Russia: A dedicated engineering subsidiary of Russia’s largest rolling stock manufacturer, TMH, was established in 2018. It consolidated fragmented, plant-based teams in different regions to a unified centre of cross-functional expertise to enhance R&D efficiency amid the growing complexity of new product development. TMH Engineering now employs over 1,200 design and process engineers.
TMH Engineering CEO Dmitry Petrakov gave a comprehensive interview on the company’s development and current operations in the latest issue of the corporate magazine Vektor TMH. He discussed designer collaboration within the holding, the intricacies of creating new locomotives—including the future mainline gas turbine locomotive 3TE30G and electric locomotive 2ES9—features of 1D-modelling, and much more. ROLLINGSTOCK publishes his quotes in full.
TMH Engineering CEO Dmitry Petrakov. Source: TMH
Let us take a step back, Mr Petrakov. What does TMH Engineering represent today? The company was built by consolidating engineering units from different plants. Could you tell us how it all began and what proved most challenging along the way?
You’re going straight to the heart of the matter. The idea of building TMH Engineering as a unified structure was originally put in place by Yuri Orlov, who now continues his work as Chief Designer. At that time, I was working as Chief Designer at the Bryansk plant, and later I took over, developing a concept that was already clearly defined.
Today, TMH Engineering operates across ten cities. From the outset, it was clear that this would not be a simple task. Our main structural feature is geographical dispersion. Historically, engineering teams were deeply embedded in the day-to-day business processes of their respective plants. Their work rhythms, priorities and decision-making were shaped by the needs of large manufacturing sites. Bringing these teams together meant more than formal integration—it required a change in mindset.
At its core, engineering is about deploying the strongest available expertise exactly where it is needed. In practice, this means that electrical equipment specialists or electric machine engineers from Novocherkassk can work seamlessly with rolling stock designers from Tver on the same project—for example, a diesel locomotive developed for the plant in Bryansk. That level of cooperation is only possible when everyone operates within a shared framework: common procedures, unified standards and a single set of core documents.
Meeting at TMH Engineering’s unit. Source: TMH
Building that framework has been one of our key priorities over the past several years. And the process is still under way. In truth, it will never be fully complete. Engineering evolves constantly: processes are refined, approaches adjusted, solutions improved. That said, if we assess the maturity of our unified methodologies and documented procedures today, I would put it at around 95%.
Another challenge lays in regional diversity. Engineering teams in St Petersburg, Russia’s cultural capital, and in Novocherkassk, the historic centre of Cossack traditions, operate in very different cultural contexts. Communication styles differ, as do attitudes to feedback and criticism. People performing identical functions may interpret the same task in completely different ways. Coordinating designers from multiple regions within a single large-scale project is anything but straightforward. But we have learned how to manage this complexity, so we continue to move forward, step by step. Like TMH as a whole, we operate within a philosophy of continuous improvement. There is no final point—only the next stage of development.
When engineers come from different schools and professional cultures, disagreements are inevitable, sometimes even on fundamental issues. How do you manage situations like that?
In fact, disagreements like those arise all the time. And not only between different schools; they can just as easily occur within the same team, when two designers hold opposing views.
That is why we rely on a clearly defined decision-making hierarchy. At the top, we have a chief designer, supported by deputies responsible for specific areas: passenger transport, urban transport and special-purpose equipment, locomotives and service. Working closely with the business divisions, they are accountable for the products and supervise project chief designers. These project leaders take full responsibility for their products from meeting technical specifications and customer requirements, both internal and external, to ensuring that solutions are delivered to market on time. Within each discipline, chief designers as competence leaders make key technical decisions. For example, the chief designer responsible for braking systems decides what type of compressor should be used, oil-free or oil-based, screw or piston. It is precisely within these decision frameworks that most disagreements tend to arise. But that is not a problem in itself. A controlled, professional disagreement can be productive. As we often say, truth emerges through debate. And if an issue cannot be resolved at a given level, it is escalated further, ultimately to the chief designer. At that level, every question finds a resolution.
There is also a degree of professional rivalry not only between sites, but within teams themselves. That, perhaps, is simply human nature. The challenge is to build processes that allow people to rise above it. One of the ways we address this is through mentoring. Each experienced senior engineer works closely with younger colleagues, passing on not just technical knowledge, but also a professional way of thinking and behaving. This helps shape a culture where individual ambition does not dominate. In the end, we are engineers. We deal with kilonewtons, amperes and volts. That reality leaves little room for subjectivity— decisions have to be objective, balanced and constructive.”
TMH is known for its high level of integration across the entire chain, i.e., from development to manufacturing to service. Close feedback from Russian Railways, Russia’s national operator, also allows you to observe real world operation. How does this influence your work? Can you give examples of projects that benefited from this approach?”
Even at the prototype stage, our locomotives are already demonstrating higher reliability than their predecessors. And that is a direct result of this end-to-end feedback loop. The effect is particularly clear in projects such as the 3TE28 and TE26.
TE26 freight and passenger diesel locomotive. Source: Alexey Stolchnev/ROLLINGSTOCK
We have recently taken this approach a step further. Together with the locomotive division, we carried out a joint serviceability inspection of a TE26 prototype, something that had not been done before. Engineers worked side by side with representatives from service and other TMH units. By reviewing and validating operational feedback together, we were able to see that we are moving in the right direction.
Are there examples of specific design improvements introduced as a result of feedback from operation and service?
Yes, and one example may seem modest at first glance, but it is actually extremely important. We redesigned the diesel engine cooling system drainage mechanism on the 3TE28 and TE26 locomotives. Previously, maintenance crews had to carry heavy hoses several metres long, connect them and spend up to two hours preparing the system. Now imagine doing that in winter—somewhere in Russia’s Far North, say, near the town of Tynda, in temperatures of minus 50 degrees Celsius. We introduced a direct connection to the pneumatic system, and today service personnel can purge the system in minutes simply by closing two valves. The solution itself is not particularly complex. Yet for forty years this problem remained unresolved. As a result, something that took two hours now only takes two minutes.
There are many other examples. We moved to a traverse-free brush assembly design in traction motors, installed maintenance-free batteries, and replaced plain axle-hung motor bearings with rolling bearings. This eliminates the need to add oil every 120 operating hours. For service teams, this represents a significant improvement. And there are dozens of such targeted changes.
Let’s talk about science. TMH operates the Perspective Technologies Centre. How do you work together, and what kinds of projects have you delivered jointly?
Let us begin with a simple but essential principle: engineering design and research and development are inseparable. Any engineer is, by definition, a researcher. It is vital for designers to remain involved in R&D, even though the reality of the profession is less romantic than it might sound: around 80 per cent of an engineer’s work is routine, and only about 20 per cent leaves room for true creativity.
In practical terms, we divide R&D into two distinct streams. The first covers research with a high probability of real-world implementation. These projects may not lead to fundamental scientific discoveries, but they can be brought into production quickly—sometimes using our own capabilities, sometimes in partnership with external organisations. This is largely our responsibility. A good example is our R&D work on identifying complex structural elements for a next-generation traction gearbox, which we carried out in cooperation with Bauman Moscow State Technical University.
The second stream involves research with the potential to deliver a genuine technological breakthrough—but with a much higher degree of uncertainty. This is where the Perspective Technologies Centre plays a leading role. We work closely to ensure a clear division of responsibilities, while at the same time jointly introducing promising technologies into practical engineering.
One of the projects currently under way is the TE26 diesel locomotive. Together with the Centre, we are testing a new roof design and working with composite materials. If the results meet our expectations, this will give us a new engineering solution that can later be scaled across other platforms. In simple terms, the Centre’s task is to explore, experiment and take risks, while ours is to turn those ideas into reliable, efficient rolling stock.
EMKA2 contact-battery shunting locomotive. Source: TMH
One of the most notable examples of this collaboration is the EMKA2 contact-battery locomotive. It was developed on a new platform originally created for the TEM23 shunting locomotive. For the first time, this allowed us to combine diesel and electric traction within a single platform. The development of the traction energy storage system—a key element of the project—was a direct contribution of the Centre for Advanced Technologies.
You are currently working on a wide range of innovation projects: from a hydrogen train to a new electric locomotive to a gas-powered diesel locomotive. Which of these are closest to completion, and which are the most compelling from an engineering perspective?
We really do have a substantial portfolio of projects that are approaching a high level of readiness—and several of them, in the best possible sense, will come as a surprise to the industry. Among the projects that have already moved from concept to reality are the TEM23 shunting diesel locomotive and the EMKA2 contact-battery electric locomotive. These are the first shunting locomotives in Russia to feature asynchronous traction drives and, more importantly, a modular architecture. The degree of unification is so high that, to a non-specialist, the difference between them is almost imperceptible: one has exhaust stacks, the other a pantograph in their place. By the way, the TEM23 is serving as a test platform for the Avtomashinist autonomous driving system.
For engineers, these modules are not just design solutions—they are a strategic tool. They allow us to accelerate the development of entire locomotive families. In effect, we have built a technological foundation that shortens the path from concept to production for each new generation.
As this platform continues to evolve, we will expand the catalogue of modular components, making each subsequent version faster and more cost-efficient to bring to market. On the mainline side, we have completed the full design documentation for a new generation of heavy-haul locomotives—the 3TE30 diesel and the 2ES9 electric. Both are intended for operating heavy freight trains on the Eastern Railway Network of Russian Railways. Production has already begun, with body structures now being fabricated, and we expect the first prototypes to appear next year.
Render of gas-powered diesel locomotive 3TE30G. Source: TMH
What sets these locomotives apart is not only the use of asynchronous traction drives or their superior tractive performance—surpassing both previous generations and competing designs. The real breakthrough lies in a fundamentally new level of operational resilience. Our aim is to eliminate in-service failures as far as possible. This is being achieved through extensive redundancy in critical systems, combined with advanced diagnostic technologies. Taken together, these measures represent a radical change, moving our rolling stock into a different reliability class.
The 3TE30 also introduces a multi-fuel concept. It can operate either on conventional diesel fuel or in a gas-diesel cycle, using a blend of diesel and liquefied natural gas (LNG). Its deployment could become a significant milestone for the rail sector as a whole.
Will operating on LNG have any negative impact on service life?
In theory, it should not. From a physical standpoint, LNG has a higher calorific value, which does increase thermal loads on the internal combustion engine. That is exactly what our engine development teams are addressing now—ensuring that this factor has no noticeable impact in real-world operation.
Is it fair to say that you are prioritising technologies that can be scaled across different classes of rolling stock?
Absolutely. Any increase in series production works in everyone’s favour: it improves cost efficiency, makes reliability easier to manage and simplifies service. But there is more to it than that. Our goal is to make the control environment identical across an entire locomotive family—whether diesel or electric. We are very close to achieving this. Step into the cab of the EMKA2 or the TEM23, and unless you notice the presence or absence of a pantograph at the door, it is not immediately obvious which locomotive you are in.
This approach serves a very practical purpose. It simplifies training for operating and maintenance staff. Russian Railways can, if needed, redeploy train operators between different routes without friction, and service companies benefit from a much more straightforward maintenance model.
You have spoken about the goal of minimising in-service failures. What exactly enables that? And can these solutions be scaled to other types of rolling stock?
The answer lies in a combination of several design principles. The core idea is redundancy—duplicating systems that are critical to operation. For every product, we clearly understand which components would cause a locomotive to drop out of service if they failed. Those risks are analysed one by one. Where necessary, we hedge them by design. In control systems, for example, we use duplication: different hardware sets operate in parallel. If one fails, the other immediately takes over.
Then there is redundancy at a deeper level. Take equipment such as compressors or cooling fan motors. We use auxiliary converters that allow us to manage energy resources more flexibly. If one converter fails, the system automatically switches auxiliary machines to another. And as a last line of defence, direct connection always remains available. In simple terms, the rolling stock itself actively maintains its own survivability—and operation continues.
A particularly telling example is the 3TE30 diesel locomotive. For the first time, we have implemented redundancy at the level of the main power unit—the internal combustion engine itself. Two sections are linked through a common intermediate electrical circuit. If one engine fails, the train continues its journey on the other. Yes, at a lower speed—but it keeps moving. The mainline is not blocked, and there is no need to wait for a rescue locomotive. This solution is our own know-how. As far as we are aware, no other manufacturer currently applies this approach.
At the Kolomna-based unit of TMH Engineering. Source: TMH plant in Kolomna
We also build in a reserve of tractive effort. For example, hauling a 7,100-tonne freight train requires approximately 85 tonnes of tractive force. Our locomotives are designed for a continuous tractive effort of 100 tonnes. This means that even if one axle fails, the locomotive still has enough power in reserve to pull the train through.
That margin makes all the difference in real operation. The final element is predictive diagnostics. We are learning to interpret tens of thousands of parameters recorded by the control system. This work involves neural modelling and artificial intelligence. When data is analysed correctly and in time, it allows us to predict failures before they happen. Once these procedures are fully proven on new rolling stock, we will be able to scale them across the existing fleet as well.
Does the added complexity that comes with redundancy and duplication inevitably drive up production and service costs?
Not quite. Duplication does add cost—additional assemblies and components inevitably make a design more expensive. Redundancy also requires investment: extra cabling, additional switching equipment. But when you step back and look at the locomotive as a complete system, these costs are marginal compared with the efficiency they deliver. We never introduce redundancy blindly. Every such decision is backed by a detailed economic assessment. And time after time, the outcome is the same: the gains in reliability, availability and operational stability outweigh the additional costs of production and maintenance. In the long run, redundancy pays for itself.
How deeply has artificial intelligence entered your work? What do you see as the risks and the benefits?
At this stage, artificial intelligence remains a tool—a powerful one, but still just a tool. Its role today is limited to processing large volumes of data and identifying anomalies. In the foreseeable future, it will not replace engineers.
We already see that AI makes mistakes, which is why it has to be used with great care. Like everyone else, we are preparing for wider adoption as the technology matures, but we are very clear about where the boundaries lie.
For us, rolling stock is not about attractive images on a magazine cover or impressive displays at trade shows. It is the backbone of the national economy. Russian Railways operates a transport system that stretches across a vast country. Ultimately, this is also about human lives. I want people who board our trains at point A to arrive safely at point B — on time, in good health and satisfied with the journey. The price of error here is exceptionally high. At its extreme, it is measured in lost lives and economic disruption. That is why I am not prepared to hand over critical decisions to artificial intelligence. For now, we see AI as appropriate in very specific areas: supporting routine business processes, assisting with documentation and analysing large datasets.
When it comes to digital engineering, however, the real breakthrough for us lies elsewhere—in the rapid adoption of 1D modelling.
Today, engineering design is almost entirely three-dimensional. What makes 1D modelling different, and why do you describe it as a revolution?
The strength of 1D modelling lies in its ability to predict product performance with much greater accuracy at the very earliest stages of a project. In effect, it gives us a digital twin long before any physical design takes shape. We can already see how effective this approach is through a pilot project currently under way. On the basis of those results, we have decided to roll out 1D modelling across TMH Engineering in full by the end of this year.
“Once a simplified digital twin is in place at the initial stages of development, we gain far more precise control over product characteristics—including reliability. This approach significantly reduces the risk of early-stage errors and cuts down the number of intermediate engineering iterations, particularly when working with suppliers. In practical terms, it means that even before a single line is drawn on a blueprint, 1D modelling allows us to formulate precise technical requirements for components.
Designing new diesel locomotives at TMH Engineering. Source: TMH
At later stages, when a fully developed digital twin is available, we can clearly see how changes in components or operating conditions affect performance—including reliability— and carry out what is essentially a degradation analysis. In other words, we gain the ability to manage the entire life cycle of the product.
There is one more important dimension to this. 1D modelling makes it possible to automatically generate software code. This dramatically simplifies development, reduces time and lowers the overall effort required to create software. From a productivity standpoint, this is a major step forward.
This is a genuine breakthrough. Our pilot programme for implementing 1D modelling will be completed by the end of December. From January 2026, it will become part of our standard operating practice across every project. For us, this marks the beginning of a new chapter.
Where should the emphasis be placed today: on unique solutions or on scalable technologies?
In reality, engineering is always about trade-offs. There is no such thing as a perfect design: engineers are constantly weighing cost, manufacturability, service life and reliability against one another. This balancing act never stops—and different specialists often arrive at very different conclusions.
When it comes to managing reliability, costs and the full life cycle of equipment, modular and platform-based approaches have proven their value. This is the direction the global industry has taken, and we are very much part of that trend. However, standardisation does not rule out individuality. Well-conceived, stand-alone solutions continue to have strong market potential. Our new TE26 diesel locomotive is a clear example of how a truly unique product can successfully complement a platform-based portfolio.
Interviewer: Konstantin Dorokhin
Based on Progress in Unstoppable of Vektor TMH No. 4 (63) 2025
