|
Abstract
|
New assistance systems and the applications of autonomous driving of road vehicles imply ever-greater requirements for perception systems that are necessary in order to increase the robustness of decisions and to avoid false positives or false negatives.
In this sense, there are many technologies that can be used, both in vehicles and infrastructure. In the first case, technology, such as LiDAR or computer vision, is the basis for growth in the automation levels of vehicles, although its actual deployment also demonstrates the problems that can be found in real scenarios and that must be solved to continue on the path of improving the safety and efficiency of road traffic.
Usually, given the limitations of each of the technologies, it is common to resort to sensorial fusion, both of the same of type sensors and of different types.
Additionally, obtaining data for decision-making does not only come from on-board sensors, but wireless communications with the outside world allow vehicles to offer greater electronic horizons. In the same way, positioning in precise and detailed digital maps provides additional information that can be very useful in interpreting the environment.
The sensors also cover the driver in order to analyze their ability to perform tasks safely.
In all areas, it is crucial to study the limitations of each of the solutions and sensors, as well as to establish tools that try to alleviate these issues, either through improvements in hardware or in software. In this sense, the specifications requested of sensors must be established and specific methods must be developed to validate said specifications for the sensors and to complete the systems.
Language: en |