The CRC for Rail Innovation recognises the importance of investing in the future of collaborative research.

To date, the CRC has awarded 77 CRC Research Scholarships to PhD and Masters Candidates across our seven partner universities to enable post graduate students to carry out further research in topics that span all of our themes. Many of the student topics are aligned to our CRC projects.

To gain an appreciation of the breadth of this research, click on the names below to learn more.

Safety & Security

Annette Sommerville - Informing the development of baseline near miss metrics for rail level crossings

Masters Candidate - Central Queensland University

Despite industry attempts to reduce the number of collisions at rail level crossings they still occur and are costly in both monetary terms and in human lives. These collisions may be analysed to inform prevention understanding. However near misses at rail level crossings are anecdotally understood to be much more prevalent and present a greater opportunity to evolve understanding of causation. Unfortunately near misses are also known to be under-reported by train drivers. Accordingly this research aims to identify factors that contribute to the under-reporting of near misses, establish why train drivers choose not to report in particular situations, establish driver rationale at the moment when the decision is made that a near miss has occurred, and determine if that decision is based on regulatory standards or driver experience. In this study train drivers will be surveyed to ascertain their opinions in relation to near misses at rail level crossings. Their responses will be grouped to reflect common opinions and focus groups of rural and metropolitan drivers will be facilitated to further explore the underlying causes leading to the under-reporting of near misses. Outcomes of these focus groups will be used to illuminate the future improvement of near miss metrics.

Aligned to project Baseline Rail Level Crossing Video

Dr Hamideh Anjomshoa - Optimal placement of passing bays in underground mines

PhD Thesis Abstract - University of South Australia

In underground mines, haulage vehicles carry ore from underground loading bays to the surface. The vehicles travel in narrow tunnels, called declines, with occasional passing bays that allow descending empty vehicles to pull off the main path an wait for ascending laden vehicles to pass. The number of passing bays and their locations influences the delays to descending vehicles, and hence the haulage productivity of the mine. Constructing passing bays is costly, and as there is a trade-off between the cost of building additional bays and the productivity of the haulage system.

Optimising the locations of passing bays in underground mines has been largely ignored in the literature. Most mines use trucks which can reverse into existing bays or lateral infrastructure, though reversing is undesirable. Some mines are considering using road trains (trucks with trailers), which cannot reverse, and so the locations of purpose-built passing bays is very important for haulage efficiency. Railways face a similar problem of finding the best locations for passing on single-line corridors: most studies use a simulation to explore a limited set of options.

The thesis investigates two approaches to analyse the placement of passing bays in underground mine: simulation, and optimisation using a Mixed Integer Programming (MIP) model. Although our case study is for an underground mine, the same methods are applicable to the study of passing loops on a single-line railway.

Our simulator models the movements of vehicles on a decline with specified bay locations. Vehicle movements are controlled to avoid deadlock, a situation where vehicles cannot move forwards without at least one vehicle reversing. Our deadlock avoidance scheme is less restrictive than other schemes described in the literature.

Results from the simulator confirm that the spacing of passing bays can have a significant impact on haulage productivity. If bay spacing is large then it is sometimes best to send all vehicles down to the loader, then bring them back to the surface. Decreasing the spacing of passing bays allows descending vehicles to be interleaved with ascending vehicles, increasing productivity. We show how to calculate critical spacing of bays to allow interleaving without delays.
We formulate an MIP to determine the placement of bays that minimises the time taken to complete a specific number of haulage trips. The number of vehicles and bays are specified. In addition to finding the optimal placement of passing bays, the MIP gives a schedule of vehicle movements. We consider two different approaches to reduce the number of alternative optimal solutions, favouring solutions that have delays at the surface rather than underground. The MIP show that the optimal schedule sometimes requires vehicles to delay at the loader after loading.

We use results from the MIP to conduct an extensive analysis of al scenario with three vehicles operating on a single decline with two passing bays, and show how the optimal placement of bays and the minimum makespan varies with loading and unloading durations on the optimal placement of bays, and consider an alternative declined topology.

Overall, the thesis shows how proper optimisation of bay locations and vehicle schedules, tailored to the network and vehicle parameters, can have a significant impact on haulage productivity.

Anjomshoa, H (2011) Optimal placement of passing bays in underground mines (Doctoral Dissertation, University of South Australia, 2011). Retrieved from Trove.

Aligned to project Dynamic Crew Allocation

Hajananth Nallaivarothayan - Human Motion Analysis form multi-camera surveillance video for automatic observation of risky pedestrian behaviour at railway crossings

PhD Candidate - Queensland University of Technology

Inhi Kim - Railway safety using traffic simulation.

PhD Candidate - University of Queensland

In my research, the integration of traffic simulation and driving simulator will be implemented in order to accomplish a better picture to reduce crashes at railway crossings. Various Intelligent Transportation Systems will be evaluated through a test of driving simulator. All relevant data from driving simulator will be an input to traffic simulation, and vice versa. Taking into account the new approach, it is expected that new safety systems will be evaluated to be chosen for a railway crossing based on economic and safety point of view.

Aligned to project ITS for safer level crossings

Mayya Spiryagina - Socio-technical analysis of level crossing design for pedestrian use.

Master Candidate - Central Queensland University

My research aims are conducting a study on the pedestrian behaviour at railway level crossings that includes the investigation and the analysis of the complex socio-technical system which combines human, engineering, legislation and interaction issues involved in the process. In my research, in-depth analysis which involves quantitative and qualitative research methods is one such aspect, and it is going to be applied by means of the introduction of some new approaches used in the aviation industry. Another aspect is the application of the prevention accounting technique which has been commonly used in the field of occupational health and safety previously in order to improve existing cost-benefit analysis, but has not been used to assess the impact of human factors. The realisation of these two aspects will give a possibility for better understanding of pedestrian behaviour not only from a psychological point of view, but will also allow judgements to be made about economic efficiency of existing technical solutions based on the return on prevention or the return on investment criteria.

Aligned to project Understanding Pedestrian Behaviour at Level Crossings

Sina Aminmansour - Video analytics for the detecting events of interest on approach to railway level crossings

PhD Candidate – Queensland University of Technology

Recent modelling of socio-economic costs by the Australian railway industry in 2010 has estimated the cost of level crossing accidents to exceed $116 million annually. To better understand causal factors that contribute to these accidents, a data collection system is being developed by the CRC for Rail Innovation to capture quantitative data around near-miss incidents at railway level crossings. The system will collect data from a range of different data sources including train logs, geo-referenced asset data, operating schedules, and video data from forward-facing cameras installed on trains.

My PhD research project aims to design and implement Computer Vision and Pattern Recognition algorithms for automatically detecting near-miss incidents from the forward-facing cameras installed on trains. A Video Analytics module will be developed as the result of this project for the CRC for Rail Innovation’s data collection system. The module will provide algorithms for detecting near-miss occurrences related to the speed, trajectory and position of vehicles and pedestrians at railway level crossings; and the detection of signal state and signs to facilitate the investigation of factors that contribute to Signal Passed at Danger (SPAD) incidents.

Aligned to project Baseline Rail Level Crossing Video

Teodora Stefanova - Cognitive and motivational factors contributing to pedestrian's risk-taking at level crossings according to environmental characteristics and differences between Australia and France.

PhD Candidate - Queensland University of Technology

Aligned to project Understanding Pedestrian Behaviour at Level Crossings

William Thomsen - Towards affordable safer level crossings: Human factors specifications for assessing new technologies for passive crossings

Masters dissertation - Central Queensland University

The Australian rail industry has declared that eliminating the loss of life at level crossings and all that lies behind those tragedies are its highest priority. Federal and State governments have made great efforts to enable the industry bodies to research, develop and trial new systems which may enable road users to respond more safely, especially at passive level crossings where no flashing lights, audible alarms or boom-gates operate. Passive crossings may be controlled by either ‘Give way’ or ‘Stop’ signs at the crossing, with the train pictogram advance warning sign being the first indication of a crossing available to road users. The second advance warning is a crossing geometry sign for ‘Give way’ crossings, or a ‘Stop sign ahead’ warning for those crossings controlled by ‘Stop’ signs.

This study has been a part of the Cooperative Research Centre for Rail Innovation’s Project R3.111 – New Affordable Level Crossing Protection Systems. It has investigated the issues underlying road user error from the Human Factors Engineering perspective. This has revealed a wide range of factors which have an impact on what road users see as they approach a level crossing, how they understand the system of warning and regulatory signs provided at passive level crossings, what influences their decisions about obeying the mandatory requirements of ‘Give way’ or ‘Stop’ signs at the crossings, and what new systems need to provide to bring about responses which help to eliminate collisions at passive level crossings.

Much has been explained by researchers over recent decades about the need for improvements in the alerting stimuli provided at crossings. The most significant step taken in the Australasian setting was the 1992 approval of the inclusion in Australian standards of the ‘train pictogram’ image used for passive crossings to differentiate them from actively controlled crossings. There is a strong reason for this – if a different response is required, a different stimulus needs to be provided. This differentiation is critical for road users to realise that passive crossings do require a different suite of responses from those required at actively controlled crossings. Previously, all advance warning signs for level crossings were the same.

A survey of over 1200 Queensland road users to determine their self-reported responses to ten different passive crossing situations was carried out. While this revealed predominantly appropriate responses, there are still too many road users choosing risky options. This is a powerful driver for developing systems with user-centred design to achieve a reduction in level crossing collisions. The other important point to emerge from analysis of these responses is that there are two strong factors – perception and behaviour – which polarise the problems needing to be addressed in the human factors design criteria, the primary output from this study.

Another survey of over one hundred road users from the Rockhampton district provided important insights about their awareness of, and responses to, level crossing signs frequently seen at level crossings in the area. This showed that there was widely varying knowledge of the standard signs and road users’ required responses to them. That survey also gave helpful insights to road user opinion about several human factors issues which have a bearing on level crossing designs and user response to them.

A third part of this study was the direct observations of eight passive crossings in different urban and semi-rural settings. More than 1600 vehicle passages through crossings were observed and recorded. The actual behaviours observed showed a very disturbing difference from the self-reported data. The results indicated that 38% of observed passages through the eight crossings did not comply with the ‘Stop’ requirement correctly, a much higher proportion than indicated by the first survey.

Based on these data, and human factors design elements listed in Appendix 2, the Design Criteria have been compiled in Appendix 12. This set of criteria identifies requirements both for the rail industry, government and statutory authorities, and for design engineers who will be assessing new technologies for human factors compliance and technical reliability in Stage Two of the CRC for Rail Innovation project. There is a clear need to re-evaluate the present ‘puffing billy’ for the advance warning sign, and more attention-grabbing enhancements to improve road user perception of the three sign system at passive crossings need to be trialled.

Speed limits and distance settings should also be reviewed with a view to increasing the time available for road users to respond safely to the protection system.

Thomsen, W R (2011) Towards affordable safer level crossings: Human factors specifications for assessing new technologies for passive crossings (Masters Dissertation, Central Queensland University, 2011) Retrieved from Trove.


Aligned to project Affordable Level Crossings