Active travel modelling

Teemu Sihvola, Ramboll - Simulated bicycle flows from mobile network data

Martijn Hollestelle | Ramboll

Martijn Hollestelle is a data scientist and urban planner in Ramboll's Digital Mobility Lab. He applies data science methods to understand the impacts of mobility and other technical and societal developments on society the built environment. For many years he has been working on various innovative modelling methods. He has developed various (agent-based) modelling applications in the field of land-use transport interaction, mobility, traffic safety and renewable energy transitions. 

Teemu Sihvola

Senior Consultant (Smart Mobility) | Ramboll Australia

Teemu Sihvola is the Head of Department of Ramboll Digital Mobility Lab. Teemu is a pioneer in developing on-demand transportation platforms. He developed the concept for Kutsuplus, the first dynamic ride-pooling service in the world. Teemu has experience developing digital mobility concepts as a consultant, researcher and as a startup entrepreneur.
Teemu has strong experience in personal transport, public transport, tendering processes, and impact evaluations. Teemu has a good view of society's passenger and freight transport and the related economic, social, and service-quality factors.
In addition to the role of Head of Unit, Teemu has held several responsible roles as CEO (Ajelo Oy, and MOIA Finland Oy), member of the Board (Ajelo Oy, Split Technology, Inc, Split Finland Oy), Head of the product development unit (Split Finland Oy) and member of the extended management team (MOIA GmbH).

Ian McCarthy

Head of Smart Mobility, Australia | Ramboll

Leading Ramboll's Smart Mobility business in Australia, Ian is a recognised leader in the development of transport and mobility strategies, with a focus on helping organisations to build an evidence base to support equitable, multi-modal transport investment.
With 15 years of industry experience across consulting, the public sector and with commercial transport operators, Ian brings a well rounded understanding across operations, planning, investment prioritisation and delivery of transport services and infrastructure.
Ian has an extensive background developing strategic frameworks and standards to support healthier, more equitable planning, allocation and operation of road space in Australasia, including a significant role in the development of the National Service Level Standards for Roads. The framework provides a consistent platform to understand the functional requirements for all roads across Australia, including considerations for place and active transport users.

Simulated bicycle flows from mobile network data

Cities and regions across the world are investing in bicycle infrastructure to increase the modal share of cycling and encourage sustainable mobility. It is important to understand the behaviour and needs of cyclists to design the best interventions and policies. 

Bicycle counting data tends to be limited and unreliable, and the overall cycling situation in the network is often hard to understand.  Solutions are available on the market to help planners understand the needs and usage of the bicycle network, such as transport models tailored towards or including cycling, or data products derived from other sources (i.e., GPS-enabled sports tracking platforms). These models are usually costly and require a considerable amount of deployment time and data resources. On the other hand, data from sports tracking platforms is usually more widely and rapidly available, but it is biased towards certain demographic groups and recreational activities. 

Recognising the need for decision support tools for bicycle infrastructure planning that are accurate and easily deployable, Ramboll developed an approach to simulate the routes and volumes of cyclists everywhere on the network based on aggregated mobile network data combined with detailed bicycle network data from OpenStreetMap. 

In this presentation, we will explain the fundamentals of the developed methodology, describe the benefits over conventional methods, and provide an assessment of the accuracy of the results based on bicycle counts and travel survey data. We elaborate on the input data and the steps we have undertaken to identify bicycle trips and increase the level of detail needed for cycling analysis in urban areas. 

Combining our international experience with our local knowledge, the presentation will include use cases in Europe and will illustrate its potential use in Australian planning projects. It will also identify steps for further utilising AI and improving the accuracy and applicability of the tool.

Ten-Zen Guh, Mott MacDonald - Getting More Value From Pedestrian Modelling

Ten-Zen Guh

Principal Transport Planner | Mott MacDonald

Ten-Zen has over 14 years of industry experience across Australia, Asia, North America, and South America in a diverse range of sectors. In the ANZ region, he has led pedestrian modelling on major projects for clients such as Sydney Metro, Brisbane Airport, and National Gallery Victoria.

Getting More Value From Pedestrian Modelling

As Australia continues to grow and governments continue to invest in cities and infrastructure, there is a growing need for pedestrian modelling as a decision-making tool to help plan and design spaces that are fit for walking and movement of people. Pedestrian modelling can be applied to a range of sectors including rail, aviation, leisure, sports, education, and large events, and across the full range of the asset lifecycle including feasibility study, business case, concept design, detailed design, construction, and operations. It can refer to either first principles-based static analysis or agent-based dynamic microsimulation modelling.

On the surface pedestrian modelling may seem familiar and intuitive, however there are layers of intricacies and nuance that require an understanding of in order to make better informed decisions. As pedestrian modelling is a relatively niche discipline which has only recently begun to become more widespread in use on major projects in Australia, it is not uncommon for project team members, including either or both practitioners (modellers) and/or end users (including transport agencies, architects, contractors, developers, etc), to be unfamiliar with some of these intricacies and nuance. This may lead to diminished pedestrian modelling analysis value at best and wasted time and effort at worst, and possibly a risk of poor design and pedestrian experience outcomes.

In his 14+ years of industry experience, Ten-Zen Guh from Mott MacDonald has worked on projects all around the world in a variety of different sectors. He has come across some common themes and pitfalls and in this presentation, he discusses lessons learned that would improve the modelling process for practitioners and provide more value to the end user, thereby benefiting both parties.

Dr Afshin Jafari & Dr Dhirendra Singh, RMIT University - The Activity-based and agent-based Transport model of Melbourne: a virtual environment to test the impact of built-environment change interventions on active travel

Dr Afshin Jafari

Research Fellow | RMIT University

Afshin is a computational urban researcher and a Research Fellow at RMIT Centre for Urban Research with expertise in large-scale transport modelling, agent-based simulation, and active transportation.

Dr Dhirendra Singh

Principal Research Fellow (Associate Professor) | RMIT University

Dhirendra is a Principal Research Fellow (Associate Professor) in the AI Discipline of the School of Computing Technologies at RMIT University and is a technical lead on the ongoing CSIRO/RMIT Evacuation Modelling project.

The Activity-based and agent-based Transport model of Melbourne: a virtual environment to test the impact of built-environment change interventions on active travel

Active transportation offers a promising solution to several key challenges of today’s unhealthy and unsustainable lifestyles in rapidly urbanising cities, from obesity and chronic diseases to congestion and air pollution. Increasing the share of active modes of travel requires a supportive environment and well-designed built-environment interventions. Given the complexities of designing such interventions, there is a need for a virtual environment to test interventions and to analyse their potential impact on active travel.

Over the last few years, a multi-disciplinary team of researchers at RMIT University, Melbourne, developed a city-wide and open-source transport model for greater Melbourne known as the Activity-based and agent-based Transport model of Melbourne (AToM). AToM is designed and calibrated to simulate a single day of Melbourne’s transportation system with a 10% sample population and is designed to be suitable for modelling four main travel modes of car, public transport, bike and walk.

AToM is aimed to serve as a test-bed for planners, policy-makers and researchers to test the impact of different built-environment change interventions, such as adding a bike lane, on travel mode share and individual travel behaviours and to inform prioritise infrastructure investments.

In this presentation, an overview of the AToM and its components will be provided. Additionally, we will discuss what factors were included in the model to ensure the suitability of AToM for modelling active transport trips and provide an example case study of using the model to test a city-wide cycling intervention for Melbourne.