Explore CTU's Digital Library

Access research publications, theses and academic works

Faculty of Civil Engineering Faculty of Mechanical Engineering Faculty of Electrical Engineering Faculty of Nuclear Sciences and Physical Engineering Faculty of Architecture Faculty of Transportation Sciences Faculty of Biomedical Engineering Faculty of Information Technology Klokner Institute Masaryk Institute of Advanced Studies Central Library Acta Polytechnica Acta Polytechnica CTU Proceedings CTU Publishing Activity

Recent Submissions

Weakly-supervised Computer Vision Methods for Autonomous Driving
(Czech Technical University in Prague. Faculty of Electrical Engineering. Department of Cybernetics, 2026) Neumann, L.
The thesis summarises the author’s post-PhD scientific work in the field of computer vision for autonomous driving. The common theme across the presented methods is the reduction of the reliance on costly human annotation, which is one of the main obstacles to training reliable perception models at scale. Instead of labelled data, the methods draw supervision from signals that are essentially free to obtain – the laws of physics, projective geometry, the temporal consistency of a scene observed over time, or the predictions of simpler, readily available off-the-shelf models. Across several autonomous driving tasks, models trained without domain-specific human labels are shown to match or closely approach the accuracy of their fully-supervised counterparts, while being substantially cheaper to supervise.
STRENGTHENING THE BEARING AREAS OF STEEL BRIDGES USING UHPC
(Žilinská univerzita v Žilině, 2025) Boutar Y.; Ryjáček P.; Starýchfojtů M.; Stančík V.; Kolpaský L.
Steel railway bridges are critical components of transportation infrastructure but often suffer from degradation due to long-term exposure to environmental factors such as corrosion and the ageing of protective systems. Over time, legacy design features, including inadequate moisture drainage and poor detailing at key joints, contribute to localised deterioration, further compromising the structural integrity of these bridges. In light of these challenges, this study investigates the potential of ultra-high-performance concrete (UHPC) as a reinforcing material to improve the performance of steel bridge structures. The focus of this research is on the role of shear connectors in facilitating the composite action between the steel structure and UHPC. By improving the bond between these materials, shear connectors help optimise the overall performance of the reinforced system. The study examines various shear connector configurations, aiming to identify the most effective solutions for different types of steel bridge designs. The findings provide valuable insights into the behaviour of shear connectors, including their role in enhancing shear resistance and minimising slip, which are critical factors in the long-term stability and safety of retrofitted bridges.
From Zdiby to Prague: Designing Tram Logistics for Alza
(IEEE eXplore, 2026) Harant M.; Šedivý M.; Horák T.; Svítek M.
Urban logistics faces increasing pressure to reduce emissions and traffic congestion while maintaining high service levels for e-commerce. This paper develops a general framework for evaluating cargo-tram integration and applies it to a Prague case study linking the planned Zdiby extension with four major Alza branches. The routing problem is formulated on a directed tram graph and solved by a state-extended Dijkstra’s algorithm with turn penalties and stop-visitation states. A second screening step evaluates night-time timetable compatibility on shared passenger infrastructure. For the Prague case, the feasible loop serves Vltavská, Dejvická, Anděl and Pankrác, departs at 03:10, and returns to Zdiby at 06:21. Under the assumed vehicle layout, one tram can transport up to 19 roll cages, corresponding to an upper bound of 4.75 t or approximately 1,710 parcels per trip. The results confirm technical feasibility for night operation, while also identifying noise, timetable coordination, infrastructure access, and
Principles of Object–Process Methodology for System Architecture Modeling in Complex Transport Systems
(IEEE eXplore, 2026) Kremlík V.; Svítek M.; Merunka V.; Kofránek J.
The paper introduces Object–Process Methodology (OPM; ISO 19450) as a unifying approach for designing a modular functional architecture of Mobility Hubs. OPM integrates the structural dimension (objects, assets, roles) and the behavioural dimension (processes, information and energy flows, control) into a single, coherent model. The paper outlines the core principles of OPM and its bimodal representation: graphical Object–Process Diagrams (OPDs) are paired with an equivalent textual specification in Object–Process Language (OPL). This duality supports intelligible communication with non-technical experts, facilitates requirements validation, and enables early detection of errors such as incompleteness, contradictions, or missing links between functions and resources. The practical value of OPM is demonstrated through a simplified case of passenger rail transport within the Czech Republic, where the end-to-end passenger journey is hierarchically decomposed. The example illustrates how OPM integrates processes with the required objects and resources (e.g., the app, ticket, travel plan, station, train, signaling) within a single coherent OPD/OPL model, enabling systematic checks of linkage completeness, decision conditions, and cross-level consistency.
Synergy of Semantic Modeling and Multi-Agent Systems for Resilient Smart City Utility Networks
(IEEE eXplore, 2026) Walletzky L.; Svítek M.; Kozhevnikov S.; Hruška Z.
Bridging the gap between conceptual service design and technical execution remains a critical challenge for Smart City resilience. This research presents a novel integration of the 4DocMod semantic framework with Multi-Agent Technology (MAT) for autonomous utility network management. Using the ”Diamond See” methodology, we deconstruct technical ontologies into dynamic semantic maps. The proposed framework enables ”paradigmatic resilience,” allowing AI agents to switch operational logics—from economic optimization to crisis survival—via a ”Context Bridge.” This approach ensures systemic stability in complex urban ecosystems without requiring modifications to the agents’ core programming.