Road safety and road transport sectors acknowledge the need to address heavy vehicle safety, both in terms of reducing the number of crashes and the flow-on impacts on productivity. Advances in technology now enable transport operators to strengthen their ability to measure and monitor in-cab driver performance in real-time as a way of complementing existing company safety policies and further ensuring they meet OHS requirements. This paper outlines a Commonwealth-funded program, the Advanced Safe Truck Concept project, led by Seeing Machines in collaboration with the Monash University Accident Research Centre, Ron Finemore Transport and Volvo Group Australia. This three-year program aimed to better understand the real-world risks faced by trucking operations and their drivers, and then through this high quality research to generate new technological solutions.
Priority on Australian roads is the safe and efficient movement of people and goods with an emphasis on the movement of motor vehicles. Comparatively there has been little consideration of safety when we are vulnerable road users, particularly cyclists. Fundamental to safety when we ride on the roads, is safe interactions between cyclists and drivers. This study involved an investigation into a specific interaction is not well understood – when a cyclist is travelling straight and a driver is turning left. More specifically, this study aimed to provide new insights to understand how engineering and legal aspects influence safety issues related to this interaction. To achieve this, definitions of safe and legal were explored and then applied video observations from interactions at intersections in metropolitan Melbourne, Victoria (n=275). Two thirds of interactions (188, 68.4%) had elements of non-compliance and potential contributing influences included infrastructure/road design (i.e. road markings, bikes lanes, pedestrians and number of vehicle lanes) as well as observed behaviours (cyclists and drivers). Findings highlight that while there is a connection between engineering (road design) and law (road rules), the two currently operate in parallel and are likely to be contributing to inconsistent behaviours and confusion among road users.
Many heavy rail systems rely on Park and Ride to (PNR) for car-based station access. Providing parking adjacent to stations is expensive, can create local congestion and discourage alternative access modes. Travel surveys in Melbourne, Australia are used to examine current station access mode choice decisions and the potential to reduce reliance on PNR. Station access mode choice decisions vary depending on socio demographic factors, access distance/travel time and line haul travel time/distance. Current PNR user decisions are strongly influenced by parking prices at the ultimate destination and the perception that the car is a faster station access mode than other options. The bicycle is perceived to be cheaper and easier to park access option, but concerns emerge about safety and its travel time, reliability. Nearly half of respondents were interested in accessing the station by bicycle highlighting the scope for tailored programs to influence current station access decisions.
The replacement of conventionally fueled passenger cars by electric vehicles (EVs) has long been expected across the world as developed nations move away from fossil fuels and towards energy use that is cleaner, cheaper and renewable. However, where some nations have shown a clear path towards rising EV uptake during the past decade, Australia has steadfastly refused to follow: EV uptake has been extremely limited, while established patterns of liquid fuel consumption have continued unabated. The potential virtues of EV adoption have been extolled from all the usual perspectives: environmental benefits have been measured, low economic running costs have been highlighted, and even improved levels of energy security have been indicated. But these arguments to encourage EV uptake have failed to gain any traction sufficient to replace the status quo of a national road transport system that has grown to rely on a diet of imported conventional liquid fuels. A recent survey of industry, research and government operatives by Murdoch University researchers probed public awareness of the complex issues surrounding transport fuel economics and the reasons for consumer reluctance to change to EVs. The main aim of this paper is to use information from the survey as evidence to support the theory that the path to improving EV uptake in Australia will continue to fail unless EVs become available under a certain set of conditions, or possibly until it happens by default when The Fourth Wave arrives.
The development of infrastructure delivery is a well-trodden path from concept to business case to detailed design and opening. Processes exist at both state and national levels to outline justification for value of an investment. Many of these concept solutions are larger scale and require development from government bodies, while design elements sit with the local delivery teams. These two arms produce a dichotomy of challenges – the delivery for the city and state against the requirements for a well-defined project scope. The end result can produce for a project with focus on the new infrastructure but understated holistic network view to plan and deliver complementing projects for access and unloading to the new investment. VicRoads as part of Department of Transport required a model to explore wide range of questions of short term to midterm delivery needs of this includes how and to what extent the proposed projects influence traffic movements and journey patterns in terms of various tier of impact. The organization also needs traffic management solutions that requires to be implemented that can strategize to mitigate the impact of disruption as an indispensable part of construction of new projects. This ensures that projects can be delivered. The Domino model was developed as an innovative new approach to estimate and predict the response of network to various projects ranging from change in signal plans to construction of new infrastructures such as highways or bridges. This instrument assists VicRoads in immediate delivery of network projects.
Vision Zero and the Safe Systems approach are part of the recent shift in road safety engineering towards networks that are more forgiving of human errors. These new approaches challenge earlier attitudes where ‘driver error’ was considered a major, but unavoidable cause of road trauma. However, it is unclear whether the researchers and practitioners developing these new approaches have successfully engaged with the legal profession to bring traffic law along, together with the field of road safety, though this transition. Traffic enforcement and the legal processes to deter and punish violations are an important input to the “alert and compliant road users” (PIARC 2015) at the centre of the Safe System approach. However, road rules exist in an adversarial justice system built on concepts of negligence and duty of care. Whether road safety thinking aiming to be “more forgiving of human error” (PIARC 2015) is compatible with, or is being adopted by, the current legal system remains unclear. This paper explores a recent, high profile crash at the Montague Street bridge in South Melbourne. It uses case study methodology and a textual analysis of the judge’s sentencing remarks to explore how the laws of negligence might overlap or conflict with research knowledge about human factors and the driving task, and the Safe Systems approach. The paper does not seek to judge, comment on or otherwise give detailed opinion on the legal system or the outcomes of the case in question. Rather, it finds that current road safety research may not be fully informing or have been fully incorporated into traffic law or the way the legal system generates outcomes. Conclusions and directions for further research are also detailed, including a suggestion for greater engagement by the field of road safety engineering with that of law. (words 295) 1 Note to readers: please forgive the somewhat indulgent reference to Romeo and Juliet (Shakespeare 1594) in the title. However, in the context of the Montague(s) Street bridge, and themes related to fate, chance and human nature, it was just too hard to resist this allegory. Plus, a (hopefully) catchy and engaging title might just help to make you want to keep reading (Delbosc 2017). ATRF 2019 Proceedings
The accuracy of dwell time estimation is crucial for both tactical and operational practices of public transport. This paper aims to make a systematic review of the dwell time models that have evolved over the past 40 years. The scope of this study is limited to the dwell time models pertaining to the passenger rail. Studying the literature concerning train dwell time models, similarities and differences were analysed and discussed based on the modelling approaches and assumptions made in the development of the models. For instance, models were categorised based on modelling philosophy, time-period, key variables involved, data collection and validation methods. Through the comparison, common interests and future trends were then identified. It is found that there is no perfect model that fits all scenarios. The best outcome relies on the effort of choosing the most appropriate model, calibrating the parameters, making some ad-hoc adjustment and continual improvements.
Reliability is one of the critical success factors for both passenger and freight rail service delivery. One major factor that significantly impacts reliability performance is delays spanning over spatial and temporal dimensions. One way to increase reliability is to avoid systematic delay propagation through better timetable design to reduce the interdependencies between trains caused by route conflicts and train connections. In this paper, we aim to predict the propagation of delays on a railway network by developing a conditional Bayesian delay propagation model. In the model, the propagation satisfies the Markov property that determination of delay propagation for the future of the process is based solely on its present state, and that the history does not have an influence on the future. For the cases of delay caused by cross line conflicts and train connection, throughput estimation is considered in the model. The proposed model benefits from scalable computing time and complexity advantages over the Markov property. Implementation of actual operational data shows the feasibility and accuracy of the proposed model when compared to traditional probability models. The proposed model can be used for timetable evaluation and operations management decision support.
A road transportation network and its strategic utilization has a crucial role in emergencies occurring after natural disasters. After most natural disasters, such as floods, hurricanes, tornadoes, earthquakes and tsunamis, one of the most important emergency responses is to provide or deliver relief goods, such as water, food or medicinal supplies, to the affected areas. The complication is that in determining the routes to take for deliveries to affected areas, one has to take into consideration, at the very least, the costs, duration of trips and the availability of the routes. Also the supply and demand situation of the relief goods has to be taken into consideration before choosing the most preferred routes for the deliveries. In this paper, a Monte Carlo approach is applied for the emergency relief goods transportation strategy problem. Monte Carlo simulation has been used for varied applications in including project cost estimation, project schedule estimations, risk assessments, benefit cost analysis and selecting risk response strategies. The Monte Carlo model developed in this paper integrates costs, duration of routes and availability together with the supply and demands requirements to generate the most preferred routes. The results of the Monte Carlo simulations can be used by decision makers (emergency response team) to facilitate the decision making process while choosing the preferred and practical combinations of routes for various deliveries. The proposed approach is then applied to several simple situations to illustrate the simplicity, versatility and practicality of the approach.
Bluetooth MAC Scanner (BMS) based traffic data is widely utilised to estimate travel time (speed) on the road network. The seamless availability of BMS data from large urban networks (such as Brisbane) provides opportunities to visualize congestion on the network. However, the baseline road network cannot be directly used for congestion mapping as the BMS scanners are offset from the road network. Thus, it becomes necessary to snap scanner points on the road network thereby creating a BMS based network. The BMS based network lines are currently manually assigned which are inefficient as well as time-consuming. This paper provides a technique that can be adopted for any large-scale network to define the links between the scanner locations. The strategy expresses a procedure based on the restricted path matching technique. As a case study, the proposed methodology is applied on real Brisbane network and utilised for congestion dashboard development.
Agglomeration benefits are usually the largest category of wider economic benefits of transport projects. They are estimated by applying productivity elasticities to forecast changes in effective densities. These productivity elasticities are obtained by regression analysis to fit production functions that include effective density as an accessibility measure. In most cases, the functional form and parameter values for the distance decay curve in the effective density specification are simply assumed. This paper discusses how different decay curve assumptions affect productivity elasticity and agglomeration benefit estimates. It is shown that an agglomeration benefit is comprised of a large number of terms, each affected in multiple ways by the decay curve. Numerical simulations for hypothetical cities and projects are employed to further investigate the effects. Generally, a faster rate of distance decay leads to lower productivity elasticity and lower agglomeration benefit estimates. It is recommended that the decay curve functional form and parameters be estimated from productivity data when estimating productivity elasticities rather than imposed by assumption.
Australia has seen a steady rise in the number of car-passenger trips made by children to school, and a decline in walking-to-school. Australia differs from most nations in that it has one of the highest rates of private schooling in the world at around 34%, supported by high levels of funding support from the Commonwealth Government. Little is known about the effects this has on travel behaviour and whether it is a factor in our high rates of chauffeuring. This paper looks at journeys-to-school in South-East Queensland. Two research questions were posed: i) how do students in private and public schools travel to school? and, ii) is there any relationship between school type and mode choice? The methods involved advanced geo-spatial matching to allocate all trips made to schools in the 2017-2018 South East Queensland Travel Survey to either public or private schools. The resulting dataset included 617 trips from home to private schools and 2,539 to public schools. Private school students are less likely to walk to school and more likely to be chauffeured, than public school students. For those chauffeured, trip distances are much greater for private secondary school students (median = 7.7km) compared to public secondary school students (4.4km). We estimate that private secondary schooling alone was responsible for around 56 million km of additional private motor vehicle travel on the SEQ road network in 2016. Australia’s policy settings supporting high rates of private schooling appear to be a modest but important contributor to traffic congestion and declines in child physical activity and independent mobility. These impacts should be considered in any holistic evaluation of the costs and benefits of Australia’s school funding model.
The forthcoming integration of autonomy into social and economic systems is a priority area of research. While non-trial research exercises such as public confidence surveys, traffic simulations or policy reviews remain important, they fail to include the ‘on the ground’ learning. This paper presents user experience data collected from passengers whilst they were travelling on-board an AV shuttle travelling at low speeds on the UWA campus. The user responses indicated that approximately half the users thought the shuttle to novel but not a practical form of transport, leaving half that did indicate a willingness to adopt the service as part of their day to day travel.
Understanding the differences in multi-modal travel demand can help transport planners to improve the sustainability of a transport system. Thus, this study aims to develop a multi-step methodological framework to identify gaps in demand between different modes and apply on a realistic large-scale network. The framework includes three methods. Method 1 is carried at a coarser level of spatial resolution, while method 2 and 3 are carried at one level finer resolution than that of method 1. The proposed framework is demonstrated using car and transit OD matrices developed from observed Bluetooth and smart card data, respectively for the Brisbane City Council region. The gaps in transit service usage are estimated between different sections of the network by identifying OD pairs that have low transit usage but high car demand. The findings from this study show that there are significant number of OD pairs that might require further investigation in order to improve overall transit patronage for Brisbane city. For instance, Method-1 showed that SA4 (coarser level) OD pair of Brisbane North- Brisbane East needed the most attention for transit improvement, and method-2 further identifies the SA2 (finer level) zones within Brisbane North- Brisbane East (for example, Eagle Farm – Pinkenba) that needed to be further investigated. Although the techniques are only applied to car and transit matrices, the proposed methods are generic in nature, and therefore can be applied to compare other modal combinations.
Public Transit (PT) Supply Index (SI) is primarily estimated to prove the differences in demand and supply to provide an understanding of social issues related to PT, such as equity, and social inclusion. The definition of SI is yet not clear from the literature; thus, it is usually mixed with accessibility, and availability of transit service. This study revisits the supply index estimation problem with the more detailed and robust approach. A route-based equation is developed in which the variables include walk buffer areas of stops, the total area of the zone, straightness of route provided to that zone, and average Available Seating Capacity (ASC) of transit services. The equation provides Origin-Destination (OD) based SI, that is SI for each OD pair. Furthermore, the equation also incorporates the number of transfers needed to travel from the origin zone to the destination zone and speed between those zones. Most of the data required for analysis are taken from either static Google Transit Feed Specifications (GTFS) or smartcard data. Though the analysis is only carried out at a few OD pairs, when the method is applied to all zones it is expected to provide detailed insight into the transit supply, therefore can be helpful for planners to visualize and improve the complex and cumbersome networks in a more effective manner.
This paper presents an open source framework with key logic units design to assist with the acquisition, elaboration, storage and visualization of General Transit Feed Specification (GTFS) real-time data. It bridges the gap for researchers and planners wanting access to GTFS real time through providing the resources to understand and begin extracting GTFS data. To showcase the applicability, the framework is applied on the real data from Translink, Brisbane, Australia and the extracted data is utilised for performance measures and dashboard visualization. The codes are open sourced and are available at https://github.com/darronlim/GTFS_LiveFeed_Extraction.
