Your doctorate research focused on the development of data-​driven statistical models for the analysis and optimization of the maintenance decision-​making process in railway track and wheelset maintenance. What is your research focus?

Being an Industrial Engineer with solid industry background (in the areas of logistics, supply chain and maintenance), I was able to transfer a lot of the acquired knowledge to my academic research in the transportation area. Transport is an interdisciplinary field, and researchers are required not only to have different backgrounds and skills but also to be exposed to problems in different stages of urbanization, infrastructure, regulation and budget so that the proposed solutions and enhancements can make a real impact. In this sense, I like to think of my research as having the focus of proposing solutions to transportation problems, whether those come from railway maintenance, as most of my doctoral research, or public transport operations and planning, which is a topic I have been working more closely. My goal is to become a specialist in quantitative methods for transport problems within the fields of statistics, machine learning and operations research, as tackling real-world problems usually involves a combination of multiple methods, where collecting, understanding and making proper use of the data is mandatory for delivering scalable, practical and cost-effective solutions.

What developments do you see in the specialist area of maintenance?

As in many other fields, maintenance is developing as a result of the introduction of new technologies ranging from the design of new products and materials to monitoring methods and equipment. In the past decades, maintenance has experienced a shift of paradigm, leaving its original role of a mere repair activity to a strategic role in the organization, directly affecting the operational and financial results. Companies in the maintenance paradigm of the past may still see it as “too much time spent on fixing things”, but companies that recognize the capital importance of maintenance, especially in our globalized world, will likely thrive. As a sector that directly impacts quality and productivity, maintenance must act proactively. Therefore, many developments in the area are related not only to new technologies, but also fundamentally to processes and methods, as well illustrated by predictive and condition-monitoring methods, and other maintenance-improvement programs such as Reliability-Centered Maintenance (RCM) and Total Quality Management (TQM). The increasing complexity and diversity of physical assets within an organization increases the demand for better models and tools. Associated with unprecedented levels of data, this creates a perfect environment for specialists to develop intelligent and cost-effective solutions to make maintenance a strategic sector.

What can science do to advance the rail industry?

A few weeks ago, I watched a video from a YouTube channel (Numberphile) that is famous for explaining some cool math problems. The video was on ‘how to make railway timetables with graphics’, and it gave an interesting perspective on how the control of the train operations was made before computers, and how science dramatically improved the safety and efficiency of this process. I guess most travellers do not realize how much science has influenced railways over the years, but I would say that it had (and has) a giant impact on the rail industry. Imagine a sector that has vital importance to society, is characterized by complex, expensive and long life-cycle assets, operates 24/7 and has to comply with a series of regulations: the rail sector obviously urges for scientific developments. In my perspective, as both a researcher and a former railway maintenance engineer, science is evolving at a much quicker pace than the one that can be absorbed by the rail industry. Ranging from complex scheduling problems to the use of signal processing for degradation modelling, machine learning and AI techniques for defect monitoring, just to cite some examples, a lot is being proposed by many enthusiastic and diligent researchers worldwide, some of them employed directly by the railways. However, the inherent complexity and magnitude of the sector, inserted in a highly regulated environment, imposes many obstacles and some hard restrictions to the “freely” implementation of many of these developments. On top of that, many methods that typically privilege mathematical elegance are not practical, with some model assumptions that do not hold true in the actual environment, and many of them ignore uncertainty dynamics, which is a key element of railway operations. With all that being said, my opinion is that science is on the right path in terms of pursuing new methods and solutions, but it is crucial to keep strengthening the relationship with the industrial partners, through cooperation and mutual support, so that the chances of successful implementations are greatly enhanced.

Where is the railway industry heading in the next few years/ decades?

My opinion is that railways are moving towards becoming the most sustainable, innovative and safest mode of transport in the world, all that thanks to the massive investments in technology, digitalization and automation. It was not too long ago that we witnessed the economic gains from moving from paper archives to completely digitalized systems, or production gains from manual railway sleeper replacements to automated railway track laying machine replacements. Although some may not think of these as disruptive technologies, they surely had a great impact on the sector. Today, we are only scratching the surface of where we will be a few years or decades from now. I believe the increasing demand for railway transportation, both passenger and freight, will fuel the need to accelerate communication technology and improve railways efficiency through automation. The blend of Internet of Things (IoT) and Artificial Intelligence (AI) will allow unprecedented levels of data to be recorded, processed and analysed, leveraging the automation of processes and tasks, improving the quality of services and reducing the need for human intervention, with better risk management and, consequently, better decisions.

With only 22% of women, gender is not balanced in the transport sector - according to the European Commission for Mobility and Transport. How do you see the status of women in science?

Throughout history, discoveries and innovations of pivotal importance were made with the irrefutable mark of women. However, their names are still under-celebrated and what we witness is a high asymmetry of opportunities between genders. This asymmetry is even more pronounced in STEM (science, technology, engineering and mathematics) and related fields, although this scenario appears to be slowly changing. For example, according to the U.S. Bureau of Labor Statistics, in 2019 women were 48% of the U.S. workforce and, although they were still underrepresented among the STEM workers, they went from being around 8% in 1970 to 27% in 2019. On a global level, the reality is not more encouraging: the most recent survey from UNESCO Institute for Statistics (UIS) highlights the gap in science, revealing that less than 30% of the world’s researchers are women. Apart from merely acknowledging these numbers, it is crucial to understand the qualitative factors behind these discrepancies, which include, but are not limited to, financial considerations, family decisions, discriminations and cultural aspects. The conclusion is that, as many recent studies suggest, biological differences or psychological attributes cannot fully explain the gaps observed in numbers and salaries. Fundamentally, social and environmental factors have played a pivotal role in determining the current gender inequity. I believe women have come a long way over the last few decades, and although there is vast evidence pointing out that we perform as well – if not better – than our male counterparts in various roles, it still seems that we need to prove our competency, resilience and proactiveness to the rest of the world. When thinking about women in science, and especially in male-dominated fields like STEM, my perception is that women are technically as prepared as men (if not more), however, cultural biases against women and stereotypes are still present, even if they are unconscious. Strange as it may sound, sometimes these biases come in the form of the impostor syndrome, or the researcher doubting her own capabilities for assuming a given role. For the scientific community, it is time to create opportunities and engage girls and women in science by promoting social support, especially in communities where STEM careers are contrary to cultural expectations.

What advice do you have for women and girls who want to pursue a career in science - especially in a male-dominated field like railways?

The first, and possibly best, piece of advice I could give is saying “just go for it”! I am sure a lot of us overthink about “what is it going to be like if I am the only woman there?”, but, at the end of the day, it is not only about gender equality, it is about occupying a deserved position based on skills and competences. Always remember that you are there because you are qualified to be there, do not ever underestimate yourself. This may seem like an oversimplified way to deal with the many possible obstacles you may encounter, but having a strong mindset around your capabilities will prevent you from suffering from a lot of common issues women face at work, especially when the environment is hostile. As sad as it may sound, sometimes even basic infrastructure issues, like not enough female sanitary installations, is a reality, and we need to conquer that space so the reality can change for others in the future.

Some other important advice: talk to your peers, know what they are working on and support them whenever you can, this is definitely the most effective way of engaging other women because it does not only create a supportive environment (which we all need) but also corroborates with inclusion and participation of others. 

Finally, yet importantly, know that science is an amazing world, do not be afraid of joining it!


Über Mariana de Almeida Costa
[ETH Zürich über Mariana de Almeida Costa] Sie ist eine ehemalige Wirtschaftsingenieurin mit einem soliden industriellen Hintergrund, die direkt in den Bereichen Beratung, Lieferkette und Logistik tätig war. In den Jahren 2015-2017 erwarb sie einen Master of Science in Operations Research am Georgia Institute of Technology (Atlanta, USA), an dem sie 2020 auch ihren Doktortitel in Wirtschaftsingenieurwesen ("Statistics track") erwarb. In ihrer Doktorarbeit befasste sie sich mit der Entwicklung datengesteuerter statistischer Modelle für die Analyse und Optimierung des Entscheidungsprozesses bei der Instandhaltung von Gleisen und Radsätzen, wobei sie die Untersuchung von Kompromissen zwischen Lösungen mit minimalen Kosten und der einfachen Umsetzung der empfohlenen Strategie miteinander kombinierte (vgl. ETH Zürich).