Motor vehicle and method for motion planning of an autonomous motor vehicle

Abstract

The invention relates to a method for motion planning of an autonomous motor vehicle (10) in situations where a person (16) is moving near another vehicle (18). The method comprises the following steps: acquiring (S10) environmental data by a sensor device (12) of the motor vehicle (10), wherein the environmental data includes at least the person (16) and the other vehicle (18); determining (S12) an activity being performed on the other vehicle (18) based on the environmental data by a control device (14) of the motor vehicle (10); determining (S14) a predetermined, situation-specific movement space (20) of the person (16) by the control device (14) depending on the determined activity; and providing (S16) the motion plan for controlling the autonomous motor vehicle (10) by the control device (14), in which the autonomous motor vehicle (10) navigates around the situation-specific movement space (20) of the person (16).Furthermore, the invention comprises a motor vehicle designed to carry out the method.

Description

[0001] The invention relates to a method for motion planning of an autonomous motor vehicle in situations where a person is moving in the vicinity of another motor vehicle. The invention further relates to a motor vehicle configured to carry out the method.

[0002] When planning the movements of autonomous vehicles, the movement patterns of people, such as pedestrians, pose a problem because they cannot always be predicted. Particularly in confined spaces, such as parking lots or gas stations, a person's path may overlap with a planned trajectory of the autonomous vehicle, requiring it to be aborted and recalculated. This can lead to delays and increased resource consumption for the autonomous vehicle.

[0003] US Patent 10,304,335 B2 discloses methods, systems, and computer program products for detecting available parking spaces in a parking environment. Radar systems are used to collect data about the parking environment. The radar data is provided as input to a neural network model. Algorithms of the neural network can be trained to detect parked vehicles and conflict data, which includes debris, shopping carts, streetlights, traffic signs, and pedestrians. The neural network model evaluates the radar data to estimate parking space boundaries and approximates these boundaries as splines. The neural network model outputs the spline estimates to a vehicle computer system. The vehicle computer system uses the spline estimates to detect available parking spaces. The spline estimates are updated as the vehicle navigates to the parking environment.

[0004] From DE 10 2014 224 101 A1 a method for operating a parking lot is known, wherein traffic of several vehicles within the parking lot is coordinated in such a way that a predetermined minimum distance between two vehicles is maintained in order to avoid collisions between vehicles.

[0005] A parking assistance device is known from US patent 11,840,216 B2. The parking assistance device comprises a parking space recognizer that detects a plurality of parking spaces in which a vehicle is parked, a parking space candidate calculator that calculates one or more parking space candidates from the plurality of parking spaces based on an evaluation criterion of a driver and information that indicates a route from a moving object around the vehicle, a display controller that displays the one or more parking spaces on a display, and a driving controller that controls the driving of the vehicle to the parking space candidate that has been selected from the one or more parking space candidates.

[0006] From DE 10 2019 203 334 A1, a method is known for initiating a reaction of a first vehicle to a person in the vicinity of a second vehicle, in which at least one person on a roadway traveled by the first vehicle and at least one second vehicle are detected, wherein measurement data determined by at least one sensor unit are received by a control unit, the control unit performs a classification and registers the evaluated data as a person and as a second vehicle, movement vectors of the person are determined and expected movement of the person is calculated, a position and width of vehicle doors of the second vehicle are determined or estimated on the basis of the measurement data, a probability of a vehicle door being opened by a person is calculated, wherein a reaction of the first vehicle is initiated by the control unit depending on the calculated probability.

[0007] German patent DE 10 2021 001 519 A1 discloses a method for avoiding collisions when vehicle doors are opening. This method uses a communication system of an ego-vehicle to establish an ad-hoc connection with a vehicle parked at the roadside. The ego-vehicle receives a door opening signal emitted by the other vehicle due to an initiated or anticipated opening of a roadside vehicle door. Furthermore, the ego-vehicle's trajectory is checked for collisions with the hazardous area that would arise if the vehicle door were to open. In the event of a potential collision, a warning is issued and / or the ego-vehicle's movements are intervened.

[0008] The object of the invention is to improve the motion planning of an autonomous motor vehicle in situations in which a person is moving near another vehicle.

[0009] This problem is solved by the independent patent claims. Advantageous embodiments of the invention are disclosed in the dependent patent claims, the present description, and the figures.

[0010] The invention is based on the idea that, based on a detected activity being carried out on the other vehicle, a person's movement space is limited to a specific movement space, whereby the movement planning excludes this movement space when determining the movement of the autonomous vehicle.

[0011] One aspect of the invention relates to a method for motion planning of an autonomous vehicle in situations where a person is moving near another vehicle. The method comprises the following steps: acquiring environmental data by a sensor device of the vehicle, wherein the environmental data includes at least the person and the other vehicle; determining an activity being performed on the other vehicle based on the environmental data by a control device of the vehicle; and determining a predetermined, situation-specific movement space for the person by the control device, depending on the determined activity. Furthermore, the method comprises providing the motion plan to the control device for the autonomous vehicle to navigate around the situation-specific movement space of the person.

[0012] In other words, a sensor device in the vehicle can detect that a person is near another vehicle and performing an action there. The sensor device can include imaging sensors, such as a camera and / or lidar, that capture the person and the other vehicle as environmental data. This environmental data can then be evaluated by a control device. The control device can, for example, include a computer trained to evaluate the environmental data. The evaluation of the environmental data can be performed using image recognition algorithms that can detect an action by the person and / or the state of the other vehicle. In particular, artificial intelligence can be used to evaluate the environmental data, and this AI can be trained using image data of activities that can be performed on vehicles.For example, artificial intelligence can be trained to recognize boarding and alighting processes, loading and / or unloading and / or refueling, so that these activities can be recognized from the environmental data.

[0013] The control device can, in particular, recognize an activity based on actions taken by the person and / or on the condition of the other vehicle, such as open doors, an open trunk, and / or an open fuel filler cap. Based on these indicators, the control device can narrow down and define the activity. Preferably, predefined map data can also be used to restrict activities. For example, if it is detected that the vehicle is in a supermarket parking lot, other activities, such as refueling, can be excluded. Conversely, the vehicle's location at a gas station can indicate refueling as the activity.

[0014] Once the activity has been identified, a predetermined, situation-specific movement space for the person around the other vehicle can be defined, corresponding to the identified activity. This situation-specific movement space can be defined based on a movement pattern typical for the activity. For example, different activities or actions of individuals can be analyzed, particularly using predetermined movement data, thus yielding the respective situation-specific movement space for each activity. This movement space can encompass an area around the person's current position or a defined area at the position of the other vehicle, within which a possible movement of the person is anticipated.

[0015] Finally, the control device can perform motion planning for the autonomous vehicle, excluding the situation-specific movement space of the person during motion planning. This means that the autonomous vehicle's movement is not planned through the situation-specific movement space.

[0016] The invention offers the advantage of providing an accurate prediction of a movement area, thus improving the spatial definition for the autonomous vehicle's motion planning. This reduces the frequency with which motion plans need to be discarded, as it becomes less likely that a person will cross the vehicle's trajectory, necessitating the aborting and recalculation of the autonomous vehicle's movement. In particular, this saves time and computational resources, resulting in improved motion planning.

[0017] Furthermore, according to the invention, the activity is the refueling or charging of the other vehicle, and this activity is detected by means of an open fuel filler cap or charging port of the other vehicle. That is, the activity can be the refueling of the other vehicle with fuel or the charging of a vehicle battery of the other vehicle, and this activity can be detected from the environmental data by means of an open fuel filler cap or charging port. Thus, the situation-specific movement space for this activity can be defined, in particular, at a specific position of the fuel filler cap or charging port.

[0018] The invention also includes embodiments that offer additional advantages.

[0019] One implementation involves continuously monitoring whether the person remains within the situation-specific movement space. The autonomous vehicle only maintains its movement plan if the person is within this space. In other words, the sensor device continuously monitors the area of ​​the situation-specific movement space after the movement plan has been executed. If the person remains within the situation-specific movement space, the autonomous vehicle can continue its journey based on the movement plan. However, if it is detected that the person is leaving the situation-specific movement space, the movement plan can be aborted, for example, by the vehicle stopping.

[0020] In particular, it can be provided that the autonomous vehicle's motion planning is re-run based on a motion prediction of the person if the person leaves the situation-specific movement area. This means that after the person leaves the situation-specific movement area, a motion prediction can be made. For example, the person's direction and speed can be extrapolated to estimate how the person will move outside the situation-specific movement area. Based on this motion prediction, the motion planning can then be re-run. This offers the advantage of increased safety.

[0021] Another embodiment provides that the activity is entering or exiting the other vehicle, with this activity being detected by an open vehicle door of the other vehicle. This means that the sensor device can detect, for example, an open vehicle door or the opening of the other vehicle door, whereby the situation-specific movement space can be defined within an area around the open vehicle door. This movement space is preferably adapted to the activity of entering and exiting the vehicle.

[0022] Another configuration involves the activity of loading or unloading the trunk of another vehicle, with this activity being detected by an open trunk lid. This means the sensor device can detect the open trunk lid of the other vehicle, particularly when a person is near it, thus enabling the detection and identification of the loading or unloading activity. Specifically, the appropriate situation-specific movement area for this activity can be defined within a region around the trunk, allowing the vehicle to navigate around this area during movement planning. This offers the advantage of improved detection of the movement area required for this activity.

[0023] Another embodiment provides that the activity is the loading or unloading of the trunk of the other vehicle, whereby this activity is recognized by the presence of a shopping cart and / or the loading or unloading of the trunk by the person. This means that, alternatively or additionally to the open trunk lid, it can also be recognized whether a shopping cart is located near the other vehicle, preferably in an area of ​​the trunk, thereby restricting the activity, in particular to the loading and / or unloading of the other vehicle.

[0024] Alternatively or additionally, loading and / or unloading operations can also be detected from the environmental data, allowing conclusions to be drawn about this situation. This further improves the detection of this situation.

[0025] In a further advantageous embodiment, the refueling or charging of the other vehicle is also detected by identifying a refueling or charging point from the environmental data and / or from predefined map data. This means that the sensor device can recognize that the other vehicle is at a refueling or charging point by analyzing the environmental data for characteristics of such points. Alternatively or additionally, predefined map data can be provided from which the position of the other vehicle can be determined. Furthermore, a comparison with infrastructure features can be made using the map data, for example, to determine whether the other vehicle is at a refueling or charging point. This allows for a more precise definition of the activity being carried out.

[0026] In a further advantageous embodiment, the movement space during refueling or charging of the other vehicle is also determined based on an identified action of the person, which includes refueling or charging the other vehicle and paying at a payment point. This means that typical action patterns of the person can be recognized that indicate refueling or charging. For example, holding a fuel nozzle or charging plug can be identified as an action of the person, thus allowing refueling or charging to be determined as an activity. Additionally, a route to a payment point can also be defined as a situation-specific movement space, whereby payment points can be determined from environmental data and / or map data.

[0027] For use cases or application situations that may arise during the procedure and are not explicitly described here, it may be provided that, according to the procedure, an error message and / or a request for user feedback is issued and / or a default setting and / or a predetermined initial state is set.

[0028] Another aspect of the invention relates to a motor vehicle, in particular an autonomous motor vehicle, comprising at least one sensor device and one control device, wherein the motor vehicle is configured by means of the sensor device and the control device to perform a method according to the preceding aspect. This offers the same advantages and possibilities for variation as the method.

[0029] The motor vehicle according to the invention is preferably designed as a motor vehicle, in particular as a passenger car or truck, or as a passenger bus or motorcycle.

[0030] The invention also includes the control device for the motor vehicle. The control device can comprise a data processing device or a processor unit configured to carry out an embodiment of the method according to the invention. For this purpose, the processor unit can comprise at least one microprocessor and / or at least one microcontroller and / or at least one FPGA (Field Programmable Gate Array) and / or at least one DSP (Digital Signal Processor). In particular, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or an NPU (Neural Processing Unit) can be used as the microprocessor. Furthermore, the processor unit can comprise program code configured to carry out the embodiment of the method according to the invention when executed by the processor unit. The program code can be stored in a data memory of the processor unit.The processor setup can be based on at least one circuit board and / or at least one SoC (System on Chip).

[0031] The invention also includes further developments of the motor vehicle according to the invention, which have features as already described in connection with the further developments of the method according to the invention. For this reason, the corresponding further developments of the motor vehicle according to the invention are not described again here.

[0032] As a further solution, the invention also includes a computer-readable storage medium comprising program code which, when executed by a computer or a computer network, causes it to execute an embodiment of the method according to the invention. The storage medium can be provided at least partially as a non-volatile data storage medium (e.g., as flash memory and / or as an SSD - solid state drive) and / or at least partially as a volatile data storage medium (e.g., as RAM - random access memory). The storage medium can be located within the computer or computer network. However, the storage medium can also be operated, for example, as an app store server and / or cloud server on the internet. The computer or computer network can provide a processor circuit with, for example, at least one microprocessor.The program code can be provided as binary code and / or as assembly code and / or as source code of a programming language (e.g. C) and / or as a program script (e.g. Python).

[0033] The invention also includes combinations of the features of the described embodiments. The invention therefore also includes realizations that each exhibit a combination of the features of several of the described embodiments, provided that the embodiments have not been described as mutually exclusive.

[0034] The following are exemplary embodiments of the invention described. This is illustrated by: Fig. 1 a motor vehicle according to an exemplary embodiment; Fig. 2 an exemplary situation with the motor vehicle; Fig. 3 a process diagram according to an exemplary embodiment.

[0035] The exemplary embodiments described below are preferred embodiments of the invention. In these exemplary embodiments, the described components each represent individual features of the invention, which can be considered independently of one another and each further develops the invention independently. Therefore, the disclosure is intended to include combinations of features of the embodiments other than those shown. Furthermore, the described embodiments can also be supplemented by further features of the invention already described.

[0036] In the figures, identical reference symbols denote functionally equivalent elements.

[0037] In Fig. Figure 1 shows a schematic representation of a motor vehicle 10, which may in particular be configured as an autonomous motor vehicle 10. The motor vehicle 10 may include a sensor device 12 and a control device 14.

[0038] The sensor device 12 can include at least one camera designed to capture environmental data of the motor vehicle 10. This environmental data can be used for movement planning of the autonomous motor vehicle 10 and can, for example, record other vehicles and / or persons in the vicinity of the motor vehicle 10.

[0039] For evaluation purposes, the acquired environmental data can be transmitted to the control device 14, which can be configured to analyze the environmental data and perform motion planning based on the analysis. The control device 14 can comprise a processor or microprocessor, in particular a graphics processor, which preferably performs the motion planning of the autonomous vehicle 10 using a sampling-based algorithm.

[0040] During motion planning, the movement of the autonomous vehicle 10 can be restricted by people or pedestrians, especially in confined spaces such as parking lots or gas stations. For example, if a pedestrian crosses a planned motion trajectory of the vehicle 10, the trajectory cannot be continued, and it must be recalculated, which costs time and resources. Therefore, it may be possible to exclude areas where a person's movement is expected before motion planning. This can be achieved, in particular, by detecting situations where a person is moving near another vehicle, and by estimating the person's movement space based on the detected situation or activity.

[0041] This is illustrated by an exemplary situation involving an autonomous vehicle in the Fig. Figure 2 illustrates this situation. In this scenario, the motor vehicle 10 can move in a parking lot, for example a supermarket parking lot, while the sensor device 12 records the environmental data. In particular, the sensor device 12 can detect the person 16 at another vehicle 18.

[0042] Based on the recorded environmental data, the control device 14 can determine the activity that person 16 is performing at the other vehicle 18. In this situation, person 16 might, for example, be loading the trunk of vehicle 18. This activity can be determined, in particular, by detecting an open trunk lid of vehicle 18 from the environmental data. Alternatively or additionally, the loading action of person 16 in the trunk can be recognized from the environmental data, in particular a video stream of the environmental data, for example, from a movement pattern during loading.

[0043] Once it has been established that the identified activity is loading the trunk, a predetermined, situation-specific movement space 20 can be defined, particularly around the area of ​​the trunk or the current location of person 16. This situation-specific movement space 20 can designate an area within which person 16 typically moves while loading the trunk. This area can be predetermined from predetermined analyses of movement patterns during trunk loading.

[0044] Alternatively or additionally, the activity, i.e. the loading of the trunk, can also be identified by other features, such as a shopping cart 22 near the trunk and / or map data, which shows, for example, that the vehicle 10 is moving in a supermarket car park.

[0045] Once the situation-specific movement space 20 of person 16 has been determined, the movement planning can be carried out by the control device 14, whereby a driving trajectory 24 of the motor vehicle 10 can be determined that bypasses the determined situation-specific movement space 20. Consequently, the probability that the driving trajectory 24 is disturbed or crossed by a movement of person 16 can be minimized, thus reducing the number of interruptions or recalculations of the driving trajectory 24.

[0046] Preferably, while traveling along the trajectory 24, it can be continuously checked whether the person 16 is still within the situation-specific movement space 20, whereby the autonomous vehicle 10 only maintains the movement plan if the person 16 is within the situation-specific movement space 20. If the person 16 leaves the situation-specific movement space 20, it can be provided that the movement plan is re-executed based on a movement prediction, for example, an extrapolation of the direction of movement of the person 16.

[0047] In addition to this example of loading the trunk of vehicle 18, other activities, each with its own situation-specific movement space, can be detected. For example, an activity can also be detected as entering or exiting vehicle 18, which can be determined, for instance, by an open vehicle door, or as refueling or charging the other vehicle 18. Refueling or charging vehicle 18 can be detected, for example, by an open fuel filler or charging port, an action by person 16 such as holding a fuel nozzle or charging plug, and / or by identifying the refueling or charging station from environmental data and / or predefined map data.

[0048] In Fig.Figure 3 shows a schematic process diagram for the motion planning of an autonomous motor vehicle 10 in situations in which a person 16 is moving near another vehicle 18.

[0049] In step S10, environmental data can be acquired by a sensor device 12 of the motor vehicle 10, wherein the environmental data includes at least the person 16 and the other vehicle 18.

[0050] In step S12, an activity can be determined that is carried out on the other vehicle 18, in particular based on the environmental data by a control device 14 of the motor vehicle.

[0051] In step S14, a predetermined situation-specific movement space 20 of the person 16 can be determined by the control device 14, whereby the situation-specific movement space 20 is determined depending on the identified activity.

[0052] Finally, in step S16, the motion planning for controlling the autonomous vehicle 10 can be provided by the control device 14, in which the autonomous vehicle 10 avoids the situation-specific movement space 20 of the person 16.

[0053] Overall, the examples show how automated or assisted motion planning can be provided to avoid pedestrians moving near the vehicle.

Claims

[1] Method for motion planning of an autonomous motor vehicle (10) in situations in which a person (16) is moving near another vehicle (18), comprising the steps: - Acquisition (S10) of environmental data by a sensor device (12) of the motor vehicle (10), wherein the environmental data includes at least the person (16) and the other vehicle (18); - Determining (S12) an activity being carried out on the other vehicle (18) using environmental data by a control device (14) of the motor vehicle (10); - Determining (S14) a predetermined situation-specific movement space (20) of the person (16) by the control device (14) depending on the detected activity; - Provision (S16) of the motion planning for controlling the autonomous vehicle (10) by the control device (14), in which the autonomous vehicle (10) avoids the situation-specific movement space (20) of the person (16). - wherein the activity is refueling or charging the other vehicle (18), wherein this activity is detected by reference to an open fuel tank or charging port of the other vehicle (18). [2] Method according to claim 1, wherein it is continuously checked whether the person (16) is still within the situation-specific movement space (20), wherein the autonomous motor vehicle (10) only maintains the movement planning if the person (16) is within the situation-specific movement space (20). [3] Method according to claim 2, wherein the movement planning of the autonomous motor vehicle (10) is re-executed based on a movement prediction of the person (16) if the person leaves the situation-specific movement space (20). [4] Method according to one of the preceding claims, wherein the refueling or charging of the other vehicle (18) is further detected by identifying a refueling or charging point from the environmental data and / or from predefined map data. [5] Method according to one of the preceding claims, wherein the movement space during refueling or charging of the other vehicle (18) is further determined on the basis of an identified action of the person (16) which includes refueling or charging the other vehicle (18) and paying at a payment point. [6] Motor vehicle (10) comprising at least a sensor device (12) and a control device (14), wherein the motor vehicle (10) is configured by means of the sensor device (12) and control device (14) to perform a method according to one of the preceding claims.

Images