|Source: Sydney Morning Herald|
Patterns are also fundamental to the design of effective transport networks. A pattern is necessary to make the timetable legible for passengers, to use resources efficiently and to link lines in a way that make connections convenient.
In transport, patterns are measured by time not distance.
Even if the network comprises a single line, passengers expect departures by the ferry (or train or bus) to be at regular intervals - say once every 30 minutes. This can only happen if the time required by the ferry to complete the outbound and inbound leg is a whole integer multiple of the service interval. The less frequent the departures, the more challenging this requirement becomes.
The simplified network below is a 60' minute round trip with four intermediate stops at B, C, D and E and 30 minute headways. Vessels depart the starting point (A) two minutes past the hour and half hour and also depart from the end of the line (F) two minutes past the hour and half hour. As stop C is positioned 15 minutes from both the starting point (A) and the end of the line (F), inbound and outbound vessels always cross at C.
- A,C and F are nodes which are potential interchange points with buses or trains, because vessels travelling in both directions converge at these points at the same time. A single bus could connect at C with both the inbound and outbound vessel.
- If a regular interval is maintained all day, and the vessels are punctual, they will never cross at B, D or E. This means these terminals do not need to be dual berthing, unlike C where the vessels cross all the time.
Realistically, most networks have several lines and they need to connect with each other. This is when patterns become really important.
In the following slightly more complex 30' interval network, four lines connect at a hub, "B". All routes "through line", so passengers travelling from A to C (or from D to E) do not have to transfer at the hub. Although the length of the lines vary, they all conform with the rule that each cycle time is a whole integer multiple of the service interval.
Two lines do not have intermediate nodes (B-C and B-E). Line B-D has one intermediate node and the other has three intermediate nodes.
This network has the potential for linkages with bus lines with timed connections at the nodes. Suddenly, from a very basic underlying pattern, following a very simple mathematical rule, it is possible to create a comprehensive, multi-destination transit network:
Another feature of an integrated regular interval timetable, like this one, is that the pattern leads to other surprising benefits:
- all vessels arrive at the hub at the same time, two or three minutes before the hour and half hour, and all depart two or three minutes after the hour and half hour. This makes rostering much more efficient, because crib breaks can be scheduled in neat, modular blocks.
- as the same pattern of vessel movements is repeated throughout the day, it is easier for ferry masters to develop a regular tempo or cadence. Potential safety risks are more predictable and punctuality is easier to manage.
- it is not difficult to add capacity for special events or to meet peak demand if the extra departures are scheduled at half the interval (eg 15 minute interval instead of 30). This does not disturb the underlying stability of the network structure. Also cruise ship movements can be scheduled mid-ways between the pulse times to avoid interfering with ferry operations.
A pulse pattern in network planning is so important, it should be viewed as the fundamental DNA. This is well understood in Swiss public transport agencies, but sadly neglected in Australia.