New autonomous positioning technology will benefit metro operators and passengers.
Autonomous trains have attracted a lot of attention over the past couple of years. It’s easy to see why. Autonomy promises driverless operation, higher capacity, lower costs, greatly increased reliability – and happier passengers.
Of course, modern CBTC (Communications Based Train Control) already offers those benefits. What’s more, CBTC is highly automated. So what’s the attraction of autonomous technology for metro operators and their passengers?
Taking CBTC to the next level
What makes autonomous technology so valuable is that it can be used to enhance the reliability of CBTC systems. It can even make them easier to deploy. One capability – autonomous positioning – is of particular interest.
Accurate positioning is the bedrock of CBTC operations. As well as governing safe separation between trains, positioning data makes it possible to minimise headways and therefore to increase capacity.
In conventional CBTC systems, the position and speed of trains is determined using a combination of a 2D guideway map, track-mounted transponder tags and train wheel odometers.
This system of train positioning is safe and reliable. But as with any system that depends on moving parts (odometers) and track level components (tags), there are limitations. So how does autonomous positioning work – and what are the benefits?
A brand new way to position CBTC trains
Autonomous positioning is a completely new way of doing things. In fact, it uses technologies more often found in aircraft than trains. These include inertial navigation, radar and radio ranging sensors – trusted technologies that provide exceptional reliability.
Outputs from sensors are combined using sensor fusion algorithms, which evaluate the data to provide a continuous real-time train position. This approach enables an extremely high level of accuracy – centimetre-level precision is achievable.
Thales has already successfully demonstrated this technology in a nine-month pilot project on the New York City Subway. The new positioning system was integrated seamlessly with an existing CBTC system.
How does autonomous positioning help?
The beauty of autonomous positioning is that it accelerates CBTC deployment and minimises disruption. This applies to any type of project: brownfield resignalling, new signalling deployments on greenfield projects and upgrades of existing lines. And it can be used on any type of CBTC network: metros, light rail transit and people movers. Benefits include:
Installation is easy because all sensors are fixed to the front of the train – not underneath. The entire system can be fitted in around two days per train. At the wayside, the only devices are radio ranging beacons. Transponder tags are not required, so there are no devices on the track bed.
Positioning is achieved in several different ways, making the system highly fault tolerant. For example, if radar is lost for a few seconds, positioning can be carried out by the inertial system instead. Sensors suites are duplicated at each end of the train, further contributing to resilience.
Research shows that the fusion of radio ranging, inertial navigation and radar outperforms conventional positioning. Additionally, slip-slide reading errors generated by wheel odometry are eliminated. With autonomous positioning, only the true speed of the car body is measured.
Autonomous positioning gets trains into stations faster, so trains no longer need to crawl for the last few metres to find the correct stopping position. This has the potential to shave several seconds off station-to-station run times, boosting system throughput.
Onboard sensors are inherently reliable and require little routine attention. Sensor performance and health is easy to monitor, even remotely, using predictive maintenance software. Transponder tags are no longer required, so wayside maintenance is minimal.
In fully-automated unattended systems, autonomous positioning provides a fast and accurate way to localise trains when they are initialised at the start of each working day. This can be achieved using depot radio ranging or GPS – with no need for tags.
Autonomous positioning puts operators on a migration path towards fully-autonomous operation in the future. A number of new use cases are envisaged based on train-based artificial intelligence. These include detection and response to different weather conditions, signal aspect recognition, train localisation based on computer vision and obstacle detection.
The journey is just beginning.