People
Ing. Pavel Mašinda

doc. Ing. Martin Leso, Ph.D.; Ing. Pavel Mašinda; doc. Ing. Tomáš Tichý, Ph.D., MBA

The proposed project focuses on the complex issues of rail transport automation, particularly transitioning from the current automation level GoA 2 to the fully autonomous GoA 4. The project will thoroughly analyse the processes and safety requirements essential for this transition while assessing the operational efficiency and feasibility of full autonomy. The core objectives include evaluating the technical and safety aspects of fully autonomous rail systems, determining the suitability of eliminating human operators, and collecting data for optimizing Automatic Train Operation (ATO) systems. Special attention will be given to decision-making modules within on-board ATO systems that utilize data from ETCS (or other ATP - automatic train protection) and ATO trackside components. These modules are not being developed but leveraged as tools to collect and analyse critical data about the behaviour of autonomous systems. The project builds on the current research in autonomous and intelligent transportation systems, reflecting global efforts to modernize railway infrastructure, increase capacity, reduce human error, and improve environmental impacts. The outcomes will be valuable for academia, regulatory standard development, and practical implementation in autonomous railway systems. Moreover, the project will support education and training in traffic management and automation at the Faculty of Transportation Sciences.

2025 - 2026

Studentská grantová soutěž ČVUT - SGS25/104/OHK2/2T/16

Ing. Lukáš Kacar; doc. Ing. Tomáš Tichý, Ph.D., MBA

The proposed project deals with the current issue of traffic detectors used in cities with a focus on their appropriate selection and the use of traffic data measured by these detectors. Cities seek to improve traffic flow and the passability of individual intersections. For this purpose, data from traffic detectors and other detection methods such as FCD or C-ITS are used. These are used to measure traffic flow, environmental load, as well as other necessary characteristics to apply new benefits for sustainable urban mobility. They are also used in intersection areas for their dynamic control. Nowadays, especially in intersections, induction loops are most commonly used, which often do not work and are expensive to repair or the data they provide is not further used. It is an effort to make more efficient use of this existing data while installing sustainable detectors that will provide more traffic data that can then be used by more entities. The aim of the project is to optimise the appropriate deployment of detectors in public space for future predictions of individual and public transport, pedestrian, cyclist, supply and potential autonomous mobility behaviour, including the trajectories of all road users, using new AI approaches. Important elements of AI are not only language models based on deep neural networks such as YOLOv8, which have been successfully applied in video image analysis or thermal imaging, but also other approaches such as multi-agent systems, fuzzy logic, etc. All these innovative approaches have a major impact on future traffic optimization as well as the use of alternative detections such as data from mobile operators, FCD, C-ITS and other innovative approaches within area management. The effort will be to model and optimize the proposed solution, with subsequent evaluation of specific designs and usecases in the proposed area that will be usable for more accurate management and information in the area, with links to other telematics sys

2025 - 2026

Studentská grantová soutěž ČVUT - SGS25/105/OHK2/2T/16