PREMOVE (PROYEXCEL_00684)

Movement prediction of traffic participants for the safe integration of self-driving cars in urban areas 

Abstract

Thanks to recent advances in the development of intelligent control systems, autonomous vehicles are beginning to emerge and are being tested on the road. But applications in urban areas, such as city centers, with pedestrian areas, bicycle lanes and high traffic, are more demanding for the development of these intelligent systems, and in particular, the safety subsystem for collision avoidance. In this type of environment there is a traffic pattern in which mostly pedestrians, skaters and cyclists mix with private and goods vehicles sharing the same space. This can compromise their safety, even taking into account that speed limits in these urban areas are usually lower than in other unrestricted urban areas.
 

The objective of this project is to develop techniques for detecting and predicting the movement of traffic participants for collision avoidance, which will contribute to the safe integration of self-driving cars in restricted urban areas. A multi-sensor system will be designed to obtain information from traffic participants and intelligent techniques, such as deep neural networks, supported for tuning in simulated environments, will be applied, transferred to the real application and experimentally validated.

Background

Eloy Vergara Gómez (2020). Desarrollo de un simulador para el diseño de sistemas de percepción y control de vehículos autónomos en el entorno de la ampliación del Campus de Teatinos. Master thesis in Mechatronics. Universidad de Málaga. Link RIUMA.

 

Publications

Daniel Steven Gamba Correa (2023). Incorporación de sensores realistas en la simulación en CARLA de la navegación de coches autónomos. Degree thesis in Electronics, Robotics and Mecatronics. Universidad de Málaga. Link RIUMA.

Contact

Jesús Morales Rodríguez. Tel. +34 951952323, email: jesus.morales@uma.es

Jorge L. Martínez Rodríguez. Tel. +34 951952322, email: jlmartinez@uma.es