AITPM New Members – July 2017

AITPM was pleased to welcome the following members in July 2017:

First Name Last Name State Membership Level
Northern Beaches Council NSW Corporate (5)
Oliver Chambers QLD Student
Paolo Moscheni WA Student
Hong Ki An SA Student
Christopher Hew VIC Student
Matthew Creighton QLD Student
Monita Wambrauw SA Student
Mitchell Young VIC Student
Navid Muttaqui QLD Student
Sigrid Pembroke QLD Associate
Michael Willson NSW Member
Rebeka Giana TAS Member
Shivani Jordan TAS Member
Steven Northey NSW Member
Rebecka Gunnarsson VIC Member
Kate Pratt VIC Member
Andrew Morse  NSW Member
Gi Sun Hong QLD Member
Giles Smith QLD Member
Ronja Trapp QLD Member
Vishnu Anne NSW Member
Rodney Markotis WA Member

Transport Surveys: Don’t just look at the width

In our December 2016 newsletter, we had an article that began with the words:

It is said that big data is a kilometre wide but only a millimeter thick. It is providing us with a lot of numbers but the information can lack depth. There are many factors involved in why people choose to travel, the method they use to do it and the path they take. We need a comprehensive range of data to ensure we understand fully the environment in which we work.

In a recent ITE Talks podcast, Lucy Saunders a public health specialist, at the Greater London Authority in England and Dale Bracewell, manager of transportation planning for the City of Vancouver in Canada spoke on the topic that Care must be taken when using traffic data as evidence.

Lucy gave a great example of how just counting some numbers does not give the full story.

The AITPM national conference in August 2017 will have a session on transport data that looks at the latest methods of data collection but also how we have to look very hard at the real meaning of the numbers we collect.

A link to the conference web page and to the ITE podcast is on the AITPM Web site.

https://www.aitpm.com.au/national-conference/

http://www.traffictechnologytoday.com/video-audio.php?v=saunders

Developing a Good Practice Guide on Travel Forecasting

Primary Author : Ian Clark

Organisation : Flow Transportation Specialists

 

This paper will document work carried out by Ian for NZMUGS, leading to a “good practice guide” being prepared, which was presented to the NZMUGS conference in September 2016. The work was in response to concerns expressed, both at AITPM and NZMUGS conferences, about inadequacies in the way the transport modelling industry was dealing with uncertainties in forecasting.

 

The work included a literature research on the accuracy of forecasts and the main sources of inaccuracy. It included surveys among NZ practitioners on current practices, leading to commentary on the topics of: toll roads, conservatism, transparency, optimism bias, length of forecast, the effects of new technology, looking back as well as forward, trip generation, consistency, and risk.

 

Aimsun Mesoscopic Calibration

Primary Author : John Bennett

Co-authors : Brian Betts

Organisation : AECOM

 

Mesoscopic level modelling for the purposes of project assessment has increased in prevalence across Australia and is considered to provide a bridge between macroscopic modelling and microsimulation modelling.

Aimsun provides a software tool to bridge this gap by facilitating integration between macro, meso and micro levels of detail. The mesoscopic assignment is based upon a simplified application of the car following model, which is more commonly used for microsimulation.

Despite the rise in use of Aimsun as a tool for project assessment, very little research exists to guide practitioners on best practice for calibration at the mesoscopic level. This paper investigates the key parameters that influence the calibration of Aimsun mesoscopic models and recommends a process that achieves comparable traffic operations between the mesoscopic and microsimulation level models.

 

The simplified car following model used in Aimsun mesoscopic assignment does not account for key factors in traffic operation, including vehicle acceleration/ deceleration and detailed lane changing/ merging/ weaving behaviour. This lack of detail often leads to difficulties in achieving representative levels of congestion on approach to intersections – a common finding is that the mesoscopic assignment underestimates congestion in comparison to a microsimulation assignment within the same network.

The analysis presented in this paper utilises an existing Aimsun model to identify a best practice process to calibration of traffic conditions at the mesoscopic level. The process focuses on how vehicle reaction times and section (link) speeds can be used to sufficiently calibrate operation within a mesoscopic assignment and to provide comparable operation with a microsimulation assignment within the same network. Recommendations on a range of values to adopt for vehicle reaction times and vehicle speeds are provided for a range of situations, to be used as guidance for industry practice.

 

External Factors That are Likley to Affect Travel Patterns

Primary Author :Steven Piotrowski

Co-authors : Hugo Wildermuth

Organisation: Consultants, Department of Transport, WA

 

As part of the Perth Transport Plan for a City of 3.5m people, long range transport modeling was undertaken using strategic travel models.The strength of conventional transport models is in short-to-medium term projections, in realistically distributing demand among the thousands of origin-to-destination pairs, and in differentiating between various activity centres and between peak and off-peak travel. The weakness of the current models is that they do not explicitly allow for changes in people’s travel behaviour in response to “external” factors not directly related to the transport system. Two travel behaviour scenarios were considered: a “Base Scenario” employed traditional modeling and a ‘business as usual’ situation. The “External Influence” scenario was expected to include changes to travel demand as a result of expected changes in the external operating environment for transport such as the introduction of autonomous vehicles.

 

A workshop was held with 20+ Perth-based transport experts who were asked to rate qualitatively the expected impacts on travel demand of changes to several “external influences” including: socio-demographic, employment, education, shopping, ITS, and other technology. The anticipated impacts on land use, road capacity, daily trip rates, car usage, public transport usage, active transport, peak period travel, commercial vehicle trips and value of travel time savings were estimated and tabulated.

 

Modelling pedestrian and vehicle interaction in Auckland

Primary Author : Christian Arkell

Organisation : Jacobs

 

This topic outlines the lessons learnt and the findings from using Legionfor Aimsun to model the interaction between pedestrians, cars and light rail vehicles. This type of modelling is likely to play a large role in transport infrastructure planning going forward, particularly in city centre areas with significant pedestrian volumes.

 

Access into Auckland’s City Centre is becoming more and more constrained and requires a significant public transport intervention in  order to cater for rapid population growth currently occurring. Bus corridors are already over-capacity so Auckland Transport has begun to develop plans for a Light Rail network. Jacobs is involved in a joint venture which is currently engaged as technical advisor and concept designer. Several of the light rail stations will be located in areas of high pedestrian activity which raised questions about:

  • The impact of turning cars giving way to large numbers pedestrians
  • The required capacity of light rail platforms and adjacent footpaths

In order to answer these questions, Legion for Aimsun has been used to simulate the interaction between cars, light rail vehicles and pedestrians at various points along the proposed route.

Issues during modelling included:

  • Safety and operational issues when integrating light rail pre-emption signals with pedestrian crossings
  • The simulation of barnes dance/scrambled pedestrian crossings
  • What is the impact of increasingly high light rail vehicle frequencies.

The main outcomes and lessons learned:

  • Legion for Aimsun is a useful tool for modelling interactions between all road users
  • The impact of increasing frequent light rail service on pedestrian amenity is significant and important

 

Freight modelling in Australia

Primary Author : Pedro Camargo

Co-authors : Lauren Walker

Organisation : Veitch Lister Consulting

According to a 2008 study by Ibis World, non-bulk freight in Australia is expected to grow by 350% between 2008 and 2050, while bulk freight is expected to grow by over 280% in the same period. Despite this substantial projected growth, the freight modelling capabilities of many road agencies have not kept pace with planning needs (BITRE 2014).

This paper proposes path for state agencies and the federal government to expand their freight modelling capabilities, addressing the pressing need better tools to plan for future freight movements.

 

There is consensus that understanding the impact of the freight movement growth will allow for the development of policies to address issues such as decaying infrastructure, greenhouse gas emissions and social equity. The availability of a consistent forecasting model will provide a tool that allows for the appropriate analysis of these policies.

This paper explores the current state of freight modelling in Australia and the data available for the development of new models. It is divided in four

parts:

A thorough investigation of the datasets available for freight modelling in Australia and identification of major gaps in available data
A review of the freight models being developed abroad
A proposed model structure suitable for statewide and countrywide implementations, including potential ways of addressing data gaps without costly surveys
A presentation of freight submodel components that include matrix disaggregation, trip generation and trip distribution, as well as advanced visualization on existing datasets and model outputs.

 

Perth Freeway Modelling

Primary Author : Antony Johnstone

Co-authors : Rafael Carvajal Cifuentes

Organisation : Aurecon and Main Roads WA

 

In May 2015, Aurecon was engaged by the Main Roads Western Australia (MRWA) Network Operations Directorate to develop the largest transport simulation model ever built in Western Australia. The initial model was 31km in length, 170 km2 in area, and covered of Perth’s Northern Freeway and surrounding suburbs. The model took seven months to develop and required extensive adjustment. A second stage was then commenced which further extended the model area to also cover the majority of the Perth’s Southern Freeway, resulting in a total modelled Freeway length of 82km.

The conference paper will be presented by both MRWA and Aurecon and will cover the purpose of the development of such a model, the complex project methodology, correlation outcomes, and the lessons learnt.

 

Based on a much larger spatial area than such models typically cover, the Freeway Simulation Models provided MRWA’s Network Operations Directorate with a holistic view of the Freeway performance and its surrounding areas. It enabled MRWA to gain crucial insights into how the network’s infrastructure will perform in future years, and to understand the impact of proposed upgrades or developments. On an ongoing basis, MRWA will be able to design to manage constrained flows, and ensure that effects on the wider network are understood and mitigated as much as possible.

At the time the project pushed Aimsun’s pre-existing capabilities. Multiple software versions were re-released during the project. TSS has selected the Freeway Simulation Model as one of its flagship projects to present at its user group meetings around the world.

The Aurecon and TSS teams were able to leverage global time zones to achieve 20-hour working days to meet the project timeframes. Austraffic were also engaged and undertook one of their largest data collection exercises in WA, with MRWA and Aurecon undertaking the site visits by helicopter to cover the large area.

 

Overview of the New Greater Brisbane Travel Demand Model

Primary Author : Benjamin Pool

Co-authors : Jaco Van Den Berg

Organisation : Queensland Dept of Transport and Main Roads

BNE Model development is complete and model launched. This is a quick review of model features and performance that improve on previous practice

 

In 2011 a review and Audit of the existing BSTM-MM found a number of issues. The BNE Model Development Program addresses these issues. Model performance to date has been significantly improved A Soft Launch us underway through to end March when a hard launch is expected. A Review comparing the BNE to the BSTM-M will have finding by EFY.

 

Revolutionising travel time analysis, Is Google the future?

Primary Author: Premraj Dorai Rajoo

Organisation : Jacobs

 

This paper will analyse the viability of Google Maps travel time data extracted from Google’s vast collection of crowd sourced data. This paper will also cover the suitability of this data to replace or supplement the use of other travel time data such as bluetooth and floating car surveys in transport planning and modelling projects.

 

Collection of quality data is one of the major risks to traffic modelling projects, both with respect to time and cost. We are now seeing many examples where remote sensing technology is collecting “big datasets” that can provide critical data in a more cost effective and often more accurate manner.

This study investigates the viability of Google Maps travel time data by using a data mining tool to extract the live travel time estimates from the Google Maps API on selected road corridors in Brisbane while also conducting floating car travel time surveys on the same day. These datasets were compared to establish the accuracy of the Google Maps data that was extracted. The corridors selected for this study were chosen based on their differing characteristics to give a robust analysis of travel time on varying traffic conditions.

In addition to the Google Maps validation exercise, with access to a large database of travel time data, this paper will report on the quantification of travel time and travel time reliability across the network using the Google data.

This paper will also include discussion on software implications, legal implications, licensing and budgetary considerations.