A conversation with Nikolay Bedzhev: Leading International Expert in Maritime Simulation Technology

Captain / Doctor Nikolay Bedzhev is a distinguished maritime professional with a career spanning navigation, engineering, and simulation technology. A Master Mariner and an active Marine Pilot at the Port of Varna, Bulgaria, Dr. Bedzhev holds a Ph.D. in Ship’s Theory, along with two master’s degrees in Electrical Engineering and Marine Navigation. As a lecturer at the Nikola Vaptsarov Naval Academy, Captain Bedzhev has been shaping the next generation of maritime professionals since 2007, sharing his expertise in ship dynamics, artificial intelligence, and computational modeling.

Beyond academia, Capt. Bedzhev is the founder of NauticBlue, a company pioneering the next-generation maritime simulators. Capt. Bedzhev’s innovations include the DNV-accredited Class S ViMarS Virtual Reality Simulator, a state-of-the-art VR training solution for the education and assessment of marine navigators. NauticBlue integrates artificial intelligence, real-time analytics, and proprietary physics-based modeling to enhance vessel simulation, reduce operational risk, and lower cost.

Dr. Bedzhev’s current post-graduate initiative, Simulation Engineering, bridges traditional maritime expertise with modern computational tools. By combining artificial intelligence, dynamic simulations, and computational fluid dynamics, Dr. Bedzhev is redefining maritime training and engineering. With a deep passion for innovation, he continues to push the boundaries of maritime technology, ensuring Bulgaria remains at the forefront of maritime-related digital transformation.

JT: First of all, thank you for doing this interview. I’ve known about you for many years, and honored to finally meet you.

NB: Nice to meet you as well.

JT: Can you describe your journey from a youth to where you are now?

NB: I come from a long line of mariners. My grandfather was an engineer, and my father is both a captain and a pilot. Naturally, I followed in their footsteps. However, the situation in Bulgaria at the time was complicated. During the post-communist period — after the fall of the regime, everything changed. The economy collapsed, and the country faced immense challenges.

During this time, being a captain was one of the most respected and well-paid professions. It provided stability and a good quality of life at a time when Bulgaria’s economy was in ruins — factories shut down, businesses disappeared, and opportunities on land were scarce. As a result, many of the best and brightest chose careers at sea, particularly as captains and engineers, which became a well-known path in Bulgaria.

Initially, I didn’t plan to go to sea. Instead, I pursued a degree in electrical engineering, which resulted in a diploma. But I soon realized it wasn’t the right fit for me. Despite my degree, I felt unprepared — I lacked fluency in English, had little experience with computers, and felt that my education was incomplete. I decided to continue studying.

That decision transformed my life. I redirected my profession to that of a seafarer and gained valuable experience working as a watch officer on car carriers. K Line, a renowned Japanese company, played a key role in shaping my maritime experience. Over the years, I worked with various other companies, including Bulgarian firms and brokers.

Eventually, I reached a milestone in my sea career, becoming the captain of a mega yacht — 100 meters long managed by a crew of 36. It was a prestigious and unforgettable experience.

Despite my career at sea, my ultimate goal was to become a marine pilot. Since around 2011 or 2012, I have been a pilot in Varna. Initially, I tried balancing my work at sea with my pilot duties on land, but it proved too demanding. Around the same time in the late 2000s, I also embarked on an academic path, inspired by my mentor and main teacher. I started as an assistant professor, which opened a new world for me, eventually leading me to pursue a Ph.D. By 2016, I had completed my doctorate.

JT: What was your focus of study for your Ph.D., and what was your motivation for pursuing it?

NB: My Ph.D. research focused on ship maneuvering, and I decided to develop a simulation. Initially, it wasn’t a full simulation — it began with my expertise in maritime pilotage, an area where I felt confident. Step by step, I built it from the ground up. I started with differential equations, then moved on to modeling hull resistance. Next came propeller and rudder dynamics, followed by environmental factors like wind and currents. The result was a detailed simulation model, developed in Matlab. After completing my education, I discovered Python, and my perspective completely changed. I was amazed by how the entire scientific community was embracing Python — it was eye-opening. Although I still respect Matlab, I transitioned to Python and open-source science. It was an incredible shift, one that expanded my capabilities and approach to research.

NauticBlue’s ViMarS Virtual Reality Simulator.

JT: Your work has resulted in some significant milestones, such as your NauticBlue product ViMarS obtaining the first national certification by Det Norske Veritas (DNV). How did this recognition impact your career and further research?

NB: Developing a simulation system, particularly one that integrates real-time vessel maneuvering, is challenging. However, our team persisted and our current software has received some attention from high-profile figures, including the Bulgarian President during a visit to Varna Naval Academy. I am deeply appreciative of my team which included experts in programming, animation, and Unity development, all of whom contributed their deep expertise to the project. The overall experience deeply shaped my approach to integrating technology into maritime education and maritime operations. We still have work to do, but feel that we have made substantive progress in 1. merging the dynamic nature of simulation with current artificial intelligence tools, 2. making the simulators low cost and accessible, and 3. using the software to support education initiatives that focus on emerging technology.

JT: You have mentioned integrating different frameworks, software, and programming languages in with your current simulator. Could you speak more to this?

NB: We used PHP for our instructor panel, but transitioning to Django provided more flexibility and scalability. We integrated Python into Unity in order to allow us to leverage advanced computational tools for real-time simulations. One key challenge was migrating projects from Python to C# in Unity. The newer versions of Unity support Python scripts, but I chose to write everything in C# for Unity. While I do have an external engine in Python and the option to use the same logic in C#, I ultimately migrated everything to C#. Additionally, the process of designing seabed simulations using soundings and charts required special consideration: for example rendering a 3D mesh based on real-time soundings. Virtual reality also played a role in enhancing these projects, allowing for more immersive training environments. In the end, regardless of the technologies used, ensuring the accuracy and realism of simulations is paramount, particularly when integrating AI tools for maritime applications.

JT: How have data-driven approaches have impacted your simulation development process?

NB: Data is the foundation of effective simulations. We rely on real measurement data rather than artificial datasets to enhance the accuracy of our models. This allows us to make informed predictions and validate simulation outcomes against real-world scenarios. The ability to analyze and interpret data efficiently is a crucial skill in maritime technology. Over the years, I feel I have developed an intuition for evaluating simulations based on data trends, much like a pilot assessing vessel performance. This approach has been instrumental in refining our simulation models and improving their reliability. Moreover, our experience in machine learning has allowed us to develop new AI-based models that enhance maritime risk assessment and decision-making.

JT: How do you manage real-time data integration or use of disparate data for vessel simulations?

NB: Real-time integration remains challenging. Real-time work involves sensors, all of which require precise synchronization. Recently, we have been exploring the use of physics-informed neural networks to improve simulation accuracy. We have also developed demonstrations showcasing real-time integration, which has been useful for evaluating system performance. Balancing theoretical research with practical implementation is essential. This approach ensures that our models are not only theoretically sound but also applicable in real-world scenarios.

With maritime data generally, employing neural networks can be incredibly useful for solving differential equations and performing regression, making every dataset valuable. I’ve worked with extensive scientific sources and empirical data, resulting in unique outcomes. Today, with powerful computers, we’re not dealing with artificial data — it’s real measurement-based data, which provides a significant advantage. I’m highly data-driven, and after working with it for so long, I instinctively recognize when outcomes are accurate. My programmer often asks how I can tell just by glancing at the figures, but for me, it’s second nature.

For example, we can predict the behavior of a 50,000-tonne vessel from raw data, making the [NauticBlue] simulations highly precise. Maritime datasets offer endless possibilities. Data is like gold — if you have it, you have treasure. The key is not just collecting data but also reading and understanding it. I often say machine learning is 99% data science; the rest is just a few layers of frameworks like TensorFlow or PyTorch. Data science is evolving, especially in handling big data and massive datasets. That’s why computational power is so crucial.

But every problem is inherently complex — there’s no universal solution. The real challenge for the future lies in managing both data and computation effectively.

JT: Can you give me a practical example of your work with machine learning and how it has shaped your understanding of the maritime space?

NB: At the Naval Academy, I was involved in fleet and ports management, but I chose to explore machine learning and deep learning. With a colleague, we worked on predicting the Baltic Index — an enriching experience that combined SARIMAX (Seasonal Autoregressive Integrated Moving Average with eXogenous factors) with machine learning techniques. We also developed a multivariate model to enhance accuracy.

Next, we applied machine learning regression algorithms like YGBM and HGBoost from Scikit-learn to analyze the Baltic Index. Additionally, we trained a Long Short-Term Memory (LSTM) neural network, which played a central role in my master’s degree research. It was a significant journey, reinforcing my belief that this field holds immense opportunities for the future.

JT: What’s your opinion on current state of artificial intelligence?

It’s important to recognize that while AI may seem like magic, it is fundamentally a scientific discipline — governed by principles, not limitless speculation. Some fear AI could spiral out of control, but in reality, it represents a structured approach to problem-solving, a new way of programming. Traditionally, computing involved providing data and algorithms to generate answers. Now, we have data and outcomes, and our task is to uncover the underlying algorithms. This shift is reshaping computing and programming — not through magic, but through scientific advancements.

JT: What are your thoughts on the future of AI in maritime operations?

NB: Artificial intelligence has immense potential to enhance maritime operations, but it should be seen as an advisory tool rather than a replacement for human expertise. While AI can provide valuable insights for navigation and risk management, human oversight remains essential. Mariners possess deep domain knowledge that AI systems must be trained to complement rather than override. The industry needs robust benchmarking and evaluation frameworks to ensure that AI systems are safe, reliable, and effective. Collaboration between maritime professionals and the tech community is crucial to developing AI tools that enhance safety and efficiency without compromising operational integrity. Additionally, the role of AI in predictive maintenance and fuel optimization is an exciting area of development that could significantly improve operational efficiency.

JT: What’s you opinion of autonomous surface ships?

NB: Shipping won’t be fully autonomous, but vessels might be managed remotely. Watch keeping officers might work from an office onshore. This shift could make the profession more attractive. However, I don’t believe that artificial intelligence alone will revolutionize shipping. Environmental factors play a significant role, which are often unpredictable. Therefore, I don’t see a fully autonomous future without human oversight.

JT: Where do you see AI landing first within the maritime space?

NB: Life on board comes with many challenges, and one of the biggest is paperwork — there’s just too much of it. When I’m on board with my colleagues, we are constantly dealing with endless documents. It’s overwhelming. If these processes could be streamlined or reduced to just the essential parts, it would make a huge difference. A more efficient system for handling paperwork would improve daily operations and, ultimately, life at sea.

Working in shipping is already demanding. Reducing paperwork and finding ways to digitalize or automate these processes would be a major step forward. If the industry can facilitate this shift, it would greatly improve working conditions on board.

JT: How do you see passing on your expertise to the next generation of maritime leaders?

NB: Maritime training must adapt to the technological advancements shaping the industry. Traditional education, including subjects like calculus and linear algebra, remains vital for developing strong analytical skills. However, training programs should also incorporate emerging technologies like AI, machine learning, and simulation-based learning. The future of maritime training will likely involve more hands-on, data-driven approaches that bridge theoretical knowledge with practical applications. Enhancing accessibility to advanced simulation tools will be key in preparing the next generation of maritime professionals. In addition, integrating augmented reality for remote training could be a game-changer, allowing students to gain hands-on experience without being physically present on vessels.

Artificial intelligence is simply a new paradigm in programming — nothing more. It’s based on the same statistics and algebra that have existed for decades, now applied with modern computational methods. When people talk about AI as something mysterious, unpredictable, or out of control, it’s misleading. In reality, it’s just another evolution in scientific methods, a different approach to programming. There’s nothing magical or uncontrollable about it — just structured, data-driven problem-solving.

JT: Can you speak to the future of the maritime industry generally?

Looking ahead, I believe the future will be shaped by highly educated professionals — engineers, captains, and maritime specialists who are well-prepared for the evolving industry. Shipping is highly regulated, and while there are minimum standards, they should not be seen as the final goal. These standards are just a baseline, not a true measure of what is needed. The industry must continue pushing for higher standards and better working conditions.

Current advancements in shipping are remarkable. We now have technically advanced vessels, reliable equipment, and better infrastructure than ever before. Regulations, vessel design, onboard organization, and shore-based support systems — such as search and rescue and vessel traffic services — have all contributed to an improved industry. I’ve never seen such a high level of sophistication in maritime operations before.

Shipping is fascinating, and the changes happening today are shaping its future in ways we can only begin to understand. While I have some ideas about where things are headed, one thing is clear: the direction is positive. My team shares this optimism. AI, in particular, is a transformative tool that will be a significant advantage for the industry. We have discovered a new way of programming and problem-solving, and it holds incredible potential for the future of shipping.

JT: Thank you Dr. Bedzhev, it has been an honor.

NB: My pleasure.

A Master Mariner and active Marine Pilot in port of Varna, Bulgaria, Capt. Nikolay Bedzhev possesses a PhD Ship’s Theory; M.Sc. in Electrical supply and equipment; M.Sc. Marine Navigation and is also acting as a Lecturer in Ship’s Theory in Naval Academy Vaptsarov — Varna, Bulgaria. He has profound understanding and experience in working with Dynamic Simulations, Marine Simulations, Data Driven models, Artificial Intelligence, Statistic and probabilistic models. He is also a Python and C# developer. Capt. Bedzhev is an owner and founder of marine simulation software company “NauticBlue” and is the creator of a DNV accredited Class S “VIMARS Virtual Reality Simulator”.