|Common Work Packages|
|BZE Work Packages|
|PTUA Work Packages|
|SDGs Sustainable Development Goals||SDG11 Sustainable cities and communities|
|Published by||Patrick Sunter|
|License||CC BY-SA 4.0|
|Translate to||Français, Español, Kiswahili, 中文, العربية, Русский, more|
|Export to||PDF, LaTeX, EPUB, ODT|
|Page views||40 [see more analytics]|
|Cite as "OSSTIP/WPBZE1". Appropedia. 2021. Retrieved 2021-09-26.|
WPBZE1- Phase 1 Display of new BZE Network in OSSTIP
This is a Work Package as part of the OSSTIP project.
Outputs: Installed instance of OSSTIP with relevant GIS data uploaded, Created map-based analysis, and updated documentation.
Estimated Time: Small-Medium
Requirements Summary[edit | edit source]
In this Work Package, the previous work to set up a display and analysis capability for Melbourne’s current public transport network in OSSTIP/WP4 will be applied to the first version of the new proposed public transport network and preliminary timetable that BZE have developed.
Note: whilst this work package is going on, we expect considerable effort will be needed to refine the proposed network into a full GTFS-compatible timetable, even with the supporting tools developed in OSSTIP/WP5. This will primarily be the responsibility of BZE team members, although user:PatSunter will support this process throughout this work package.
Results[edit | edit source]
Creating the BZE GTFS files[edit | edit source]
|Time Period||Service Headway (min)|
|05,00 AM - 07,30 AM||20|
|07,30 AM - 10,00 AM||5|
|10,00 AM - 4,00 PM||10|
|04,00 PM - 06,30 PM||5|
|06,30 PM - 11,00 PM||10|
|11,00 PM - 02,00 AM||20|
As part of creating the timetable, after some initial tests revealing that the manually created bus stops in QGIS were quite 'sparse' in some areas resulting in long walk times, we also ran a script to create bus stops at a specified maximum distance, if they weren't there already. We chose 500m for this maximum distance. The gallery below shows the difference between the previously manually placed stops in QGIS, versus those that resulted after running the update script.
Checking a bus route that is part of the BZE revised bus network plan in the Google Transit Data Feed library ScheduleViewer tool.
Loading these into OpenTripPlanner along with other modes, and routing tests[edit | edit source]
Then, a new OpenTripPlanner graph was created using this new bus timetable in combination with the existing tram and train timetables, using the same process as in OSSTIP/Setup of OpenTripPlanner to display Melbourne PT network.
We were then successfully able to request journey plans via the web interface on the new network, combined with the existing train and tram networks, and also calculate isochrones of accessibility via a certain time..
Below is an example of comparing a routed journey between the old and new networks.
And here is a comparison of accessibility to the Moonee Valley area in the inner N-W region on a weekday morning, on the old and new networks. This was before we added higher-density stop spacing, so notice that while the green 30-minute travelshed hasn't increased greatly, the yellow 1-hour travelshed has increased significantly, due to the enhanced bus network.
Note that these type of comparisons can be done manually using the interactive web-application aspects of OpenTripPlanner. But perhaps more interestingly, since OTP uses a well-defined REST AP for requesting calculations via URL, it is quite feasible to create a client script in a language such as Python to automatically select and compare travel time, location accessibility etc between network configurations.
Some caveats and lessons learnt[edit | edit source]
- OTP by default when doing a route in the journey planning interface returns the 'total travel time', but this doesn't include the wait time at the start. But one of the benefits of the BZE network is the proposed much-increased frequency and thus reduced starting waits, so in interpreting journey plans you need to carefully read the start and end time of trips, not just the total travel time.
- The walking trips at the start and end of journeys have a big impact on the results. So choosing the density of stops is important, as is making sure there is reasonable coverage over the entire area. One additional issue is that walking calculations do depend quite a bit on the accuracy of the OpenStreetMap street network and tags - including the digitisation of walking access to stations.
documenting the server-creation process[edit | edit source]
Setting up a server to display the new network is something we're working on documenting better for future reference, in OSSTIP/WPBZE3.
Relationship to later WPs[edit | edit source]
Completing this WP now lays the foundation for OSSTIP/WPBZE3, setting up a live web-server running both old and new networks, to communicate the benefits and differences between them.
[edit | edit source]
Re-run process for phase 2 BZE network[edit | edit source]
Depending on if work in OSSTIP/WPBZE2 and other analysis suggests network refinement is needed, we may want to re-run the process of creating an updated GTFS timetable, and displaying the results.
Testing of OTP Analyst time of day issues[edit | edit source]
Detailed checking of OTP analyst results has revealed some potential issues.
See OSSTIP/OpenTripPlanner Issues for tracking of those.