Developers: | StarLine (NPO StarLine) |
Last Release Date: | 2018/11/22 |
Branches: | Transport |
Content |
2024: Launching trucks without people in the cab on the Moscow-Petersburg highway
The Sber company in September 2024 launched tests of unmanned trucks on the Moscow-Petersburg highway without the presence of a person in the cockpit. The tests were carried out as part of regular flights along the M11 highway. The introduction of new technologies allows the company to significantly improve the management of cargo transportation on long routes.
The StarLine truck operates unmanned flights between St. Petersburg and Moscow on a weekly basis. The tractor, working on the basis of artificial intelligence technologies and neural networks, uses a digital road map developed by NPO StarLine. This system allows the car to choose a lane, adjust speed and take into account road signs. The new stage of testing was launched in accordance with a decree of the Russian government, which regulates experimental legal regimes for unmanned vehicles.
The main task of the tests was to confirm the possibility of fully autonomous control without the participation of a person behind the wheel. To do this, during the flights, the operator, who was previously in the cabin, was moved to the passenger seat, which made it possible to exclude his interference in the control of the vehicle. Engineers of NPO StarLine note that the first test results were positive, and the system fully coped with the task.
According to Ilya Nikiforov, head of the development department of NPO StarLine, the introduction of such technologies was an important step for the entire transport industry in Russia. In the summer of 2024, a new system for automatic cleaning of truck sensors and cameras was developed, which allows drones to operate effectively in harsh climatic conditions. These technologies help prevent pollution of lidars and cameras, which previously caused serious problems on long routes.
2023: Scheduled flights on the Moscow-Petersburg highway
Trucks without a driver behind the wheel began to make regular flights along the Moscow-Petersburg highway. This was reported in early October 2023 by the Starline NGO, which is implementing this project.
The company carries out unmanned cargo transportation between the two Moskov and St. Petersburg weekly. A highly automated truck moves on a route mission, using technologies in the fields of unmanned driving: artificial intelligence, neural networks, as well as the development of St. Petersburg scientists from NPO Starline - a digital model of the road, that is, a special map for local maneuvering along the route, choosing a lane, speed, obtaining information about road signs and restrictions.
To move along the highway in highly automated mode, recognize other participants in the movement, predict their behavior and localize the car in space, the drone uses a number of special sensors: GNSS-RTK receiver, lidars, video cameras, inertial sensors. All data was processed by several computers with their own software installed in cars.
In early October 2023, the Starline NGO reported successful tests of the StarLine unmanned tractor during a storm. The truck drove 20 km along the highway in unmanned mode at a speed of 65-67 km/h, rebuilt the route and returned back.
On the eve there was a storm warning, in such extreme weather we have not yet traveled. We decided to check how the unmanned truck behaves with a gusty wind of 16 meters per second and rain. It was interesting to see how the steering system works it out. The speed is due to the detection range in weather conditions. If we suddenly have something in front of us that needs to be urgently slowed down, we should have enough braking distance for a complete stop, provided that we detonate, "said Ilya Nikiforov, head of the development department of NPO Starline.[1] |
2021: Start testing unmanned trucks on the Moscow-Petersburg highway
On October 20, 2021 electronic engineers NPO Starline , a Russian car developer announced the testing of its pilotless truck. He will be tested on the track - MoscowSt. Petersburg (, " M11 Neva").
The company creates a digital model of the highway and equips the truck with equipment that will solve the problems of localizing the car in space, recognizing objects and building traffic routes. How long the experiment will last is not reported.
According to the NGO Starline, the company is developing a project that is a universal unmanned platform - a comprehensive solution aimed at integrating unmanned technologies into any vehicle and interacting with an intelligent transport system (ITS). By October 2021, Starline has two unmanned passenger cars that have previously passed a series of state tests and are actively tested on the roads of St. Petersburg and other cities of Russia.
Unlike cars, trucks most often move along the highway, driving tens or even hundreds of kilometers without a single intersection and traffic light. This greatly simplifies the task for developers of self-driving cars, mainly reducing it to keeping in the lane and tracking passing cars in their and neighboring lanes.
KamAZ and Evocargo are also engaged in the development of unmanned trucks in Russia. The first has different prototypes by October 2021: most of them are based on a regular truck, and last year the company showed an unmanned van without a driver's cab. Evocargo also created a truck without a cab, but with a different design resembling Einride trucks. Another difference from almost all cargo drones is that it runs on hydrogen rather than gasoline or diesel.
2020: Transition to common road testing
At the end of May 2020, the Starlin NGO announced the passage of state road tests of its self-driving StarLine car. Thanks to this, the company washes up to start testing the drone on public roads, which Yandex is already doing.
According to the decree of the Government of the Russian Federation, in order to go to public roads, all cars with autopilot must first pass certification. Certification and assessment of unmanned vehicles is carried out by the state scientific center of FSUE NAMI, which develops and tests vehicles.
Testing on the roads of the city will allow debugging and improving the algorithms of the drone in a real road situation. This will help make the unmanned vehicle a reliable and safe vehicle, "said Boris Ivanov, head of the StarLine Unmanned Vehicle project. |
By the end of May 2020, the Starlin development team will conduct preparations for testing the car on the roads of St. Petersburg.
The tests are planned to start with Lesnoy and Bolshoi Sampsonievsky avenues, these zones are located near the office of the NGO Starline. Large-scale testing is also planned on Nevsky Prospekt. During testing , a test engineer will be in the cabin of the drone, who, if necessary, will take control.
By the end of May, the StarLine unmanned vehicle can move quite confidently in a low-load city. It defines road signs, traffic lights and pedestrians, bypasses obstacles, can be rebuilt into another lane. Earlier, the drone was already tested on closed sites, on the territory of Skolkovo, participated in the race at the Crimean bridge and became the finalist of the Winter City drone race, which took place in the winter of 2019, the NGO said.[2]
2018: Travel on the route "St. Petersburg - Kazan"
On November 22, 2018, the NPO StarLine reported that the journey of 2,500 kilometers along the route "St. Petersburg - Kazan" was made by a smart unmanned car StarLine. For 10 days, residents of 13 Russian cities took part in his test drive. In all cities through which the drone passed, test drives were organized, during which everyone could get acquainted with the capabilities of unmanned technologies.
The developers of the NPO StarLine noted that when driving along the highway, the car manages to develop a speed of about 80 km/h. This confirmed the hypothesis of the developers that an unmanned vehicle can move autonomously faster than 40 km/h - such a maximum speed was recorded during a test at the entrance to the Crimean bridge.
"On the way, the system behaved stably and reliably, successfully coping with poor marking quality, its temporary absence or the presence of a lane on only one side." ' Boris Ivanov, Project Manager "StarLine Smart Self-Driving Car" ' |
In the smart unmanned car StarLine of the 3rd level of autonomy, two automated control modes are implemented. The first is keeping in the lane. With the help of the vision system, the StarLine drone recognizes the road situation, determines the lane, and the control system calculates the route and "directs" the car along it. Sensors in this mode are video cameras and radars. The second mode of autonomous control allows the car to move in the absence of markings, in bad weather conditions, in urban traffic conditions with a large number of interference and unforeseen factors, the developers say.
To improve the operation of the automatic control system, StarLine developers create a high-precision road map with waypoints, which are generated according to the principle of operation GPSnavigator-. Vehicle location on map is determined by high-precision satellite navigation receiver, inertial navigation system and odometry. Moving along the built route, the car constantly analyzes the situation around it, recognizes obstacles and makes decisions "how to behave." To perceive the environment, a vision system, lidar and radars are used.
"On the track, the car moves without the use of a road map and a route planner. If you need to change the direction of movement - at the intersection, at exits and exits from highways and other maneuvers - the system requires operator intervention. To solve this problem, during the trip we collected data to build a road map and automate these activities. " Andrei Alekseev, project development engineer |
To ensure road safety, the operator drove an unmanned vehicle on the most difficult sections of the road and when driving around the city.
StarLine developers noted the difficulties they encountered during the test:
- Recognize the roadway.
- Process large amounts of data.
Since the journey turned out to be long, due to the large amount of data and, as a result of the high load on equipment, data processing was periodically not performed quickly enough.
In order to make the experiment as open as possible, the StarLine self-driving car development team broadcast its journey in a group on Instagram and on the official StarLine youtube channel. Hundreds of Russian Internet users joined the experiment, StarLine reported.