When looking at technology in the world of transportation, lately the news has been dominated by stories about autonomous or ‘self-driving’ vehicles…

It is tempting to look at the numerous driverless train systems around the globe and conclude that this technological stage in the urban rail industry was passed several years ago. But looking deeper, you will notice that those rail systems are ‘automated’ not ‘autonomous’. Essentially, they represent well orchestrated control systems that manage network elements and coordinate movements of trains from a central control location. Yes, they are ‘driverless’, but there aren’t any actions at the train level that can be defined as autonomous.

So, what would it take to apply autonomous vehicle technology to trains, and why would we want to do it? There are many enabling technologies converging at the moment, such as advanced sensor technology, artificial intelligence, cloud computing, IOT connectivity, and big data. All of this would point to a future where a train could eventually have similar decision-making capability to a self-driving car. But does that make it a good idea, given the fact that we already have the driverless train capability? This paper explores the challenges of autonomous trains and the underlying benefits and improvements that will drive its adoption.

What is autonomous train technology?

Advanced sensor technology is now available such as LIDAR, RADAR, cameras, and other sensors that can give a train ‘vision’. Through the use of advanced AI processors, a train can determine its location based upon identifying the environment around it without the need for additional infrastructure. The train can use that vision to determine the state of signals or other equipment along the wayside as input to decide its movements. It can also identify potential danger situations, weather impacts, and maintenance concerns. At a high level, the train can ‘think for itself’ and consistently operate in an optimized, safe manner.

Does the technological equivalent of a ‘highly experienced driver’ provide justification on its own for the adoption of so much advanced technology? A train with a vision capability would definitely provide additional safety capabilities for detection of trackside workers, objects on the track and other potential guideway dangers. For certain this will bring benefits in terms of operational flexibility and efficiency, but overall, some type of signaling system is still required to coordinate the movements amongst trains.

The initial significant impact of introducing autonomous features will be achieved by the ability to greatly reduce the amount of wayside equipment required and provide enhanced recovery capabilities. If autonomous trains use train-train (V2V) communications so they can communicate and coordinate with each other, it eliminates the need for wayside systems designed to track train location and ensure safe movements. Trains will still need to communicate to wayside equipment such as switches and platform screen doors to determine their status and coordinate their functions but the need for dedicated communication infrastructure is being superseded by the advent of other new technologies. This is where the importance of advanced wireless communications such as 5G will start to make an impact on the urban rail industry.

Eventually, through the application of artificial intelligence and Big Data approaches, the entire system of coordinated trains can be optimized according to the operator goals – optimum performance, energy savings, and so on. By incorporating the ability to access alternative data feeds such as weather data or sporting event calendars, the system can become more predictive in nature and anticipate issues before they occur. The entire rail network becomes autonomous in nature, making its own decisions to meet operational goals. (It should be noted that these techniques can be applied to automated networks as well).

Challenges in adapting this technology

If all of that sounds too good to be true, it is definitely not without challenges. Sensor technology has to be perfected to safely and securely enable vision at distances that provide safe stopping distances for trains. A distributed network of autonomous trains introduces an entirely new layer of complexity into system design, with important concerns around cybersecurity, reliability analysis, system maintenance and monitoring. The system response to failure scenarios and fallback planning approaches need to be well thought through. Implementing systems using artificial intelligence to the same safety level requirements as existing systems remains a challenge.

So, it won’t happen tomorrow. Initially we expect to see the technology appear in specific functions such as train localization and line of sight recovery functions for trains in degraded mode, for example. But just like other technological advancements, one day you wake up to see your morning train thinking by itself.

About Thales

Thales is a leading multi-domestic electronics and systems group, addressing defense and security, aerospace, space, digital identity and security, and ground transport markets worldwide. State-of-the-art technologies combined to the expertise of over 80,000 employees in 68 countries make Thales a key player in assuring security of people, assets, infrastructures, cities and nations. Thales has the capability to provide the entire service package, from design to commissioning, and offer its customer comprehensive prime contracting with responsibility for every link in the logistics chain. Similar approaches are applied to infrastructure management, simulator-based training and business process outsourcing.

The Group’s solid financial standing with revenues of more than €19 million (US$21 million) guarantees a far-reaching and long-term presence as a multi-domestic and international partner. Technological innovation is a recurring theme in all Thales businesses. Thales maintains its leading technology with focused investment of close to 20 per cent of its annual revenue in R&D. Some 20,000 researchers are involved in research and development at more than fifty sites in a dozen countries, sharing information and best practice as an integrated network of qualified R&D engineers. Thales’ current R&D efforts are focused on new design tools and methods, secure software and intelligence technologies (data fusion, knowledge bases, image processing, etc.). Another basic priority is to secure sources of high-performance electronic components and critical technologies for the long term.

About Thales in transport

For ground transportation, countries, cities, and transport operators rely on Thales to adapt to rapid urbanisation and meet new mobility demands whether locally, between cities or across national frontiers. Thales’ expertise in signalling and communications fare collection and cybersecurity gives people and goods a connected journey moving them safely and efficiently. No matter how challenging the project Thales stays by its customer’s side until the job is done. Whatever it takes.

About Thales in urban rail signalling

All over the world, Thales’s SelTrac™ Communications Based Train Control (CBTC) solution helps urban public transport operators to carry three billion passengers every year. Thales’ meets customers’ needs to improve the throughput of new or existing systems, maximize infrastructure investments, and improve passenger journey and experience. As inventor of the CBTC technology, continuous innovations are embedded in product evolution based on lessons learned with Thales’ customers. Thales has 1,400 employees in Toronto, its global competence center, and a total of 2,800 globally with the rest of the world. The company has worked on very complex existing networks in New York, London, Paris, Hong Kong, Singapore, and also helped cities build new networks in Dubai, Doha, Santiago, Vancouver, Kuala Lumpur, and Seoul.

Tel: +1 (416)7484401 

Email: [email protected]

Visit: www.thalesgroup.com