Digital Twin Oriented to the Validation and Optimization of Perception, Planning, Localization, and Control Algorithms in Autonomous Vehicles Using Advanced Artificial Intelligence and Specialized Hardware

Contemporary mobility faces critical challenges, such as high road accident rates and the need for inclusive transport systems. Given the limitations of physical testing for the ANTA autonomous vehicle, the Universidad Politécnica Salesiana proposes the creation of a digital twin. This advanced virtual environment, based on artificial intelligence and specialized hardware, will allow for the validation and optimization of perception, planning, localization, and control algorithms in controlled and complex scenarios. The project addresses the need for robust simulation tools that consider the specificities of the Ecuadorian road environment, such as variable weather and lighting conditions. The methodology integrates historical-descriptive, experimental, and analytical approaches, using open-source simulators and distributed computing architectures. Expected results include the consolidation of experimental capabilities, the generation of at least four indexed publications, one intellectual property, and an operational digital twin. This development will not only strengthen applied research in autonomous mobility but will also drive technology transfer and contribute to innovative solutions for safer, more efficient, and inclusive mobility in the country.<br/><br/><b>Goal</b>: <br/>Develop a digital twin to validate and optimize perception, planning, localization, and control algorithms in autonomous vehicles using advanced artificial intelligence and specialized hardware. This virtual environment will allow for testing algorithms in complex scenarios, reducing risks and experimentation costs.<br/><br/><b>Research lines</b>: <br/>Maintenance engineering applied to transportation<br/>Transportation and mobility