Transforming trains
How digital engineering helps design the trains of the future
Travelers of decades past would recognize the trains of today, even if the design, speed, interiors, and the way people use them are all somewhat different. And today’s travelers would probably recognize tomorrow’s trains - the fundamental idea of a locomotive running along a track hasn’t changed, and probably won’t.
But look closer - a time traveler to 30 years from now will see many differences that meaningfully change what a train is. Future trains will likely have much more autonomy, use zero emissions propulsion systems, and be made of new materials. They will be faster, greener, and more autonomous.
Improvements must be made, even though much of today’s rolling stock is required to last another decade (at least). But designing this new generation of trains will impose significant engineering challenges, as more complicated technologies need to be developed and integrated. Yet, even as complexity increases, we will need to do this engineering and manufacturing faster.
A new train takes roughly five years from the initial design stages to qualification. That will likely be more if this train has to integrate brand new technologies. Twenty years ago, car designs also took five years to hit the road, but automotive manufacturers have managed to compress that timeframe to just two years, largely thanks to digital approaches to design, development and testing. Rail has not seen the same time compression yet, though it could. Doing so would speed innovation, lower costs, and help the rail industry meet its net zero goals.
Once a train design is approved, it can be very difficult to alter due to strict safety rules. So, we must also make trains more modular and flexible, for example, through the softwarization of components so that trains are not constrained by their hardware and can evolve through software updates to meet new needs.
Let’s start with an example of train innovations, and the role of digital in making rail greener, better, faster and stronger.
Alstom is developing the Coradia iLint, the world’s first hydrogen fuel cell powered train. Another major rail player is developing hybrid propulsion systems that combine diesel and battery power. Both can operate on non-electrified tracks more cleanly.
Such projects can benefit from digital modeling to optimize the propulsion systems for weight reduction and energy efficiency. Moving up another layer, researchers in Korea have used a digital twin to model 30 years of operation of a proposed train operated by solar and hydrogen, showing it would be a good substitute for the diesel locomotives currently operated in South Korea.
Similarly, digital simulation allows researchers to model the behavior of new materials – such as those that are lightweight, stronger, or with greater impact absorption – and predict how they will perform in real-world applications. For instance, the Vehicle Technologies Office at the Department of Energy (DOE) in the United States uses computational materials science to enhance the understanding and development of advanced materials like high-strength steel, aluminum, and carbon fiber composites, supporting the design of lighter and more durable components for a range of vehicles.
Similar modeling approaches support digital crash tests, allowing engineers to simulate the impacts and stresses that materials and components will face during collisions. These can identify potential weak points and help optimize the design to improve crashworthiness.
Digitizing trains
Whilst digital engineering technologies are helping with the physical design, the next generation of trains will themselves become partly-digital thanks to sensors, control software, and algorithms.
These new sensor and connectivity technologies could be highly valuable for getting more from current trains. So, let’s start by looking at examples of how technologies could be put to use.
Predictive Maintenance: Various sensors can be installed on trains and tracks to monitor the condition of equipment continuously. For example, vibration sensors and accelerometers on bearings and wheels, temperature sensors on motors, and acoustic sensors for detecting anomalies in the railway track. Data collected by these sensors can be analyzed using machine learning algorithms to predict when parts might fail or require maintenance, allowing trains to schedule repairs before breakdowns occur, reducing downtime and maintenance costs.
Energy Management: Energy consumption sensors can help monitor and optimize electricity use. This can include monitoring the efficiency of the train's propulsion system and the use of regenerative braking systems that capture and reuse energy, as well as onboard systems like lights and doors. Algorithms can analyze fleet energy usage patterns and optimize energy consumption, for example, by adjusting speeds or maximizing the use of regenerative braking.
Safety Enhancements: Technologies like LiDAR, cameras, and infrared sensors can be used for obstacle detection to prevent collisions. Track health monitoring sensors can detect track faults or failures that could potentially lead to accidents. Automatic Train Protection systems can use these sensor inputs to further enforce safe operation.
Passenger Experience: Sensors can monitor onboard conditions such as temperature, air quality, and occupancy. This data can be used to adjust environmental controls automatically, both for comfort and cost efficiency.
Digital technologies can vastly compress timeframes for engineering design and delivery of new and upgraded trains, without compromising safety and quality, as has been shown by other safety critical industries. And indeed, rail can learn a lot from such industries when it comes to digital engineering, from automation in automotive to simulated testing in aerospace and defense (A&D).
We must, of course, acknowledge that many of these technologies are new to rail. Whilst there are some quick wins to be had, truly transformative technologies like digital twins will take a lot of time. We must be careful to ensure that any such project is managed right, with clear goals and deliverables in mind, to ensure it delivers an ROI, and that we don’t end up throwing money into a digitization black hole on the vague promise of some future benefit.
Nonetheless, these technologies’ potential is undeniable. Consider one example. US freight railroads test a new locomotive model for 30-50 'locomotive years’ before it is certified to use. This means a railroad may need to test ten units for 3-5 years to gain enough data to know whether to accept it or not. But rail is expecting a mandate to migrate away from diesel equipment (California’s deadline is 2035), yet there are not enough battery-electric locomotive (BEL) prototypes to test today or in the foreseeable future.
Hypothetically, a BEL manufacturer could create a digital design and have it tested virtually in less time than the physical models could hope to achieve. Based on already developed designs for hybrid battery/diesel electric multiple unit (EMU) passenger trains, we could – with a good case to regulators backed by data – probably reduce the need for extensive physical testing. If we could start collecting data, we could ensure that soon enough to make a difference, and plan for BEL being fully on track by 2030.
Since locomotives are complex machines with thousands of parts, and the work has not been done to build such high-fidelity models, this is not realistic today. But could a fully digital design and test regime someday be sufficient to satisfy operational railroad buyers, safety regulators, and emissions authorities? This remains a goal, a hope, and a major market opportunity all rolled into one.
While the fundamental concept of a train remains unchanged, tomorrow's trains will be more comfortable, more user friendly, faster, greener, autonomous, and more convenient, driven by advanced technologies. The engineering that will get us there will require faster development cycles, akin to that of the automotive industry, delivered through digital approaches that balance rapid prototyping and digital continuity with safety and rigorous testing.
