REFERENCE ELEMENT GUIDANCE OF AUTONOMOUS VEHICLES USING 3D LASER SCANNER

Abstract

Every transportation plan in the medium and long term consider autonomous vehicles as a key solution for transport, which will reduce to a minimum traffic accidents and congestion, maximizing the use of the road. However, at present, these vehicles are still under development and they are not with acceptable figures for the placing on the market. A fundamental element on today obstacle detection is the laser scanner, which provides a three dimensional mesh that can represent all objects that are at a distance of 200 meters from the vehicle. Furthermore, the combination of GPS with inertial systems allows precise positioning on the road, if necessary reinforced with computer vision to achieve lane level positioning. However, there are situations, such as tunnels or urban roads with poor reception of GPS signals that make this positioning can be complicated, so much so that the autonomous vehicle requires human intervention to maintain the trajectory. This article describes a use of laser scanner technology to support positioning and maintaining the position even when the GPS and inertial systems are not sufficient to maintain the route. This system will be complementary to conventional guidance systems and works in parallel without interference obstacle detection system for autonomous vehicles. It has also been tested and validated in real situations autonomous vehicles driving at high speeds on real roads.