web analytics

Department of Intelligent Transport Systems

Signed agreement for assigning funds for a research project of the Croatian Science Foundation


On Tuesday, December 22, 2021, at the Faculty of Transport and Traffic Sciences, the agreement for the assigning of funds from the Croatian Science Foundation under the tender “Research project, deadline 2020-02” for the research project IP-2020-02-5042 “Development of Learning Agent-based Systems for Improved Urban Traffic Control (DLASIUT)” led by Assoc. Prof. Edouard Ivanjko was signed. The signing of the contract was attended by the dean Prof. Tomislav Josip Mlinarić, Vice Dean for Science and External Cooperation Assist. Prof. Marko Šoštarić and secretary Romina Jovančević.

The project’s total value is HRK 1,088,840.00, with the first installment of HRK 599,130.00. The project will last for four years from January 15, 2021, to January 14, 2025, and is implemented in collaboration with foreign researchers Prof. Ivana Dusparic from TCD, Dublin, Ireland, Prof. Daniela Koltovska Nečoska from the Faculty of Technical Sciences, St. Kliment Ohridski University – Bitola, Northern Macedonia and Prof. René Schumann from the Institute of Information Systems, HES-SO Valais, Switzerland. In the project, local researchers Prof. Sadko Mandžuka, postdoc Martin Gregurić, Ph.D., Dino Čakija, B.Sc., Krešimir Kušić, mag. ing. traff., Mladen Miletić, mag. ing. traff. and Filip Vrbanić, mag. ing. traff. will also participate with the plan for the employment of one doctoral student.

The goals and expected results of the project are:

  1. Identified most representative classes of traffic scenarios with which the traffic controller must cope.
  2. Developed reinforcement learning multi-agent-based traffic controller for urban traffic environments.
  3. Developed framework for structured learning/testing of learning multi-agent-based traffic controller.
  4. The developed traffic controller augmented for the management of traffic flows containing connected and autonomous vehicles.
  5. The impact of the developed learning multi-agent-based controller on the improvement of urban traffic control analyzed.

Updated: January 12, 2021 — 12:13
Zavod za inteligentne transportne sustave, 2017.