Today, our blog hosts a brief interview with Pavlos Bouzinis, PhD, Research Engineer from Metamind innovations. In this short interview, he shares insights into his role, his motivation for joining the field, and how MINDS is contributing to one of the project’s key technical developments—the F-BOX tool.
Pavlos, we delighted to have this short interview with you today in the frame of #meet the D-HYDROFLEX team blog series. You bring a strong background in artificial intelligence and cybersecurity to the D-HYDROFLEX project.
Q: Would you like to introduce yourself to get to know you?
A: Yes, of course. I hold a Ph.D in Electrical & Computer Engineering from Aristotle University of Thessaloniki, where I also earned an integrated master’s degree. Currently, I work as an AI engineer & researcher at MetaMind Innovations (MINDS). My interests include designing machine learning techniques and systems within the realm of cybersecurity. I am also deeply engaged in optimization theory, and I have a strong passion for problem-solving and being involved into research activities.
Q: Could you describe your company’s role on the project?
A: MINDS role as a technology provider is to integrate and expand the F-BOX tool into D-HYDROFLEX project. F-BOX is a federated solution for detecting and discriminating various threats and cyberattacks. Based on network traffic data and system logs, F-BOX incorporates multiple deep intrusion/anomaly detection models into a federated level. Moreover, MINDS has a crucial role within D-HYDROFLEX, by overseeing the design of the D-HYDROFLEX Hydropower 4.0 toolkit. This involves creating a reference architecture and providing explicit guidelines for developing software components, as well as outlining reference scenarios with validation criteria.
Q: We would like to know more about federated learning.
A: Federated learning (FL) is a decentralized learning technique, which enables multiple entities to collaboratively build a machine learning (ML) model, with the aid of a third party. One of the prominent features of FL is the retention of the training data at the source of generation. This implies that the involved entities do not share their raw data with any other participant or third party. Hence, FL’s primary role is to provide privacy among the involved participants, while enabling shared knowledge for building a unified ML model, without the exchange of sensitive information. In the context of D-HYDROFLEX, FL will be used for creating intrusion detection AI models, without compromising user privacy.
Q: What are the main challenges of your work on the project, and how do you tackle them?
A: One of the core challenges we face is ensuring the seamless integration of F-BOX within D-HYDROFLEX. To achieve this, we embrace a cooperative strategy, collaborating with the project partners to synchronize our goals and overcome technical issues, while also identifying critical points and orchestrating a plan to carry through with them. We aim to take a proactive approach by anticipating potential challenges before they arise.
Q: What are your expectations from the project? What impact will the project have on the energy ecosystem?
A: Our expectations from the project are to have a positive impact in hydro power plant digitalization, specifically by establishing data privacy, and enhancing cyber-resilience and reliability. From our perspective, D-HYDROFLEX will contribute to strengthening cybersecurity measures within the energy ecosystem. Through its efforts, the project aims to implement advanced cyber-shielding mechanisms, towards safeguarding critical infrastructure and ensuring resilience against emerging cyber threats.
Pavlos, really thank you for this discussion. It was a pleasure having you with us today!
Another # meet the D-HYDROFLEX team blog story is completed. Stay tuned to learn more on our workforce team!
Stay connected to learn more on D-HYDROFLEX workforce and activities. Followed us on LinkedIn and on X.
Please subscribe here to our newsletter to keep updated on our activities.
