Portal de Cartagena

www.portaldecartagena.com

• Home
• Business
• Hotels
• Restaurants
• News
• Advertisement

Researchers from the Polytechnic University of Cartagena (UPCT) are developing for the General Directorate of Traffic (DGT) an intelligent decision making system for autonomous vehicles based on the behavior of different types of drivers.

"The objective is to obtain more realistic behavior models and associated with the type of road and real traffic situations to improve autonomous vehicles", explains Pedro Javier Navarro, head of the Intelligent Vehicles Laboratory of the UPCT, which has a car, called Cloud Incubator Car (CIC), which has high definition sensors that capture 1.3 million points per second with an accuracy of two centimeters, as well as vision cameras and an inertial unit with GPS.

The Laboratory, which has just acquired a new vehicle to automate it with other technologies, will exhibit the ClC at the fair that takes place this Saturday at the Industrial School on the occasion of the Polytechnic Open Door Conference.

The driving study will involve drivers of different ages and gender and will be carried out in urban scenarios with real traffic.

"The decision making of an autonomous vehicle is conditioned by the type of road, environmental conditions, the behavior of other vehicles and the mood of its occupants, these last two factors being difficult to estimate and high uncertainty," Navarro details. .

The driving tests will collect millions of data from the actions, behavior, biometric parameters and dexterity of the drivers, including physiological sensors to measure the driver's condition and study their influence on driving behavior

The database obtained will be used to train an artificial intelligence system based on supervised learning in order to improve the decision making system of the vehicle in urban environments and reduce the uncertainty associated with the behavior of drivers.

"In the coming years, autonomous vehicles will have to coexist with a global automobile fleet estimated at 1.2 billion vehicles with driver," argues Navarro.

The decision making system will be implemented in the autonomous vehicle of the UPCT, integrating it with the vehicle's control algorithms and will be evaluated in real traffic scenarios.

Source: UPCT