Dr Nicole Ronald wrote an article for the Melbourne School of Engineering news website about the discussions around shared-use mobility from the Annual Meeting of the Transportation Research Board 2015.
Ronald, N., Thompson, R.G., and Winter, S. A comparison of constrained and ad-hoc demand-responsive transportation systems, Proceedings of the 94th Annual Meeting of the Transportation Research Board, Washington, DC, January 2015.
Planning public transport services for areas of low population density is important to enable those without convenient travel options to travel. In these areas, transit vehicles frequently travel with low numbers or even no passengers on board, therefore incurring more cost to the transport providers. Demand-responsive transportation (DRT) services are a potential efficient mobility solution to this problem.
The choice of DRT scheme is important as different types of schemes might produce different performances in the same area with the same demand. While many DRT schemes have some constraints, for example, a fixed route or a fixed time, these impose constraints on users who are already constrained, for example, due to lack of access to a car or limited times to undertake activities. Removing the fixed constraint on time leads to evaluating the performance of an ad-hoc system.
The aim of this paper is to investigate the change in performance between two different DRT schemes — a fixed-time but flexible route scheme and a completely ad-hoc scheme — using MATSim, a large-scale agent-based transport simulation, and real data from an existing fixed-time DRT service in rural Victoria, Australia. Experimentation showed that the schemes produced different outcomes for the operator and passengers, however the optimization algorithm is less important in areas of low demand. Higher levels of demand lead to extensive vehicle travel for an ad-hoc service, while altering the headways between fixed-time services could achieve a middle ground for operators and passengers.
This work is the first step towards developing a decision-support tool to evaluate different DRT schemes, in particular integrated with other modes of transport.
Please email the authors for a copy.
Nicole Ronald will be representing the iMoD team at the Transportation Research Board Annual Meeting, to be held next week in Washington, D.C. She will be presenting a paper on simulating demand-responsive transport using an existing fixed-time system in Yarrawonga/Mulwala, Australia and comparing it with a flexible-time system.
If you are interested in talking to Nicole about iMoD, contact her beforehand to set up a meeting time or drop by her lectern session at 8am Wednesday morning.
Lee, A., M. Savelsbergh (2014), An extended demand responsive connector. EURO Journal on Transportation and Logistics, http://dx.doi.org/10.1007/s13676-014-0060-6.
The need for viable public transit systems has been well documented and so has the role that so-called flexible transport systems can play. Flexible transport services offer great potential for increases in mobility and convenience and decreases in travel times and operating costs. One such service is the demand responsive connector, which transports commuters from residential addresses to transit hubs via a shuttle service, from where they continue their journey via a traditional timetabled service. To access this service, commuter and service provider agree on an earliest time the commuter must be available for collection and a latest time the commuter will arrive at a transit station. We investigate various options for implementing a demand responsive connector and the associated vehicle-scheduling problems. Previous work has only considered regional systems, where vehicles drop passengers off at a predetermined station; one of our contributions is to relax that restriction and investigate the benefits of allowing alternative transit stations. An extensive computational study shows that the more flexible system offers cost advantages over regional systems, especially when transit services are frequent, or transit hubs are close together, with little impact on passenger convenience.
On 28 and 29 August 2014, Nicole Ronald participated in a NICTA-hosted infrastructure hackathon organised as part of the Australia 3.0 initiative. From our point of view, the main aims of participating were to understand more about how hackathons work and how we can encourage spatial and transport students to get involved, as well as having the opportunity to present some of our research to a different audience.
— Code for Australia (@CodeforAus) August 29, 2014
Nicole took the opportunity to work with the newly released Uber API, which permits real-time access to Uber data, and to test out a visualisation of launchpads, based on Michael Rigby’s PhD research. While Michael’s research focuses on ridesharing, where privacy is a major issue, Nicole identified convenience (shorter travel times, cheaper trips) and health (a door-to-door travel culture leads to less walking) as potential reasons why spatial flexibility is useful in the context of single-passenger taxis. As a one-person team, this provided a self-contained project that produced some early results: the diagram below showed that, when starting at the train station, walking a short distance could lead to being picked up quicker and a quicker ride. We intend to turn the static mockups into a live demo in the near future.
The winning team consisted of PhD students from our research partners in Computing and Information Systems, The University of Melbourne and a Monash student co-supervised by Mark Wallace, an iMoD investigator. They will be mentored to further develop their product and also received two return flights to Silicon Valley.
For more information about the hackathon, please visit infrahack.org.
On Friday 13 June 2014, Dr Nicole Ronald gave a presentation on simulating DRT to the Sim-VIC research community at RMIT. This community consists of researchers working in simulation, mostly coming from agent-based simulation.
Nicole spoke about three models of DRT that have been built as part of the iMoD project: the original Delphi prototype, the SUMO-based model SUMOoD, and a model currently in development based on MATSim. The aim of this work was to explore the capabilities and limitations of existing software and identify the potential of various tools that can be used for modelling DRT.
All three models have been set up with a basic DRT scheme which optimises only single trip, “travel now” requests by minimising passenger travel time. No other transport modes were included, to ensure comparison between models; the Delphi prototype models DRT only.
Traffic microsimulation software is useful for evaluating door-to-door DRT in small regions, however more work is needed in predicting travel speeds for routing. MATSim has proved useful for larger areas, however this model currently does not make use of iterations (that is, no scoring and replanning occurs). Models of DRT in isolation are useful to get a quick overview of coverage and operation, however provide no indication of interaction with other modes of transport, and depending on the richness of the transport network representation and routing algorithms, could provide overly optimistic results.
Future work includes generalising the models to handle different optimisation algorithms and DRT schemes and beginning to explore different locations, as well as continuing our work on estimating demand. The iMoD project is grateful to Sim-VIC for the opportunity to present work-in-progress.
On 4 June 2014, Chenhao Fan, a final-year student from Southeast University, Nanjing, China, presented his graduation project on estimating demand for DRT services.
Chenhao has spent four months in Melbourne working with Assoc Prof Russell Thompson and Dr Nicole Ronald, as well as data and policy experts from the Department of Transport, Planning and Local Infrastructure, State Government of Victoria. He analysed VISTA and ABS census data to obtain preliminary estimates of trips per household.
- a new research fellow and several PhDs starting work on topics related to the project
- publications at ACM SIGSPATIAL and other papers submitted to journals
- making connections with other researchers both locally and internationally
Our plans for the next year include developing innovative optimisation algorithms for and simulations of various DRT schemes, making use of real-world data provided by industry partners.
Urban mobility and accessibility is a problem for growing cities. New ideas are required to increase mobility and access in a sustainable way, taking congestion, fuel consumption, and environmental impacts into consideration.
One possible solution is sharing transport resources, along the lines of bike sharing, car sharing, or ride-sharing, and enhancing the benefits of shared resources by making them demand-responsive. Although these systems are gaining traction internationally, many fail due to poor implementation, planning and marketing. Being able to realistically simulate these systems to evaluate viability and demand before implementation is important.
A team of researchers at the University of Melbourne, Monash University and University of Newcastle is investigating the viability of novel mobility-on-demand systems. This involves estimating the demand for travel, modelling the behaviour of potential users, developing scheduling and matching algorithms, and building simulations to evaluate systems in various urban environments and scenarios.
This presentation will report on early research outcomes and work-in-progress, focusing on a thorough review of the demand-responsive transportation literature and potential simulation approaches.
Ronald, N., Thompson, R., Haasz, J., Winter, S. (2013), Determining the Viability of a Demand-Responsive Transport System under Varying Demand Scenarios, 6th ACM SIGSPATIAL International Workshop on Computational Transportation Science. ACM, Orlando, Florida, USA.
Collaborative transportation has been proposed as a potential solution to decrease congestion, reduce environmental effects of transport, and provide transportation options to those with no or restricted travel options. One such system is demand-responsive transport, in which passengers share a vehicle, usually a bus, but can be picked up or dropped off at a passenger-specified location and time. However, as these systems are expensive to implement and require long trials in order to gain traction, effective simulation is required in order to explore their viability before implementation. Although previous work has concentrated on the number of trip requests, the spatial distribution of these requests has not been considered. This paper explores four spatially-varying demand patterns — random, a many-to-one scenario, all short distance trips, all long distance trips — using a simulation of an ad-hoc demand-responsive bus system. It is shown that along with the number of trip requests and the requested trip distances, the spatial distribution of passengers does indeed have an effect on the level of service.
Please contact one of the authors for a copy.