Monday, March 08, 2010

Small usability improvements for public transport

A great deal of transport project funding is already committed through major transport plans whose projects are typically long-term and infrastructure-based. It will be a while until the major ones come to fruition. A lot of the rest is recurrent expenditure.

Our growing population still has unmet travel needs and public transport is being asked to take on a larger role.

Hence there is a real need for transport improvements that are not expensive but can increase the usability, capability and thus patronage of transit services. Such improvements can either be one-off small infrastructure projects or a few extra services inserted at times when existing timetables are lacking.

Below is a list of examples, grouped by benefit. Their benefits are typically both local (by improving access for a particular area) and network-wide (as traits such as coverage, legibility and frequency are improved).

ACCESS AND CATCHMENT ENLARGEMENT

Frankston is a case where the station entrances south of the station is remote from most surrounding houses and further than it needs to be from a shopping centre. Passengers coming from these areas often need to backtrack to board the train, increasing their walk by 300 - 400 metres.

This extra distance is a large percentage of a station's 800 metre pedshed. Adding a northern platform entrance at 'X' would speed access for thousands of users and increase the station's 800 metre pedestrian catchment to include more residential areas off Beach Street.

The map below shows how the northern part of Patterson Lakes is physically near but just outside a station's pedestrian catchment for want of direct roads or walking paths. A narrow pedestrian/cycle bridge across the channel plus upgraded bicycle storage at Bonbeach Station would cheaply extend the reach of the rail network and provide a faster journey time than changing to a bus.

CONNECTIVITY

A common problem faced by transit is how to serve major destinations (such as universities and shopping centres) that are just beyond a railway station's pedestrian catchment. Southland Shopping Centre has a very frequent bus service from Cheltenham Station. However buses may depart from any of three widely spaced stops, so getting the next bus to Southland has always been a case of pot luck. Modifying routes to serve the same stops could allow the full benefits of the high frequency service provided to be realised.

The timetable for Route 517 below is an instance where buses cannot reliably connect with trains due to their unharmonised service interval. The effect is widely varying end-to-end trip times (making the bus a last-resort transport option) and the need for passengers to carefully pick trip times to avoid excessive waiting.

Sometimes slightly less frequency can mean better service. In the example below, reducing 517's Saturday headway from 38 to 40 minutes and Sundays from 55 to 60 minutes would at least provide constant interval connections with every second or third train respectively.

SERVICE SPAN AND FREQUENCY

There are certain times where span and frequency do not represent desired or actual travel patterns.

An instance of the former is the period immediately after the pm peak. Spreading the peak increases both efficiency and patronage. A way to encourage this is to more attractive shoulder-peak train services, such as higher frequency and express services. Critical times could be 9-10 am inbound and 6 - 8pm in the outbound. As stations are already staffed and the trains already exist, the main resources needed would be drivers and maintenance staff.

One area of unserved patronage demand is Sunday morning, where many people attend suburban markets or (sometimes) early sporting events in the city. NightRider buses have finished and the first outbound arrivals in some outer suburbs are not much earlier than 9am. Melbourne is a particularly late starter compared to Sydney, whose trains start much earlier on Sunday morning.

Unlike a frequency increase, bringing the Sunday start of service forward by 60 - 90 minutes represents an increased span. This shortens the period available for track work on Saturday night. In addition the increased span increases staffing costs at premium stations, which are staffed from first to last train.

If the cost of funding the extra 1-2 hours station staffing per week is an issue, it may be possible to find some offsetting savings by closing staff service at non-safeworking stations earlier, for instance just after the last 'up' train rather than the last 'down' on (say) a Sunday night. This would reduce customer service as regards ticket purchase facilities, toilets and perceived safety through staffing on Sunday evening. However based on a greater good argument, it may be that that the new patronage attracted by the new early Sunday trains would exceed those dissuaded from riding the last couple of Sunday evening down trains.

LEGIBILITY

One of my hobbyhorses, familiar to regular readers, the route below would be better off split into two route numbers (even if services continue to through-route via Chadstone).

Legibility is improved if a standard arrangement across all routes applies for public holidays. There have been rapid recent improvements and a majority of bus routes now have similar patterns to trains. However those as yet unstandardised are some of the busiest services in Melbourne, such as the case below:

EFFICIENCY

Addressing access and usability issues such as the above are relatively cheap, although some would involve recurrent expenditure. Costs in all cases would likely be in the low millions as opposed to hundreds of millions to billions for major projects. Nevertheless inefficiencies should still be identified and removed, especially if there exist improvements that provide larger benefits for the same expenditure.

The following are examples of routes that were useful when they were commenced but appear to no longer serve a useful function as new or upgraded alternative routes now exist.

Sources: Timetables - Metlink Victoria Pty Ltd, Maps - Melway Publishing 2007 & Google Maps.

1 comment:

Dave said...

All valid - but will we see any of these? Timeframes for change seem so slow, even for the low input cost changes, eg station access (with myki, are gates required? -surely just scanners distributed along the platform or at each gate) or timetable co-ordination (admittedly, that can get difficult). But for each of these examples you've provided there will be dozens more of the same, and who is tasked with identifying and resolving the small things that cumulatively create substantial change?