GAZ self-driving vehicles provide for fully autonomous movement along a route set by the operator. The vehicles are able themselves to identify both stationary and dynamic obstacles on the route. A central computer assesses the possibility of crossing the trajectory of a vehicle and an obstacle. Depending on the selected scenario, either the route is corrected for a safe manoeuvre or the vehicle comes to a stop, including by emergency braking if necessary.
Both vehicles include navigation systems based on high precision GPS/GLOMASS modules with differential correction receivers allowing the vehicle’s location to be determined with a precision of up to 1 cm. In order to ensure more stable operation, the systems are supplemented with inertial navigation units allowing the location to be monitored in the event of a short interruption to the satellite signal (for instance, when driving under a bridge). Another element of the complex for receiving high-precision data on a vehicle’s location is the system for recognising road markings and road infrastructure.
One of the presented self-driving vehicles uses an additional localisation system consisting of laser Lidars. The vehicle has an original technology using indirect features to recognise the boundaries of the roadway if there are no road markings or they are invisible. This technology was developed jointly by GAZ and teams of scientists from the Nizhny Novgorod R.E. Alexeev State Technical University.
The other vehicle has a driving system based on a stereo-camera – passive devices for gathering information about the environment. Because of their low level radiation, these devices allow the system to orientate itself in space without interfering with other self-driving vehicles; the devices themselves are protected from interference, including against cyberattacks. The system includes several blocks of wide-angle stereo-pairs forming a “depth map” around the vehicle, an all-round view subsystem and a wide-radius ultrasonic sensor system. For making the best use of the computing capacity, the data from external sensors are processed with varying frequency, depending on the speed, interference, weather and road conditions.
The solutions developed under the GAZ self-driving projects will be used not only for creating totally autonomous vehicles but also in Advanced Driver Assistance Systems (ADAS), which it is planned to install on GAZ vehicles in serial production.
For testing self-driving cars and assessing the functioning of the ADAS, a special area has been set aside at the GAZ testing range imitating a city environment, allowing the vehicles to be tested in all-year-round conditions.
GAZ has many years’ experience of developing self-driving vehicles and training staff for high-tech developments. Since 2008, the company has been supporting a “Robotics” programme including testing of automated systems with the participation of the country’s leading higher educational institutions in the field. Today, 72 programme graduates are working for the GAZ engineering services.
In July 2018,GAZ signed an agreement with the Moscow Government, Yandex and FGUP NAMI on development and introduction of driverless transport in Moscow. Under the agreement, it is planned to develop telematic transport technologies and smart vehicle systems and services based on them. The document also provides for joint provision of cyber security during operation of driverless transport and popularisation of autonomous driving with specific effects for the city transport system.