Future urban mobility

We're aiming to develop and evaluate an innovative framework to optimise and coordinate cooperation between multiple parties that are independent organisations but have similar interests in the same business area.

We’re using the parcel delivery services from Australia Post, Toll and other delivery companies as our case study for developing a more efficient and cost-saving method for sharing the movement of goods while also keeping sensitive information such as customer lists private.

We’re aiming to develop comprehensive tests that can help regulators assess the readiness of automated or autonomous vehicles. Currently, safety readiness tests of autonomous vehicles are not available – making it difficult for regulators to determine which self-driving software can be certified for deployment.

Our research will be used to develop a hierarchy of digital tests inspired by human driver licensing examinations and computer vision assessments of artificial intelligence systems. Expected outcomes include a standardised examination method that will provide significant benefits by allowing regulators to use the proposed standardised procedures to mandate safe self-driving systems.

We’re investigating new methods for managing urban congestion and reducing emissions through innovative transport pricing policies that encourage travellers to reduce their reliance on private vehicles and opt for alternative low carbon or low emission modes of transport instead.

It’s predicted that pricing policies will have a substantial influence on travellers’ choices of transport mode and time of travel, and will have a lasting impact on travellers adjusting their behaviour and mobility patterns to meet the needs of a sustainable world. 

To achieve this, we will establish current and emerging trends in urban transport pricing and the technology solutions available to facilitate their implementation. This will be followed by testing and evaluating the impacts of these policies using traffic simulation models that replicate traveller’s behaviour under dynamic pricing policies that vary by location, distance, time of day and other similar policies.

The outcomes of our research will provide government agencies and road operators with the necessary tools that allow them to evaluate the effectiveness and impacts of various pricing interventions and selecting policies that either individually or collectively deliver the best outcomes for reducing congestion, improving amenity and reducing emissions.

We’re aiming to develop a self-operating mechanism for increasing the successful delivery rate in the crowdsourcing-based last mile delivery environment. By introducing parcel lockers into the system, the last mile delivery is not constrained to a single person; multiple people can contribute to a single delivery with reasonable resources. We’re employing machine learning methods in our research to locate the potential geometric locations to place the parcel lockers. 

Research into the impacts of autonomous vehicles has received considerable attention in recent years, particularly their benefits as a cost-effective mode of shared urban mobility and their potential to reduce the total number of vehicles required to meet people’s needs and demands for travel.

There’s more recognition today, however, that due to regulatory, public acceptance and technology challenges, the deployment of fully autonomous vehicles is likely to be decades away and that more immediate benefits can be obtained from lower levels of vehicle autonomy complemented by vehicle connectivity.

To explore this proposition and determine impacts of vehicle automation and connectivity, we’re aiming to develop simulation models that can be used to evaluate the potential safety and capacity improvements that can be realised from converging automated vehicle technologies with vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications.

We’re developing an innovative approach to estimate future traffic conditions based on historical information received from sensors on freeways. The model can be used by road operators to optimise freeway operations and avoid traffic breakdowns.

Our research uses deep learning artificial intelligence techniques and advanced methodologies of traffic control and management to identify traffic patterns and conditions that might lead to a breakdown of traffic conditions at forecast time horizons from 15 to 60 minutes into the future.

We’re aiming to develop and evaluate simulation models for dynamic on-demand ride-sharing systems, specifically a robust and scalable dynamic network model for on-demand bus services in Melbourne.

The success of these systems relies largely on adequately estimating ridership demands, which we’ll address by using artificial intelligence and machine learning to predict future demand based on behavioural surveys. The origin-destination data needed for model development will also include touch-on and touch-off Myki data covering the study area.