Transport is central to economic activity and to our lives; it enables us to engage in work, attend school, shop for food and other goods, receive deliveries and participate in everyday activities.
Transport questions require design thinking beyond traditional disciplinary boundaries. Infrastructure is traditionally associated with civil engineering, whereas vehicles are anchored in mechanical engineering and services in the domain of business. These domains come together to address big problems like congestion, pollution, safety and accessibility.
The strength of our research into transport lies in its multidisciplinary approach; that is, the ability to scientifically tackle what are fundamentally socio-technical problems with a large and growing toolbox of methods and perspectives.
Our key research areas are system impacts of autonomous vehicles, transport and land use interactions, transport system performance measures, and traffic operations and control.
Our experts: Professor David Levinson, Ang Ji
As autonomous vehicles (AVs) emerge as the dominant technology in road transport, the system is expected to change supporting improvements in road efficiency and safety. Traditional lane-changing models simulate how drivers make decisions, and sometimes these algorithms do not reflect real-world conditions. This project develops an evolutionary game theory based lane-changing model to describe the competition of multiple vehicles during a lane-change manoeuvre. The model is suitable to describe human-driven and AV traffic, and can be deployed in future AVs to produce safer, more efficient lane changes.
Our experts: Professor David Levinson, Bahman Lahoorpoor
Transit networks are born, grow, mature, and decline over time resulting in the current unique state of network structure and land use. This research aims to characterise railway network evolution over a long period of time and to evaluate its projected impacts on land use development by simulating the co-evolution of railways (trains and trams) and urban development (terraces and towers) in urban Sydney.
Differing from previous studies, the integrated development of transit networks and urban space is investigated using historical data. This simulation allows us to analyse the emergent functional hierarchy of the railway network and agglomeration patterns of land use and to explore network-land use interactions such as reinforcement or contraction.
Travel demand and urban form are highly correlated. However, the strong correlation and the high significance of network and built-environment variables in travel demand models do not explain their causal relationship. This research is aimed at understanding the nature of the causal relationship between network investments and travel demand using time-series analyses and Granger causality tests. While such analyses have been made in the context of land-use and transport networks, this research extends the approach to travel demand, particularly mode choice, and thereby offers practical results for transport planners and decision makers.
Accessibility is the ease of reaching destinations and is becoming an important performance measure for transportation systems. Accessibility measures for emerging modes of transport, including shared autonomous vehicles, need to be developed and operationalised. This project aims to define accessibility novel modes, starting with taxi-type services. Appropriate formulation of the metrics has important implications for travel behaviour, operations, and regulation of future modes of transport.
Our partner: Associate Professor Taha Rashidi (University of New South Wales)
Travel time reliability has emerged as a valued aspect of transport system performance. Most previous work has focused on typical variability – the day-to-day changes that travellers experience on a given trip. Increased data availability makes it possible to measure and value the importance of rare events (extreme delay caused by crashes, natural disasters or special events) in traveller decisions about mode, route or departure time. This project used a stated-preference experiment to explore the trade-off between the magnitude and frequency of long delays resulting in better estimates of the value of reliability.
Our experts: Dr Emily Moylan
Sydney Coordinated Adaptive Traffic System (SCATS) provides network-wide signal timings to minimise delay to road traffic, which can lead to negative outcomes for pedestrians and cyclists. Simulations (Mayhook 2017) suggest that including pedestrian delay leads to a different set of signal timings that is more inclusive. This project seeks to quantify the inequity and inefficiency of the status quo with observations of the number of delayed pedestrians, the amount of delay they incur and queue spillback. The outcome is empirical evidence to support a more equitable algorithm for green time.
Traffic operations are essential for managing congestion and supporting economic productivity. Building on control theory, traffic flow theory and empirical approaches, we contribute to theoretical and practical traffic operations.
Our experts: Dr Mohsen Ramezani, Ye Li
Traffic signal control policies are nominally designed for isolated intersection control or coordinated control in arterials. To improve traffic operation efficiency in large-scale urban networks, control policies must think more holistically about network behaviour. A recent approach to extend the spatial extent of traffic signal control to the network level is perimeter flow control based on the Macroscopic Fundamental Diagram (MFD) model.
This project advances that work in three directions: Perimeter Flow Control with Time-varying Cordon based on MFD, Day-to-day Dynamics in Perimeter Flow Control Scheme based on MFD, and Subregional Time-dependent Congestion Pricing Strategy of Large-scale Networks. The project outcome is flexible and effective control strategies to cope with large, uncertain, irregular traffic networks.
Our experts: Dr Mohsen Ramezani, Amir Valadkhani
Ride-sourcing systems such as radio taxis and Uber are a popular mode of transportation because of their convenience. Their role in urban transport is growing due to recent improvements in affordability caused by emerging on-demand services that use mobile apps as a platform to match passengers with vacant drivers.
This project will design a dynamic ride-sourcing system including spatiotemporal pricing methods, driver-passenger matching algorithms, and recommender system for waiting passengers and vacant drivers to increase the system's profit while ensuring a satisfactory level-of-service.