In October, Dr Ning Zhao, assistant professor in the Electronic, Electrical and System Engineering department at University of Birmingham and a member of the Birmingham Centre for Railway Research and Education (BCRRE), gave a presentation titled ‘Towards Integrated Driver Support and Hazard Detection in Modern Tram and Metro Systems’ for the Institution of Railway Signal Engineers (IRSE). It focused on extending the safety features used in traditional main line signalling into the world of trams and providing systems more appropriate for light rail, and featured the latest technology. The work is also supported by Professor Stuart Hillmansen and Mr Yiqing Liu.
Dr Zhao explained an optimised smart Driver Advisory System (DAS) and a real-time Intelligent Obstacle Detection System (IODS) which has been field-tested across several light rail tram networks. The systems have demonstrated significant energy savings, improvements in driver consistency, and operational safety improvements. The presentation covered the deployment results, driver acceptance data, and implications for future digital light rail systems worldwide.
UK Rail Research and Innovation Network
BCRRE has been involved with research in the rail industry for over 50 years. A team of over 145 researchers and support staff combine academic excellence with real-world innovation through industry collaboration and a focus on whole systems thinking across rail and in the context of wider transportation. BCRRE is also the lead academic partner in the £92 million UK Rail Research and Innovation Network (UKRRIN), a collaboration between academic centres of excellence and the rail industry.
It was awarded the 2017 Queen’s Anniversary Prize for Higher and Further Education. Students study railway systems, including BEng/Meng, MSc students taught in Birmingham, PGCert/PGDip/MSc students taught in Singapore, and PhD students.
The University is home to one of UKRRIN’s four Centres of Excellence – the Centre of Excellence in Digital Systems (CEDS). Together with a growing number of industrial members and managed by a coordinating hub run by the Rail Safety and Standards Board (RSSB), UKRRIN is a key part of innovation in the rail industry, bringing together academic centres with industrial partners from the rail supply chain. These include Siemens, Bombardier, SMRT, Unipart, Hitachi, Thales, British Steel, AtkinsRéalis, RIA, AECOM, Pandrol, Progress Rail, and Aggregate Industries.
UKRRIN was formed by a consortium of universities, in collaboration with existing industry testing and trialling facilities such as Network Rail’s Rail Innovation and Development Centres. The other UKRRIN centres are:
- Digital Systems (led by University of Birmingham, in partnership with Lancaster University, Imperial College London, Swansea University, and University of Hull).
- Rolling Stock (led by University of Huddersfield, in partnership with Newcastle University, Loughborough University, University of Cambridge, University of Bristol, Brunel University, and University of Nottingham).
- Infrastructure (led by University of Southampton, in partnership with the University of Nottingham, the University of Sheffield, Loughborough University, and Heriot-Watt University).
Light rail
Light rail tram and metro systems have an important role in providing urban transportation. The Office of Road and Rail (ORR) defines light rail as “an urban transportation system that generally uses electrically-powered, rail-guided vehicles along exclusive rights-of-way at ground level, on raised structures, in tunnels, and in streets. Light rail systems generally use lighter equipment that operates at slower speeds when compared to mainline or heavy rail metro/urban railways.”
The systems include:
- Tramways – operating in a highway environment or other public space.
- Metro – light rail that operates entirely on segregated tracks, but using lighter weight vehicles than found on the national railway or London Underground networks.
- Tram-train – where light rail vehicles can operate on segregated alignments and additionally, on mainline “heavy” railway lines shared with conventional trains.
- Very Light Rail (VLR) – lighter weight versions of light rail for routes where the level of patronage is lower than for many urban mass transit systems.
Signalling
The signalling for light rail systems include ‘line of sight’ principles, when operating relatively slowly in urban centres (and similar to driving road vehicles), through to conventional aspect signalling on higher speed lines. From a commercial point of view the competition to light rail tram and metro systems is from road vehicle bus transportation. Therefore, the cost of signalling systems used on traditional heavy rail systems can be unaffordable.
Smart autonomous driving systems developed for road vehicles may provide solutions for light rail systems. However, this is not always the case, as the operating and driving characteristics are different for light rail compared to road vehicles.
BCRRE’s research into autonomous systems for light rail comes in two parts, with the objectives of enhancing safety and efficiency. The first is to provide intelligent ‘smart’ driving assistance to drivers in order to save energy, with the second part providing tram protection using LiDAR or cameras to identify hazards up to 500 metres ahead to the smart driving system, helping drivers to control the trams in an emergency.
Edinburgh trams
Dr Zhao began by explaining the BCRRE smart driver advisory system development, trial test, and application. Any rail vehicle on a movement goes through four phases / modes. These are: (i) motoring or accelerating mode; (ii) cruising mode; (iii) coasting mode; and (iv) braking mode.
The aim and objective have been to find the most appropriate train movement sequence to minimise energy usage within a constant total journey time. It was found that in their training drivers had not properly learned the benefit of and the need for appropriate coasting, especially in off peak services where the journey time would not be affected, but significant energy savings could be made. A number of algorithms have been developed, implemented, and evaluated in the optimisation for different scenarios. The result is that an impressive up to 21% energy saving can be achieved according to several trial tests and applications results.
A case study was undertaken on Edinburgh Trams, operated by Transport for Edinburgh. This connects York Place in the city and Edinburgh Airport with 15 stops and a total length 14km, with a 750V overhead line power supply system. The route was modelled taking into account the gradients and curves and various driving styles evaluated to identify the most efficient.
A simple tram smart Driver Advisory System (DAS) has been developed with the optimal driving strategy embedded into the DAS for the trial testing. The driver controls the tram in accordance with the instructions displayed by the DAS and the DAS is able to work standalone or connected to the tram CAN bus system to obtain and display real-time train operational data.
From talking to other tram operators around the world it was thought that any savings would be in the order of 2%, however the Edinburgh Trams trial concluded with the following findings.
Following the excellent results obtained in the field tests, Edinburgh Trams has implemented the optimal driving trajectory in practice. Training has been carried out for Edinburgh Trams drivers to help implement the energy saving features of the optimisation system, and Edinburgh Trams is now using the optimal driving strategy in their daily services for all drivers and trams.
There was a concern that the drivers needed to observe the smart DAS screen which could be a driving distraction, so simple bespoke line side ‘coasting’ boards have been developed and deployed.
Edinburgh Trams has also won an award for deploying and using the BCRRE smart driving technology. Other trials, including Nottingham Express Transit (NET), Manchester Tram (Metrolink), Beijing Yizhuang Metro Line, and Guangzhou Metro Line No.7, have resulted in similar energy savings.
Another key benefit observed on several light rail networks is that the technology improves driving consistency across different drivers. Traditionally, timetable planning must account for variations in individual driving behaviour and skill levels. With the support of the smart driving system, however, driver performance becomes much more consistent, resulting in highly repeatable driving patterns. This allows for more efficient and reliable timetable design, achieving a level of performance comparable to that of more expensive ATO systems used in light rail.
Intelligent Obstacle Detection System (IODS)
The ultimate objective of IODS is to project the light rail vehicle using LiDAR or cameras to scan the area ahead and monitor objects in front of the train in real-time and intelligently predict their movement trend. The information can then alert and assist the rail vehicle or the driver to make judgments as early as possible, thus reducing the accident risk.
Dr Zhao explained how BCRRE evaluated LiDAR and camera technology, and developed an Intelligent Obstacle Detection System specifically for railways. It was identified camera detection technology, which is used in road vehicles and can typically detect 20 to 50 metres ahead. This may be acceptable for road vehicles, but isn’t far enough ahead for rail. Camera-only solutions can also have difficulties in poor light and weather conditions.
While research may continue with camera obstacle detection, LiDAR has already demonstrated that it can recognise vehicles and pedestrians within a 500-metre range in front of a moving rail vehicle. The system will therefore will be able to detect and classify potential track obstacle hazards such as vehicles, pedestrians, debris, vegetation buildup, and structural defects.
The technology was trialled at BCIMO before it went into administration, and on the Coventry Very Light Rail (CVLR) vehicle, and is capable of detecting pedestrians, vehicles, and other potential hazards ahead of the tram, as well as intelligently predicting their movement trajectories. This enables the system to provide timely warnings and guidance to the driver, supporting safer and more proactive driving decisions.
The combined use of camera and LiDAR technologies ensures robust performance across a range of environmental conditions, including poor lighting and adverse weather. The system has demonstrated high accuracy and reliability over a long distance, making it suitable for integration into modern light rail operations. A video comparing LiDAR and camera obstacle detection, and the benefit of LiDAR can be found here.
Future collaboration
The integrated solution is part of the broader ambition to advance train autonomy. Dr Zhao’s team is continuing the collaboration with CVLR to further develop automated tram operations, while also adapting the system to support wider sensing applications (e.g., bridge strike prevention). These efforts demonstrate the versatility and scalability of the proposed solution.
The team welcomes collaboration opportunities with industrial partners across the rail sector. Dr Zhao can be contacted at [email protected].