Pokročilé AI nástroje pro bezpečnou vzdálenou kontrolu autonomních vozidel

The project "Advanced AI Tools for Safe Remote Control of Autonomous Vehicles" focuses on the development of a system for advanced image augmentation within the remote control of vehicles using artificial intelligence. The aim is to significantly enhance the situational awareness of the teleoperator, enabling safer and more efficient operation of remotely controlled autonomous vehicles in real time. The project will result in two outputs in the form of functional prototypes. The first will be a comprehensive remote control system that integrates machine vision methods and neural networks to combine camera imagery with LiDAR data. The outcome will be a visually consistent, contextually enriched scene representation, stable even under changing lighting conditions. The system will include automatic sensor calibration, the elimination of adverse lighting conditions, the merging of camera images into a panorama, object detection, and the highlighting of hazardous situations. The second output will be a test device for network communication. This device, designed for automated execution of test scenarios simulating network faults and identifying potential issues, will be especially useful in testing the communication systems of vehicles with autonomous functionalities. Through its focus, the project directly contributes to safer and more efficient deployment of autonomous vehicles during the transitional phase toward full autonomy. The project will be implemented by Roboauto s.r.o. in collaboration with Brno University of Technology.