Knowledge-Driven Ecosystem: A Model for Successful University–Industry Collaboration

“A closer dialogue is needed between the academic and industrial spheres – not just for financial reasons, but because we can achieve so much more together. That is why we established BME’s technology transfer company,” said Rector Hassan Charaf at the opening of the second season of the Business Breakfast series at the Budapest University of Technology and Economics (BME). “While we are grateful for any donations, that is not what we strive for; our focus is on mutually beneficial, win–win partnerships,” the Rector added in his welcome address.

In line with the new concept of the Business Breakfast series, the focus of this academic year’s first event was not on the activities of a specific BME faculty but on a current, cross-faculty theme: artificial intelligence and its applications. “There is a reason why we are moving away from faculty-specific topics. We aim to strengthen internal cohesion at the university and to foster greater cooperation among faculties with a truly interdisciplinary approach,” the Rector explained.

The first presentation was delivered by János Levendovszky, Vice-Rector for Research and Innovation, titled Artificial Intelligence and Digitalisation: RDI Activities at BME. One way for companies to increase revenue is through the adoption of innovative technologies, which is costly – hence the global trend of integrating universities into corporate innovation processes. 

In addition to scientific excellence, BME must offer a service-oriented portfolio that grants companies access to its knowledge base, research infrastructure and talented human resources. 

This is a prerequisite for building a successful innovation ecosystem, the Vice-Rector emphasised.

He went on to explain that the ongoing digital transformation involves the integration of digital capabilities into real-world production and economic processes, leading to a substantial increase in value. This is a knowledge-driven transformation, rooted in scientific competences, where basic research outcomes are combined with industrial challenges – this is the domain of so-called deep tech innovation. At BME, all the necessary components are in place for deep tech innovation: foundational research, its implementation and application, and large-scale development projects, Levendovszky noted.

He also cited several specific examples of achievements in BME’s AI research, ranging from multilingual sentiment analysis and the monitoring of bridge statics to the measurement of fluid viscosity and the real-time detection of deepfake fraud. BME continues to seek new collaborators through the BME Partner Programme, which offers access to specialised competences, RDI outcomes, infrastructure, and training programmes.

A multi-fold return on investment

“In the long term, investing in an innovation ecosystem that includes the university yields measurable benefits for companies. This is a self-sustaining cycle, where funding returns many times over through the university’s value-creating processes – in the form of RDI results, highly qualified professionals, and increased revenue,” the Vice-Rector stated. (The presentation is available here.)

István Németh, Associate Professor at the Department of Manufacturing Science and Engineering, demonstrated practical applications of AI in mechanical engineering in his presentation Artificial Intelligence and Digital Technologies in Mechanical Engineering. Examples included maintenance-planning methodologies and software that estimate and predict failures using machine learning, as well as the modelling of degradation processes and the development of discrete-event simulation models. (The presentation is available here.)

Bertalan Forstner, Associate Professor at the Department of Automation and Applied Informatics, presented on the real-time deepfake detection development previously mentioned by János Levendovszky. This project – already the recipient of several innovation awards and previously featured on the bme.hu website – can detect deepfake manipulations in live video with 90–99% accuracy, compared to a 63% success rate for human observers. The technology is expected to play an increasingly important role not only in call centres but also in safeguarding internal corporate communications in the coming years.

The presentations were followed by a panel discussion titled Artificial Intelligence and the Future – Cooperation between University and Industry. Róbert Keszte, Head of Country at Aumovio; Zsolt Kenesi, Leader of AI Research at Ericsson; Sándor Imre, Dean of the Faculty of Electrical Engineering and Informatics (VIK); Vince Balázs Nagy, Associate Professor at the 
Department of Mechatronics, Optics and Mechanical Engineering Informatics; and Pál Varga, Head of the 
Department of Telecommunications and Artificial Intelligence answered questions from the audience and moderator Tamás Iványi, Assistant Professor at the Department of Management and Business Economics.

The first question was: what is expected from students? Responses included domain-specific knowledge, open-mindedness, intrinsic motivation to work in the industry and to create real value instead of merely appearing productive, as well as adaptability as a fundamental skill. Zsolt Kenesi added that in industrial settings, decision-makers often have minimal understanding of AI, and they too must be educated – a role in which BME can contribute significantly.

Lecturers also shared observations that new generations can be considered “AI natives,” with some pupils in as early as 5th or 6th grade already using such technologies regularly. At BME, three approaches are applied: in some subjects, the use of AI is prohibited; in others, it is explicitly encouraged; and in between, students are required to disclose where and how they have used AI in their work.

But in which sectors is AI expected to advance most significantly? According to the panellists, cybersecurity and cost efficiency are key areas, along with certain industry-specific domains. Sándor Imre noted that centralisation will soon be followed by decentralisation – i.e. segmentation – towards smaller-scale systems incorporating local hardware and energy production solutions. It is already evident that AI is not confined to the IT industry: it permeates all sectors, giving rise to new professions. More and more companies now have AI officers or even Chief AI Officer roles.

gp