Přístupnostní navigace
E-application
Search Search Close
The Faculty of Information Technology at Brno University of Technology (FIT BUT) hosted the 11th edition of the student conference Excel@FIT on May 6, 2025. The event traditionally focuses on innovation and new technologies in the field of information technology. This year, a record-breaking 95 projects by bachelor’s and master’s students were submitted, with expert panels selecting 77 for the final competition. The conference is an annual opportunity for students, academics, and industry professionals to meet, share experiences, and foster collaboration.
Excel@FIT 2025: A Record Number of Projects Showed the Direction of IT Development | Author: Václav KoníčekThis year again, the Excel@FIT conference connected FIT BUT students with professionals and representatives from over twenty partner companies, helping to build the much-needed bridge between academic research and its application in industry. “The conference gives students the opportunity to present the excellent results they’ve achieved. It’s also one of the events through which we aim to deepen our faculty’s cooperation with companies and industry, because as a technology faculty, a strong link to practice is essential for us,” noted FIT BUT Dean Petr Hanáček in his opening remarks. This year’s programme kicked off with a panel discussion titled “What Did Studying at University Give Me and Where Am I Now?”, featuring representatives of tech companies, many of whom are FIT BUT alumni. The panelists emphasized that in today’s job market, a university diploma alone is not enough – language skills, international experience, and selected soft skills also matter, including self-presentation, which the conference actively put to the test.
From Faulty Solar Panels to AI-Generated Fairytales
Participants presented their projects in the form of posters accompanied by live explanations. This year’s record number of submissions covered several thematic areas reflecting current IT trends and the faculty’s research focus. A strong presence was seen in the area of artificial intelligence and machine learning, not only through research approaches but also practical applications in healthcare, industry, public administration, and cybersecurity.The conference opening was attended by representatives of partner companies. | Author: Václav KoníčekMany projects also focused on image, sound, and video processing – especially when combined with data analysis and real-time monitoring. Topics included cybersecurity (e.g., deepfake video detection or comparison of privacy tools in web browsers), digital identity management, public data visualisation, educational system development, and the use of blockchain in critical applications.
One standout project was a tool for automated fault detection in photovoltaic panels using drones and thermal cameras. “The goal is to speed up solar power plant inspections. Currently, each image captured by a drone and thermal camera must be manually reviewed by an expert. I’ve implemented a system that automates the process using an orthophoto map and a trained classification model,” explained project author Matěj Gotzman. In the medical field, Eva Mičánková presented a tool for automated polyp segmentation in colonoscopy, a project that handles large datasets and requires preprocessing for machine learning – a highly specific topic, as the author herself noted. Jiří Soukup trained models to analyse CT scans of lungs to detect cancerous changes.Students presented their projects using posters. | Author: Václav KoníčekPractically oriented projects also included a tool for detecting AI-generated text (Matej Koreň), useful for combating plagiarism and monitoring language model outputs. In the field of education and digital creativity, there was a voice-controlled web platform for practising math (Dominik Horut) and a storytelling app for children where a language model generates a story and a voice module turns it into speech. Project author Barbora Šmondrková summarised her motivation: “I wanted to focus on connecting children with AI. It will be an important topic in the future, and this is a simple way to show children what AI can do and what the real outputs of large language models can be.”
Conceptually demanding topics included a project by Šimon Šmída, who explored the interpretability of black-box machine learning models and ways to increase transparency and trust in their decisions. A socially impactful project came from Tomáš Švondra, who developed a decentralized electronic voting application for national elections built on blockchain architecture.
In the sports domain, Maksim Samusevich presented a project on the automated analysis of football videos, using computer vision to track and analyse players, referees, and the ball. All conference contributions are available in the online proceedings.Winners of the public vote; from right: Matej Olexa, Dominik Honza, Jan Seibert, Pavel Štarha and Jakub Jeřábek. On the far left: Associate Professor Milan Češka. | Author: Václav KoníčekA Promising Future for the Faculty and Industry
During the awards ceremony, the high quality of student work was repeatedly praised, along with the strong applicability of the presented concepts. Projects were assessed for their research contribution, technical implementation, and commercial potential in three categories – awards given by expert panels, industry partners, and the professional public. The expert panels selected 18 outstanding projects whose authors also received scholarships. Several projects ranked among the best across multiple categories. Pavel Štarha’s project on an educational chessboard for modern teaching of the classic game succeeded in all three categories. The top five presentations in the public vote (Jiří Kunovský Award), alongside Štarha’s work, included projects by Matej Olexa (monitoring of Bluetooth Low Energy devices), Dominik Honza (an app with an interactive map of Czech fishing grounds), Jan Seibert (a game simulation of a nuclear power plant), and Jakub Jeřábek (an open-source platform for public blockchain analysis of Tron).
Excel@FIT 2025 once again confirmed that student research can be not only technologically and scientifically rigorous but also socially relevant and practically applicable. For many students, it was a valuable first public presentation of their work – and a unique opportunity to gain important feedback from industry partners and experience communicating their research.