EP3759701A1

EP3759701A1 - Method for operating at least one automated vehicle

Info

Publication number: EP3759701A1
Application number: EP19700566.3A
Authority: EP
Inventors: Jorge Sans Sangorrin; Ingmar Berger
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2018-02-28
Filing date: 2019-01-09
Publication date: 2021-01-06
Also published as: US20210122392A1; DE102018202966A1; WO2019166143A1; US11577747B2; CN111788616A

Abstract

The invention relates to a method for operating at least one automated vehicle (200), comprising the following steps: sensorially detecting road users (200, 300, 400) with the at least one automated vehicle (200) and/or by means of a sensors in an infrastructure; determining predicted traffic routes for the road users (200, 300, 400) by means of a computer device (200) based on defined criteria; transmitting control data (D) corresponding to the predicted traffic routes to the automated vehicle (200); and operating the automated vehicle (200) according to the control data (D).

Description

description

title

Method for operating at least one automated vehicle

The invention relates to a method for operating at least one automated vehicle. The invention further relates to a method of a device for operating at least one automated vehicle. The invention further relates to a computer program product.

State of the art

The field of driver assistance systems and automated driving has become increasingly important in the automotive industry. With the introduction of standard tests, such as The Euro-NCAP 2018 test guidelines will include features such as predictive pedestrian protection and the autonomous emergency brake assist system in all vehicle categories in the foreseeable future. In addition, the equipment quota of vehicles with driver assistance functions for combined lateral and longitudinal guidance will increase steadily in the future. These will have a high level of availability, especially on highways and well-developed highways, and will enable relaxed driving outside of urban areas.

Sensors such as radar, ultrasound, video cameras, lidar, etc. capture the vehicle environment. With the help of this environment detection, road users and infrastructure can be recognized in the environment of the vehicle. Other technologies, such as The navigation system and GPS-based localization systems are already standard features in many new vehicles.

Another class of driver assistance systems are the so-called parcassistance functions, which are currently available in first production vehicles. Some of them already allow you to control the parking process via a special app on your smartphone. However, the detection capability of vehicle-mounted sensors is limited in terms of environmental perception in complex scenarios, such as those often found in urban scenarios. This can lead to a limited availability of highly automated systems. For example, concealments and / or nearby objects in the surroundings of the vehicle can lead to a limited environment detection by means of the aforementioned sensors. A large number of objects arranged within a limited space can further restrict the surroundings identification.

One way to gain information about the environment is to develop vehicle-to-vehicle and vehicle-to-infrastructure communication technologies that are in the process of being developed and will be standard equipment in all vehicles in the medium to long term. In the near future, communication technologies will even enable real-time transmission of data.

At the moment, LTE is the worldwide standard for mobile communications. It is already available worldwide in many regions and allows transmission rates of up to 300 MBit / s or, in its latest version, 4G / LTE + up to 4,000 MBit / s. At the same time, the successor 5G is already in the starting blocks. This will enable real-time mobile communication with transfer rates up to 10,000 MBIt / s and with latencies <1 ms.

The smartphones, which are now equipped as standard with these technologies, have replaced the desired mobile phones within a very short time since the first iPhones® appeared. Regardless of age, virtually every adult today owns a smartphone. Even cheaper models are now equipped with GPS, Bluetooth and LTE technology.

Driver assistance systems and automated driving systems perform calculations to estimate the development (prediction) of the traffic situation. On the basis of this estimate, the system decides to drive the appropriate behavior of the vehicle in the form of, for example, a trajectory. To model the predic- tion of the situation for the individual road users, behavioral models are stored. These behavioral models contain information about the movement possibilities of different object types and evaluate their interaction with other road users and the shared infrastructure. For example, a vehicle can move much faster than pedestrians, but a pedestrian can stop almost immediately. Every prediction of the situation becomes more uncertain as the prediction time increases. The uncertainty here is essentially due to two factors:

On the one hand, the information about the other road users detected via the vehicle-specific surroundings sensor can be noisy and / or incomplete.

- On the other hand, a prediction based on behavioral models is only possible with a limited foresight into the future. This is due to the number of possible behavioral options that a road user can perceive within a very short time.

In conventional systems, all these calculations take place without any further information, which can not be detected by the built-in sensors.

EP 2 911 926 B1 discloses a method in which vehicles collect environmental information, wherein an exchange of the environmental information is provided across several vehicles.

DE 102 10 546 A1 discloses a method for automatic vehicle guidance, in which infrastructure data are transmitted wirelessly to the vehicle and commands for the vehicle guidance are calculated on the basis of the infrastructure data. It is provided that the infrastructure data for at least one immediately preceding section of the route is loaded into a vehicle's own memory, that the current position of the vehicle is continuously determined with a precise positioning system and that the commands are based on the position data and the stored infrastructure data be calculated.

Disclosure of the invention

An object of the invention is to provide an alternative method of operating at least one automated vehicle. The object is achieved according to a first aspect with a method for operating at least one automated vehicle, comprising the steps:

Sensory detection of road users with the at least one automated vehicle and / or by means of sensors in an infrastructure;

Determination of predicated traffic routes for the road users by means of a computing device on the basis of defined criteria;

Transmitting control data corresponding to the predicted traffic route to the automated vehicle; and

Operating the automated vehicle according to the control data.

Advantageously, no complex sensor technology in the automated vehicle is required in this way, because important information or data about the environment of the vehicle, so to speak "in the network". This realizes a "central intelligence" that does not lie on individual vehicles. In this way, a functionality with many services for operating the automated vehicle is advantageously provided, as a result of which an at least partial operation of the automated vehicle by means of the central intelligence can be realized, in particular in urban areas.

According to a second aspect, the object is achieved with a device for operating at least one automated vehicle, comprising:

a sensor device for detecting road users with the at least one automated vehicle and / or by means of sensors in an infrastructure;

a computing device for calculating predicated traffic routes for the road users by means of a computing device based on defined criteria; and

a transmission device for transmitting control data corresponding to the predicted traffic route to the automated vehicle.

Advantageous developments of the method are the subject of dependent claims.

An advantageous development of the method provides that the control data are designed as trajectory data. In this way, the automated Be managed by a central authority, which has an optimized overall view of the situation and thus can optimally control the traffic situation.

A further advantageous development of the method provides that the environment is sensed by the road users, wherein the sensor data are transmitted to the computing device. In this way, the central computing device is advantageously enabled to use a large amount of data of the traffic participants for an optimized determination of a predictive traffic model.

A further advantageous development of the method is characterized in that control data for defined partial routes are transmitted from the computing device to the automated vehicle. In this way, the method can be applied regionally specifically and in particular in regions with particularly high traffic volumes.

A further advantageous development of the method provides that the method in the automated vehicle is controlled by means of a man-machine interface. This supports convenient control of the method, for example with a touch screen of the automated vehicle and / or a touch screen of a mobile phone. The user can thereby be shown graphically that a network is available with the above-explained control options. By an appropriate input by means of the human-machine interface, the driver can then accept that the vehicle is automatically controlled with the aid of said network.

A further advantageous development of the method provides that a digital map is used by the computing device to determine the predicated traffic routes for the traffic participants. In this way, circumstances for the computing device can be provided, so that the determination of the predicated traffic model is performed even more accurately.

A further advantageous development of the method provides that at least one traffic infrastructure device is controlled by means of the computing device. In this way, for example, traffic light circuits and / or switching barrier systems to better manage traffic flows.

The invention will be described in detail below with further features and advantages with reference to two figures. Above all, the figures are intended to clarify the principles essential to the invention.

Disclosed method features result analogously from corresponding disclosed device features and vice versa. This means, in particular, that features, technical advantages and embodiments relating to the method result analogously from corresponding designs, features and advantages of the device and vice versa.

In the figures shows:

Fig. 1 is a schematic representation of a scenario for carrying out the proposed method; and

2 shows a basic sequence of an embodiment of the proposed method for directing a traffic flow.

Description of embodiments

In the following, the term automated motor vehicle is used interchangeably in the terms partially automated motor vehicle, autonomous motor vehicle and partially autonomous motor vehicle.

An automated or autonomous vehicle is a vehicle that manages without a driver. The vehicle drives autonomously, for example, by automatically recognizing a road course, other road users or obstacles and calculating corresponding control commands in the vehicle and forwarding these to actuators in the vehicle, whereby the driving course of the vehicle is correctly influenced. The driver is no longer involved in a fully autonomous vehicle on the ride.

Vehicle-to-vehicle communication (Car2Car or C2C) is understood to mean the exchange of information and data between motor vehicles. The vehicles in question collect data, such as ABS interventions, steering angle, position, direction, speed, etc. and transmit this data to other road users via radio (for example via WLAN, UMTS, etc.). The aim is to increase the "sight of the driver" by electronic means. Vehicle-to-infrastructure communication (C2I) is understood as the exchange of data between a vehicle and the surrounding infrastructure (eg traffic lights).

The technologies mentioned are based on the interaction of sensors of different traffic partners and use the latest methods of communication technology to exchange this information. For this purpose, autonomous or semi-autonomous vehicles exchange data with one another by means of car-to-car communication systems.

Nowadays, the calculations for automated driving are made exclusively on the automated vehicle. High-performance computer cores are usually used for this purpose in order to be able to process the high volume of data required in a timely manner. However, such computing systems are relatively expensive, take up a lot of space in the vehicle and can reach their limits due to the steadily growing number of sensor systems in the vehicle with regard to the computational load.

For offloading computations to the cloud, it may be beneficial to provide an image of a computational requirement for automated driving ("artificial intelligence") in the cloud. In this case, the image of the artificial intelligence in the cloud can correspond exactly to the artificial intelligence on the automated vehicle or be defined differently to the artificial intelligence of the automated vehicle. An identical image of the artificial intelligence of the automated vehicle in the cloud has the significant advantage that with the same input data, calculation results of the cloud do not differ from calculation results on the automated vehicle. Advantageously, an image of the artificial intelligence in the cloud for entire vehicle variants or vehicle families.

1 shows a highly schematized, exemplary scenario of a proposed method for operating at least one automated vehicle 200. One recognizes in the environment a preferably permanently installed sensor device 10 (eg in the form of a camera, radar, etc.), with which an environment scenario detected which includes, for example, the automated vehicle 200, pedestrian 300, cyclist 400, animals (not shown), and so forth. For the sake of simplicity, only one of each of the aforementioned road users is shown, but it goes without saying that the method can be used with a large number of automated vehicles 200, pedestrians 300, cyclists 400, etc.

The sensor device 10 transmits the data to a central computing device 20, which calculates from the determined sensor data a predictive traffic model with a predicated traffic route for the at least one automated vehicle 200. For the creation of the predictive traffic model defined criteria can be used, for example a consideration of congestion situations, accident situations, accumulation of road users, weather conditions, etc.

By means of a transmission device 21 arranged inside or outside the computing device 20, control data D, e.g. wirelessly transmitted to the automated vehicle 200 in the form of trajectory data. For this purpose, a high-speed or real-time data connection is preferably provided, with which a real-time communication with the automated vehicle 20 is performed. A controller 210 (eg, in the form of controllers, actuators, etc.) disposed within the vehicle 200 may control the automated vehicle 200 in accordance with the transmitted trajectory data. In one variant, it is also conceivable that, in addition to the trajectory data, other types of control data, e.g. Control data for the actuators to the automated vehicle 200 are transmitted. It is also conceivable that by means of the control data D infrastructures (for example traffic lights, barriers, etc.) are switched.

Advantageously, in this way a device 100 is provided for the autonomous control of connected road users, preferably on main traffic arteries, in order to enable drivers of the automated vehicles 200 to travel relaxed within urban areas. For this purpose, the automated vehicles 200 advantageously only have to have a small amount of equipment. In this case, the device 100 provides closed-loop control commands to automated vehicles 200 in real time, whereby the device 100 does not provide any information about the surroundings to be provided. As a result, so to speak, the sensors for all road users "centrally" designed.

In an advantageous development of the device 100, it can be provided that sensor data acquired by the road users is transmitted wirelessly to the computing device 20, wherein these data are then used by the computing device 20 for improved determination of the predictive traffic model.

The transmission of said data to the computing device 20 may preferably be based on known vehicle-to-vehicle and / or vehicle-to-infrastructure communication.

Advantageously, an uncertainty in the prediction of the traffic situation can be significantly reduced in this way. The prediction time and a certainty about the prediction can advantageously be significantly increased, as a result of which a degree of automation of the entire system is increased, as a result of which an automated operation of the automated vehicle 200 is significantly longer and more frequently available. This is particularly useful in an urban environment with a high traffic density and an at least occasionally very high number of road users.

In the further variant of the device 100, it can be provided that the computing device 20 uses a digital map for determining the predicted traffic model, in which data of the environment are stored. As a result, the determination of the predictive traffic model by the computing device 20 can be carried out even better and faster.

Advantageously, it is possible for a driver of the automated vehicle 200 to establish a communication connection to the device 100 via a human-machine interface, and to use the services provided in this way.

In an exemplary scenario, a driver of the automated vehicle 200 is traveling to a city with greater traffic. By means of a man-machine interface (eg touch screen), the driver receives the message that services of the device 100 are available. The driver then accepts the offered services on the touch screen, whereby the vehicle 200 communicates relevant data (eg, type, extent, navigation destination, etc.) to the central computing device 20.

The computing device 20 locates the automated vehicle 200 within the stored digital map and determines the trajectory that the vehicle 200 should drive in consideration of the predicted traffic situation to achieve the navigation destination. Since the central computing device 20 also determines the trajectories for all other vehicles, freedom from accidents is advantageously supported. The trajectory (e.g., in the form of position and speed of multiple times in the future) is then communicated to the automated vehicle 200. The vehicle 200 is equipped with corresponding control devices in order to start the trajectory on the basis of the transmitted trajectory data.

In an advantageous variant, environment sensors installed in the vehicle 200 or in infrastructure devices can be used to provide feedback to the central computing device 20 that no collision takes place. Sensors in the infrastructures may also be used to report that road users without vehicle-to-infrastructure equipment are in the observed area.

Advantageously, the proposed device 100 may also be used to provide paid services (e.g., by means of a software app, receiver / transmitter hardware, tariffs, etc.).

It can preferably be provided that the device 100 does not cover an entire urban area, but only main traffic arteries with a high traffic volume. In a further variant, it can also be provided that a smartphone arranged inside or outside the automated vehicle 200 carries out the communication to the computing device 20. For example, it is also possible for pedestrians or cyclists to be recognized by the sensor device 10 with a smartphone, in which case infrastructure devices (eg traffic lights, barriers, etc.) are appropriately switched by the computing device 20. Advantageously, it may also be provided to book additional services via the device 100, such as a parking space, services at the destination, etc.

Advantageously, it may be provided that the device 100 is at least simply redundant, so that an increased level of safety of the operation is supported.

2 shows a basic sequence of an embodiment of the method according to the invention.

In a step 500, a sensory detection of road users 200, 300, 400 is performed with the at least one automated vehicle 200 and / or by means of sensors in an infrastructure.

In a step 510, predicated traffic routes for the road users 200, 300, 400 are determined by means of a computing device 200 on the basis of defined criteria.

In a step 520, transmission of control data D corresponding to the predicated traffic route to the automated vehicle 200 is performed.

In a step 530, an operation of the automated vehicle 200 is performed in accordance with the control data D.

Advantageously, the proposed method can be implemented by means of a software program running on the computing device 20, whereby a simple adaptability of the method is supported.

One skilled in the art will suitably modify and / or combine the features of the invention without departing from the gist of the invention.

Claims

claims

A method of operating at least one automated vehicle (200), comprising the steps of:

Sensory detection of road users (200, 300, 400) with the at least one automated vehicle (200) and / or by means of sensors in an infrastructure;

Determining predicated traffic routes for the road users (200, 300, 400) by means of a computing device (200) on the basis of defined criteria;

Transmitting control data (D) corresponding to the predicted traffic route to the automated vehicle (200); and

Operating the automated vehicle (200) according to the control data (D).

2. The method of claim 1, wherein the control data (D) are designed as trajectory data.

3. The method of claim 1 or 2, wherein by the road users (200, 300, 400) the environment is detected by sensors, wherein the sensor data to the computing device (200) are transmitted.

4. The method according to any one of the preceding claims, wherein from the computing device (20) to the automated vehicle (200) control data (D) for defined sub-routes are transmitted.

5. The method according to any one of the preceding claims, wherein the method in the automated vehicle (200) is controlled by means of a man-machine interface.

6. Method according to one of the preceding claims, wherein a digital map is used by the computing device (200) to determine the predicated traffic routes for the road users (200, 300, 400).

7. The method according to any one of the preceding claims, wherein by means of the computing device (20) at least one traffic infrastructure device is controlled.

8. An apparatus (100) for operating at least one automated vehicle (200), comprising:

a sensor device (10) for detecting road users (200, 300, 400) with the at least one automated vehicle (200) and / or by means of sensors in an infrastructure;

a computing device (20) for calculating predicated traffic routes for the road users (200, 300, 400) by means of a computing device (200) on the basis of defined criteria; and

a transmission device (21) for transmitting control data (D) corresponding to the predicted traffic route to the automated vehicle (200).

9. Use of a device (100) according to claim 8 for an automated vehicle (200) in an urban environment.

10. Computer program product with program code means for carrying out the method according to one of claims 1 to 7, when it runs on an electronic computing device (20) or is stored on a computer-readable data carrier.