermediate position between no and full car ownership. Stackable like shopping trolleys with kerbside recharging depots dotted around the inner suburbs and near railway stations. Unlike Flexicar (which uses standard vehicles) their small size means that they're less useful for some errands that can't be done with public transport such as large-scale grocery shopping and bulky retail.

Rob Adams (City of Melbourne): Argued that nearly all non-housing infrastructure we'll have in 2020 is already built and we can't build our way out of congestion.

Blames the early 20th century Garden City Movement and later motorisation for our low densities which he says are unaffordable. Strongly advocates high housing densities (4-8 storeys) along every major road within about 20 kilometres of Melbourne CBD; this could accommodate an extra 2 million people within the existing urbanised area. These corridors would be served by trams or frequent buses. Housing in between these corridors (about 90% of the urban area) would be at existing densities but 'greener'. Cities such as Barcelona and Curitiba provided examples.

Like other speakers, high density was considered essential due to a claimed relationship between it and good public transport (similar to Newman/Kenworthy). Having single-family houses on tramlines (a common sight 5 - 10km from Melbourne) was considered a waste that this reurbanisation would address and bring more people closer to public transport.

Dr Jago Dodson (Griffith Uni) introduced us to his mapping work comparing fuel dependency with suburban location, transport infrastructure and incomes (called VIPER). When interest rates and fuel prices rose in 2007 the study got an M (for mortgage) and became VAMPIRE.

As would be expected, the lower income outer suburbs had the biggest susceptibility to fuel price increases due to more widely spaced services, poorer walkability and less public transport. Conversely the posh inner suburbs had high incomes, better public transport, lower car usage and thus high resilience to oil price hikes. Hence we had an urban structure that conspired to increase inequalities as cities grew larger and fuel prices rose. While he did not advocate a government housing program he did mention that private developers tended to seek highest returns by developing in (dear) inner city areas instead of outer areas (I assume this referred to unit development; there are plenty of private housing estates on the fringes).

Not mentioned tonight were two points; that mortgage holders are generally not the lowest income earners (most of these rent) and that the outer suburbs can be quite diverse and include high-service nodes such as central Werribee, Frankston and Dandenong where people can live without a car. Comparison between residents very close to those centres and those further afield (but in the same suburb) would have been interesting.

Dr Jago was a Melbourne 2030 skeptic for a couple reasons. Most activity centres are in the inner suburbs therefore they perpetuate existing inequalities, and much high-density development is of low quality and poor energy efficiency (which might even offset saving from reduced car use). A critic of large infrastructure projects, he saw planning, co-ordination and service as important and supported high-frequency buses. He also preferred local road/rail seperations to mega-projects such as road tunnels.

Nicholas Low (GAMUT) started off with a shopping list of transport projects that he claimed were to be in the Premier's transport statement next month. His particular list included: Tarneit railway line to Southern Cross, extension to South Morang, electrification to Melton, what was called Mees-style trains through central Melbourne, 100 new buses, Doncaster and SmartBus improvements, bike paths. Plus road projects like the Frankston bypass (already announced), Greensborough - Eastlink freeway and half the Eddington road tunnel (Sunshine to Citylink via Tottenham, Kingsville and Seddon).

Prof Low saw this list as being biased too heavily towards roads. Instead he (like Prof Currie but unlike Prof Mees) advocated Eddington's rail tunnel plus rail lines to Doncaster and Rowville. This could help stem car use in Melbourne, which had recently outstripped Perth (in average kilometres driven per year). He stated his opposition to the plan (before it has come out), instead favouring a 'no new roads' policy and bigger cuts to carbon emissions (350ppm suggested).

Cath Smith (VCOSS), like the previous speakers mentioned how lucky she was to live in a bikable inner city neighbourhood with trams, trains and buses. However with high rents and property prices such living was less available and affordable.

VCOSS' constituents comprise many who don't drive, the disabled, pensioners and the unemployed. While train crowding was important, it was clear that VCOSS' priorities were elsewhere, especially with regard to co-ordinating community transport, integrating country school buses and, in Melbourne, local bus route coverage. Industrial estates have no or little public transport and apprentices in particular can find jobs difficult to reach due to 'historical' bus routes that skip job growth areas. As with prevous speakers, the interplay between housing, jobs, transport, education and services was mentioned.

Faster DDA roll-out and higher multi-purpose taxi funding was advocated (the ministerial announcement today was praised) as well as minimum frequencies for buses. Every 30 minutes seven days a week was proposed; while this is a doubling of the existing MOTC safety net, it still may not necessarily guarantee connectivity with trains.

On the environment the point was made that transport modal shift was the key to carbon reductions, not more efficient cars. This was contradicted by Robin Batterham who said that our car fleet (amongst the least efficient in the world) could double its efficiency to European standards if only regulation was applied.

After the main speakers were two respondents; Robin Batterham (former Chief Scientist) and David Ettershank (Kensington resident and, like other speakers a self-confessed 'inner-city elite member of the chattering class').

Robin Batternham drew several threads together and saw a general consensus (at least amongst those there, an important qualification given the non-representative speakers and audience). These included support for higher density and a belief that (pace Mees) it is necessary for good public transport. This can be made to happen politically if those in power are convinced of sufficient support.

David Ettershank carried the political theme further. He claimed that the Labor Party was abandoning the inner-city (Andre Haermeyer was mentioned), and that ALP focus groups have such a construct called 'Cranbourne Man' (who always wants more roads), as well as a 'Shepperton Man' (who was not explained) to draft policy.

Questions from the audience were about housing, train over-crowding and one from left-field apparently advocating a high-speed ferry to Hobart. Jago Dodson said good houses cost no more than bad houses and all should have 6-7 stars (in energy rating). Nick Low said that more trains were needed, but we also needed more train paths as well so they could run. Rob Adams said that train infrastructure had long lead times and 100 'green' buses could be put on special bus lanes on major roads overnight. A Vicroads employee suggested an intermediate road network for human-powered and low-powered vehicles. Such a metro-wide network would cost about the same as a single freeway and overcome the main impediment to cycling - safety and cars. However one of the panel suggested kerbside lanes on all roads would be cheaper and provide a more extensive network.

Overall it was an interesting session, albeit predictable given the narrow inner-city pro-density panel. The attendance of 400 (estimated) indicates extreme interest in transport and city policy.