Autonomous Driving

Applications
Title	Autonomous Driving
Edited by	GMV
Level	Intermediate
Year of Publication	2011

The main drivers for achieving autonomous driving is the reduction of traffic accidents by eliminating human error, increasing road capacity and traffic flow by reducing distance between cars and making use of traffic management information, relieving the car occupants from driving and navigation activities and allowing them to engage in other activities or rest.

A driverless car requires the combination of several techniques among which GNSS. These techniques will enable to guide autonomously a land vehicle from one point to another using public roads. In autonomous driving, GNSS can be used for navigation by determining the vehicle location and speed. With this information the vehicle route can be decided using digital maps. Lane and attitude determination could also benefit from GNSS if the accuracy is good enough. If the location information is shared among cars, GNSS could be a part of a short-range situation awareness system (awareness of other vehicles in the road and collision avoidance) although it is not expected that GNSS is the sole means of information for short-range situation awareness.

Application Architecture

Autonomous Vehicle Prototype

A autonomous vehicle would have a GNSS receptor and antenna installed, providing positioning to the car's autopilot system. The autopilot system uses this information to decide the route to take using digital maps and eventually traffic information.

Normally autonomous car prototypes rely on short range sensor (cameras, ultrasound, radar, lidar, etc) for short-range situational awareness. Short-range situation awareness allows the car's autopilot to determine the lane where is traveling and to be aware of the surrounding environment including other cars and obstacles. GNSS can provide additional information for short-range situational awareness but it is not likely that it is used as the sole means for lane determination and situation awareness. For the use of GNSS for short-range situation awareness the accuracy would have to be at decimeter level and the information of the position of the cars would have to be shared between neighboring cars. This would be helpful but it wouldn't cover for other obstacles that might be on the road.

Using road infrastructure (such as rails or electronic markers) for navigation could make autonomous cars simpler in technical terms but would require an infrastructural cost that might be too high. Also this would limit the roads that an autonomous car could use to the ones having this infrastructure installed.

Application Characterization

Autonomous Driving is still a research area and there aren't yet vehicles approved to be driven without human supervision. The existing vehicles are research prototypes that still cannot run autonomously 100% of the time.

A restricted form of Autonomous driving is Advanced Driving Assistance Systems (ADAS). ADAS combines vehicle capabilities to improve mobility and active safety. GNSS will provide important additional data to ADAS on the vehicle’s environment. ADAS then warns the driver of imminent danger or takes full or partial control over the vehicle. For instance, the speed could be reduced by ADAS under bad visibility conditions if the car approaches a tight turn too fast. This function would be possible with accurate position data of guaranteed integrity and accurate mapping information.

Application Examples

Examples of autonomous vehicle projects are:

• Google driverless car - Project by Google that involves developing technology for driverless cars.
• EUREKA Prometheus Project - The PROgraMme for a European Traffic of Highest Efficiency and Unprecedented Safety was the largest R&D project ever in the field of driverless cars and was funded by the European Commission.
• BRAiVE - Prototype autonomous vehicle developed by VisLAb of the University of Parma .