From DSI to FILS

I did not directly witness the beginnings of foreign-language engineering education at the Bucharest Polytechnic. By 1990 I was already in Italy, and what I know about that period comes from documents I consulted later and from the accounts of colleagues who had taken part in building the new structure.

In the summer of 1990, the Department of Engineering Sciences, known by the initials DSI, was established with teaching in English. The French and German tracks were added in the spring of the following year. It was one of the ambitious projects that emerged immediately after the collapse of the communist regime: Romanian higher education was trying to reopen towards the West, attract foreign students, and build programmes compatible with those of European universities.

The official initiative was associated with Rector Virgiliu Constantinescu, while the practical organisation relied to a large extent on professors from the Faculty of Electrical Engineering, including Andrei Țugulea, Paul Dan Cristea, and Marius Preda. From the outset, DSI had been conceived as an autonomous structure, similar to a faculty and with its own leadership. It did not, however, have a sufficiently large teaching staff and relied on academics from Politehnica's traditional faculties and from other universities.

TEMPUS programmes and other forms of international cooperation made possible training periods abroad, the preparation of new courses, and the establishment of links with Western universities. The three language tracks followed different paths: the German track benefited from substantial institutional support from Germany, the French track from the support of the Francophone community, while the English track relied more heavily on Politehnica's own efforts and on individual university partnerships.

In 1991, the former Electrical Engineering programme was renamed Electrical and Computer Engineering, a sign that computer science was beginning to acquire a place of its own in the new structure. The programmes in Applied Electronics and Information Engineering would later emerge from this direction.

The DSI arrangement had been conceived as temporary, yet it lasted for more than a decade. In October 2001, the Senate of Politehnica University approved its transformation into the Faculty of Engineering in Foreign Languages, and from 2002 FILS began to operate officially as a faculty. In essence, this legalised a structure that had functioned almost as a faculty since 1990.

Faculty status did not, however, solve the problem of teaching staff. Many of the professors who had taught in the early years had their main duties in their faculties of origin. As funding and the initial enthusiasm diminished, some of them withdrew. FILS needed an academic structure of its own, capable of ensuring continuity. This led to the creation of the Department of Engineering in Foreign Languages, initially a small nucleus made up mainly of young academics.

Material resources: from makeshift arrangements to premises of its own

In the beginning, the material resources available to foreign-language engineering education were very modest. In the old building, the spaces used for practical work were unsuitable, and in some cases the available computing equipment amounted to a single computer. Students carried out many laboratory activities in the facilities of the faculties from which the lecturers in the respective subjects came. Such dependence was natural for a new structure, but it limited its autonomy and made it difficult to organise a coherent programme.

When I joined FILS in 2004, the process of building an adequate material base had only just begun. In the following years, the development of the study programmes, the growth of the faculty's own teaching staff, and the accreditation procedures required further laboratory equipment and more careful use of the available space. My own contribution concerned mainly the curricula and the formation of the teaching staff; the material base on which they relied had been, and continued to be, the result of a collective effort.

The situation gradually improved. Computers were ordered, and the faculty was allocated two rooms in the basement of Building J, beneath the faculty office, in the space of a former storeroom. They could be converted into the faculty's own laboratories, equipped with networked computers. This was not the achievement of a single person, but the result of the efforts of the faculty leadership, professors, and technical staff. Among the colleagues directly involved in organising the laboratories and planning their use was Associate Professor George Drăgoi, whose contribution must be acknowledged in this history.

The decisive change came when the entire CJ building in the New Campus complex (Fig. 1) was made available to foreign-language engineering education. It had previously been assigned to a faculty in the mechanical field, but had not been used by it. FILS thus acquired a space that could be clearly identified as its own, with lecture rooms, laboratories, and administrative offices brought together in the same building. For a faculty that had begun by borrowing teachers, laboratories, and rooms from other structures within Politehnica, this development meant more than a simple logistical improvement: it was one of the conditions of its institutional maturation.

In 2004, the history of FILS also became my history

I was appointed to FILS on 15 September 2004 as a professor in the Department of Engineering in Foreign Languages. As far as I remember, I was the department's only full professor; the other colleagues were teaching assistants or lecturers.

At that time I was still living and working in Italy. My arrangement had been established with Paul Dan Cristea, coordinator of the English track, and with the dean of the faculty at the time, Professor Adrian Pascu: I had to fulfil all my teaching obligations, down to the last hour, but I could do so in a modular system. In other words, I came from Italy for periods of two, three, or four weeks and concentrated the hours of all my courses within those intervals.

I was not the only person who taught in this way. Other professors also came for concentrated periods, but they were paid by the hour. I was a tenured employee of Politehnica. This difference would later become important when I took over the leadership of the department.

Between Italy and Bucharest

I came to Bucharest at least four times a year. I had to teach, supervise projects, and return for examinations. During my stays in Romania, I lived in my mother's house in the Militari district, which I had always kept.

During those years, my mother became ill and I had to place her in a home for elderly people. Whenever I was in Bucharest, I visited her every day. My days were divided between the faculty and those visits. My mother died in February 2010. In the following years, when I returned for the winter examination session, my stays in Bucharest remained linked to her memory as well: I arranged the traditional memorial services for her.

During the weeks I spent in Romania, I taught between four and eight hours a day. The modular system did not mean fewer obligations, but their concentration into a very short period. My teaching work, however, was not limited to the periods spent in Bucharest. In Italy, I devoted a great deal of time to preparing the courses, which I changed every year. I never believed that a professor could repeat the same material indefinitely, least of all in a field in which technologies and practices change rapidly.

The students' command of English was not uniform either. Although the courses were officially taught in English, not all students could follow a more complex technical explanation with equal ease. When I felt that the language was becoming an obstacle to understanding the subject, I continued in Romanian. The students did not object; I had the impression that they were pleased to be able to follow the explanations more easily. My purpose was not to prove that I could teach in English, but to help them understand what I was teaching.

There were also differences between the FILS students and those I had taught many years earlier at the Faculty of Automatic Control and Computers. I do not think these differences resulted only from the faculty's profile; they also reflected a generational change and a different way in which students related to the university. The gap between my teaching objectives and the learning goals of the new generations widened over the years and, after almost twenty years, contributed to my decision to stop teaching.

A “factory” of software engineers

In the early 2000s, Romania had a rapidly expanding software industry made up of private companies. It needed not only programmers capable of writing code, but engineers who understood the entire process of building a system: requirements, modelling, design, architecture, team organisation, testing, quality, and product maintenance.

Since I was in a position to shape the profile of future graduates from the computer-related programmes at FILS, and did not want their curriculum simply to copy those of the Faculty of Automatic Control and Computers, I chose to align it with the Information Engineering programme of the Faculty of Electronics, which at that time was expanding towards computer science. There I had greater freedom in constructing the curriculum. I preferred the name Information Engineering. I knew that James Martin had used the expression in the 1980s and 1990s for a methodology that claimed to cover almost every aspect of organising and processing information. I considered that an excessive extension of the term. For us, however, the name could designate a modern field, broader than programming alone and flexible enough to include new directions in computer science.

I imagined the programme as a genuine “factory” of software engineers. I used the word “factory” not to suggest uniform production—the sense in which Professor Moisil had once used the metaphor— but to emphasise that the faculty had to educate, systematically, the specialists needed by the economy.

With the agreement of the faculty leadership, alongside the traditional programming courses I gradually introduced subjects with names that were unusual in the conventional computer science curricula of Romanian universities: Software Development Methods, Software Methodologies, Software Design Techniques, and Software Architectures. Together they were intended to form a coherent path: from programming to methods, from methods to modelling and design, and then to architecture and control of the development process.

As early as 2004, I decided that the first programming language taught to FILS students should be Java, not C or C++. Those languages did not disappear from the curriculum and could be studied later, but I considered Java better suited to introducing object-oriented programming on a conceptual basis and to preparing future software engineers. FILS students did not object to this choice, which took them away from the path followed at Automatic Control and Computers, with one exception: Alexandru R., a first-year student and an enthusiast of C programming, asked me in the very first lecture, “Why not C?” Today Alexandru is an associate professor; he has not renounced his passion for the C family and now preaches Rust.

In the later years, all courses included projects. I did not want students to be left only with definitions and diagrams memorised for the examination. They had to build things, choose between solutions, justify an architecture, and understand the difference between a program that works and a software product developed according to engineering principles.

Software engineering was the core of the programme I created, but I did not want it to become a narrow specialisation. Students had to remain open to other areas of computer science, especially artificial intelligence, distributed systems, and medical applications.

Building a teaching staff of our own

At first, one of the serious problems was organising the practical activities associated with my courses. I concentrated my teaching in the periods when I came from Italy and did not have enough time to conduct all the laboratories as well.

I asked Crenguța Bogdan, my former teaching assistant at Ovidius University, to help me. She agreed and, in an effort I have never forgotten, commuted between Constanța and Bucharest to conduct the practical sessions for my courses. This continued until 2008, when Andrei Vasilățean completed his studies. I then accepted him as a doctoral student, and he took over the practical activities for my courses.

The subject of Andrei's doctoral thesis was inspired by the research I was carrying out in Italy, especially within the LuMiR project. For him, however, I modified the initial direction by introducing intelligent agents and the concept of an eHealth ecosystem. This was also one of the ways in which I tried to build a teaching staff of our own: good graduates could continue to the master's and doctoral levels, participate as assistants in practical activities and, gradually, become members of the faculty's academic staff.

I tried to bring the best graduates into the department. It was not easy. The software industry offered much higher incomes than the university, and talented young people had little reason to choose an academic career if salary was their first criterion. Those whose vocation for research and teaching was strong enough to resist the attraction of industrial earnings were few.

Sometimes those who turned towards teaching were people who had not found a suitable position in industry, or who wanted a more flexible schedule for other activities. It was difficult to explain why such a graduate was rejected. Yet responsibility for the shortage of academics with a genuine vocation did not belong only to them. We, the professors, did not always succeed in identifying the best students in time and, above all, in persuading them—through our own example—of the appeal of being teachers and researchers.

During the period when Andrei Vasilățean was becoming a teaching assistant, a combination of circumstances gave me the opportunity to meet Maria-Iuliana Dascălu, a graduate of the Academy of Economic Studies, a serious young woman with good knowledge of computer science. I was pleased to be able to approve both their applications for academic competitions. Andrei came directly from among the students educated at FILS and was continuing to distinguish himself through his doctoral work; Maria-Iuliana followed a different university path, but brought the same seriousness and desire for knowledge.

Later, both would become members of the faculty leadership. I regard them as two of the people to whom I passed the standard. The fact that they remained in the university, grew professionally, and continue the construction begun in those years gives me a real sense of continuity.

Over the course of my activity as a doctoral supervisor, three doctoral candidates came from Ovidius University in Constanța, three from FILS, and another three had followed other university paths. The fact that FILS had begun to produce its own doctoral candidates was a sign of academic maturity.

The first Software Engineering master's programme at Politehnica

The project could not stop at bachelor's studies. The subjects introduced into the Information Engineering programme gave students an engineering foundation, but software engineering was too broad a field to be studied in depth only within the first cycle of university education.

In May 2007 I was still working on different versions of the programme and curriculum for a master's degree in Software Engineering. I can no longer state with certainty whether we managed to organise admissions as early as the autumn of that year. I do know, however, that in September 2008 we held the entrance examination. The master's programme was therefore operating in the 2008–2009 academic year.

In May 2009 I organised a public presentation of the programme. The poster announced it as the first Master of Software Engineering at Politehnica University of Bucharest. The wording did not indicate the year in which it had begun, but the fact that it was the first programme of its kind created within the university.

In 2010, the programme was presented and evaluated as a research master's degree, taught entirely in English, lasting four semesters and totalling 120 credits. We did not want merely to offer our graduates a few additional courses. The master's programme had to educate specialists capable of understanding the entire chain of software product development, from the emergence of requirements to delivery, bringing together the technical perspective, architecture, project management, testing, and quality assurance.

The core of the programme included subjects such as Principles of Software Engineering, Formal Models in Software Engineering, Software Methodologies, Software Project Management, Software Architectures, Software Testing, and Software Quality Management. Around them were courses oriented towards fields that were then developing rapidly: model-driven engineering, distributed systems, human–computer interaction, the Semantic Web, software agents, games, and interactive simulations.

The programme was organised around three parallel directions: fundamental software engineering subjects, application-oriented subjects, and supervised research. Almost half of the student's work was reserved for research and preparation of the dissertation. The dissertation had to contain a genuine research element and, whenever possible, lead to the design and implementation of a software system.

The master's programme aimed to educate not merely better programmers, but future leaders of software development: architects, designers, technical leaders, and specialists capable of improving existing practices in industry.

By then I had also been a doctoral supervisor for approximately ten years. The best students could therefore follow a complete path: the bachelor's programme provided the basic engineering education, the master's degree brought advanced study and applied research, and the doctorate opened the way to advanced research. The “factory” of software engineers I had imagined in 2004 had thus acquired all the necessary levels.

Accreditations

During the period in which I headed the department, the FILS programmes had to undergo evaluation by the new ARACIS structures. The files were extensive and required an enormous amount of administrative work: curricula, course specifications, staffing data, premises, laboratories, scientific activity, and justification of the competences offered to graduates.

In preparing the documents, I received essential help from Bujor Păvăloiu and Ioana Guică. Without their careful work, organising the files would have been much more difficult.

For the English-language Information Engineering programme, I was directly involved in defending the curricular concept before the committee. Its members knew me professionally, but were surprised by some course titles that did not exist in the traditional nomenclatures. I had to explain why a modern programme needed subjects such as Software Methodologies, Software Design Techniques, and Software Architectures, and how they formed a coherent path for educating a software engineer.

The initial puzzlement gradually turned into interest. The committee understood that this was not an arbitrary invention of titles, but the introduction at Politehnica of a mature academic field, internationally recognised and required by industry.

I did not personally lead every accreditation. For the German-language Applied Electronics programme, the main role was played by Professor Istvan Sztojanov, a long-standing representative of the Faculty of Electronics and a supporter of that direction. As head of the department, I was responsible for the structure as a whole, but each programme had people who best understood its particular features.

The accreditation of the Information Engineering programme and that of the Software Engineering master's degree meant more to me than the fulfilment of formal requirements. They were the institutional confirmation of a curricular project that I had built gradually.

Students and industry

Many FILS students began working for companies before graduation. This gave them an opportunity to gain experience and earn an income, but it created serious problems for the faculty. They had to reconcile their work schedule with attendance at courses, projects, and examinations. Sometimes employment became the priority, while university was treated as an activity to be fitted in among professional obligations.

Dissertation topics rarely came directly from their work in companies. In firms, students were used mainly for programming tasks. I tried to teach them something else: to think at the level of requirements, development methods, models, architectures, processes, and software product quality.

This difference did not mean that professional experience was useless. On the contrary, many remained with the companies where they had begun working and obtained better positions after completing their studies. FILS graduates found jobs quickly, and many went abroad to work. Their careers confirmed that the chosen orientation answered a real need of the software industry.

Heading the entire department

In 2008, when the leadership of the faculty changed, a new head also had to be appointed for the Department of Engineering in Foreign Languages. I was elected to take on this responsibility—as far as I remember, without another candidate—and began the work immediately.

The position was not limited to the English-language Information Engineering programme, to which I had devoted most of my teaching activity. I was head of the entire department, which managed its own teaching activities for all three language tracks of the faculty: English, French, and German.

Each track had a coordinator who knew its professors, current problems, and institutional relationships. This organisation spared me from having to enter into every detail of the French and German tracks. I was, however, directly responsible for the English track, both with regard to its teaching staff and to its curricula and programme development.

I wanted the English and French tracks not to evolve as two unrelated structures. In my view, the two Information Engineering programmes had to develop together and have curricula that were as similar as possible. For a time, I succeeded in maintaining that correspondence. I also taught in the French track precisely in order to support the close relationship between the two programmes.

In time I realised that, given my circumstances, it was too difficult to work simultaneously in both tracks. I lived in Italy, concentrated my teaching into short periods, and also had responsibility for leading the department. I therefore gave up the courses in the French track and concentrated my teaching on the English one. Later, for reasons related to the available staff and to the French track's own orientations, its curriculum gradually diverged from the English one.

My position was therefore a dual one. As a professor and programme designer, I was involved mainly in the English track. As head of the department, I had to oversee the functioning of the entire structure, with all three tracks, coordinate positions, subjects, and the department's own staff, and maintain relations with professors coming from the other faculties.

An important part of this activity, scarcely visible outside the university, was preparing the annual staffing and teaching-load plans. I had to allocate subjects, lectures, and practical activities among the department's own academics, professors from other faculties, and hourly paid collaborators; ensure that all teaching obligations were covered; and propose the positions needed for the department's development. In a structure that was still small and dependent on external collaborators, preparing the staffing plan was, every year, a difficult exercise in balancing the requirements of the programmes, the available competences, and the existing resources.

The position also involved participating in staff evaluations, preparing competitions for academic posts, drawing up self-evaluation and quality-assurance documents, and coordinating the department's research activity. Programme accreditations were the most visible part of these responsibilities, but behind them lay continuous administrative work carried out together with the track coordinators and the faculty leadership.

The situation was all the more unusual because I continued to live and work in Italy. I relied on the track coordinators and on colleagues who were permanently in Bucharest, especially Bujor Păvăloiu, to solve operational problems. The general responsibility for the department, its development, and its relationship with the faculty leadership nevertheless remained mine.

The department still had few academics of its own and depended on many professors from other faculties or paid by the hour. It was so small that my leadership position did not carry the allowance normally granted to the head of a chair or department. The number of its own employees did not reach the required threshold.

As far as I remember, only around 2012 did the department reach the size at which its head would have been entitled to the allowance. Yet that was precisely when my term was ending and I was about to be retired. There was a certain irony in the situation: I had helped develop a structure that was too small for its leadership position to be remunerated, and when it finally became large enough, I had to leave it.

A project that had reached maturity

Looking back, almost everything had changed between 2004 and 2012. The department I had found so small had grown, and a more stable nucleus of academics had formed around it. The Information Engineering programme had acquired an identity of its own, distinct from a simple reproduction of the curricula of other faculties.

Software engineering had become the programme's principal axis without closing it off from other areas of computer science. There was a master's programme in Software Engineering, the programmes had passed the necessary accreditation procedures, and the course names and curricular conception that had initially caused some puzzlement had gained institutional recognition.

The faculty now also had its own rooms and laboratories, incomparably better than the improvised material base of the early years. Academic, administrative, and material development had supported one another and given FILS a clearer institutional identity.

There were increasingly fewer professors from outside the faculty teaching specialised courses, although collaboration with hourly paid academics remained important. The most convincing result, however, could be seen in the graduates: they found their place in the software industry, in Romania and abroad, and confirmed that the chosen orientation met a real need.

The structure I had led had become mature enough to continue on its way without me.

A bitter retirement

In January 2011, the new Education Law came into force. One of its provisions required university academics to retire at the age of sixty-five. For older professors, the measure came like a blow from a club. An entire generation that had accumulated decades of teaching and scientific experience had to leave their tenured positions almost simultaneously.

I did not dispute the need to promote younger academics. It seemed risky to me, however, that renewal should be achieved by abruptly removing the most experienced professors. I had already passed the age stipulated by law. In my case, retirement was postponed until the end of my leadership term and the university elections of March 2012.

Shortly beforehand, I was called in by one of the vice-rectors. He spoke to me in a friendly, conciliatory tone and asked me to accept retirement without protesting. The university, he explained, needed professorial positions to be vacated for the many young associate professors who were ready for promotion.

— No problem, Professor. I shall apply for retirement, I replied.

I later learned that this large group of associate professors waiting to occupy the vacated posts did not exist, at least not in the proportions that had been presented to me.

In March 2012, while many professors in the same position were trying to obtain the university's support in order to continue their work, I spoke with the rector of Politehnica. I had known her for many years, since she was a young teaching assistant at the Faculty of Industrial Chemistry, and I trusted that my situation would be considered sympathetically.

— Do not worry, Professor. We shall request a sabbatical year for you and several other colleagues, and then we shall see, she told me.

I left reassured, convinced that she would not forget me. Later, I was told what had happened after the Senate meeting. The rector's secretary, who knew me because she had been my student, noticed that my name was not on the list of those for whom a solution had been found.

— And what are you doing about Professor Șerbănați? she reportedly asked.

The reply was brief:

— Take his retirement file to Human Resources.

And so I was retired.

I say frankly that I would have preferred to continue my activity at FILS as a tenured professor. I had also retired in Italy and, for the first time in many years, would have had more time for the faculty, my courses, and the students. I did not, however, turn retirement into a drama. I knew that every professional stage comes to an end and I was satisfied with what I had managed to build.

What disturbed me was not retirement itself, but the way two people in the university leadership, from whom I would have expected honesty and support, treated me with promises and explanations that were not to be honoured. The irony was that, only a few months after my retirement, the regulations were amended and universities were once again allowed, under certain conditions, to retain professors who had passed retirement age. For me, the change came too late.

After retirement

My connection with FILS did not end in March 2012. I continued to teach on an hourly paid basis until 2023. For approximately ten years, I too belonged to the category of academics outside the permanent structure whom the faculty continued to need.

I was no longer a tenured professor and no longer held a leadership position, but the courses and the students remained part of my life. The fact that the faculty continued to grow, that it now has more academics of its own, and that people educated or recruited in those years continue the work gives me the feeling that the effort was not in vain.

I left the leadership and then the body of tenured professors, but part of what I conceived and built remained in the faculty's structure, its programmes, and, above all, its graduates.

Software Engineering, an educational profile for the future

A graft that can become rootstock

Looking back, the Software Engineering profile I tried to build at FILS may be compared to a graft. It did not emerge on ground without an engineering tradition; it was grafted onto systems engineering, from which it inherited—in the spirit of the analysis and design methods promoted by authors such as Ed Yourdon and Robert C. Martin—the rigour of modelling, a view of the whole, and concern for integrating components into a coherent system. The compatibility between the two fields proved very good. Systems engineering supplied the roots and resilience, while software engineering brought a different kind of fruit: development methods, architectures, processes, testing, quality, and the organisation of collective work.

Software engineering was not, however, a weak field incapable of existing on its own. What it needed was rather a recognised institutional and curricular support on which it could develop within a technical university. Once mature, its position can change. I believe that today software engineering can itself become the rootstock onto which artificial intelligence, data- and model-based systems, cloud technologies, security, and new generations of distributed systems are grafted.

Artificial intelligence can generate code and dramatically accelerate the creation of applications, but it cannot replace the need for well-formulated requirements, coherent architectures, verification, testing, security, rigorous validation, and accountability throughout the product life cycle. The more powerful the tools become, the more important the engineering discipline within which they are integrated becomes. This was one of the central ideas of my 1992 book, and it remains a current requirement.

A profile that must evolve

The Software Engineering profile should therefore remain in university curricula, but it cannot remain unchanged. In addition to the foundations of programming, modelling, and design, it should include AI-assisted development, engineering of data- and model-based systems, cloud and DevOps technologies, security, distributed systems, verification of automatically generated results, and collaboration between humans and intelligent agents. The software engineer of the future may write less code by hand, but will have to understand more deeply the system being built and its consequences.

As for applicants, I do not believe that the name Software Engineering alone will produce a massive influx. Artificial intelligence, cybersecurity, and data science currently have greater drawing power. A well-explained software engineering programme, taught in English, connected to industry, and built around real projects can nevertheless attract a steady flow of good candidates. Its advantage is that it does not depend on a single fashionable technology, but on an enduring need: transforming ideas and prototypes into reliable and durable software systems.

In the past, software engineering was successfully grafted onto systems engineering. In the future, it can become the rootstock onto which new technologies are grafted. From this perspective, the project begun at FILS in 2004 can continue—with different tools and different subjects—as long as the education of programmers is not confused with the education of software engineers.