Computer-implemented method and computer program for the functional safe localization of a vehicle on a roadway and an automated vehicle.

Abstract

A computer-implemented method for the functional safe localization of a vehicle (10) on a roadway (1) on, at, or in which reference elements (2) are arranged, wherein the vehicle (10) comprises sensors (3) that detect fields of the reference elements (2), the method comprising the steps: reading kinematic signals and estimating a change in position (SafetyEME) of the vehicle (10) via an interface to an on-board network of the vehicle (10) (V3); after passing at least two successive reference elements (2), correcting a position deviation of the vehicle (10) resulting from the estimated change in position (SafetyEME) by adjusting (MMS LOC) the last two detections of the fields detected by the sensors (3) of the vehicle (10) to the respective measured positions of the reference elements (2) in the reference map (Map) and obtaining a pose of the vehicle (10) within the reference map (Map) (V4);from the respective pose (MMS LOC), the last at least two sensory detections of the reference elements (2) and / or the current estimated position change (SafetyEME), predict (SafetyLOC) the next sensory detection based on the reference map (V5); if two consecutive predicted detections are not detected sequentially, mark the position of the vehicle (10) as invalid and initiate a stop maneuver of the vehicle (10) based on the last valid pose (V6).

Description

[0001] The invention relates to a computer-implemented method and a computer program for the functionally safe localization of a vehicle on a roadway. Furthermore, the invention relates to an automated vehicle.

[0002] The following definitions, descriptions and statements retain their respective meaning for and apply to the entire disclosed subject matter of the invention.

[0003] EP 2 360 544 A1 discloses a system for determining the position of a vehicle by means of magnetic marking elements installed in a roadway and magnetic sensors arranged on the vehicle.

[0004] US 11 604 476 B1 discloses a guidance system for a vehicle, wherein the vehicle is operated in a first operating mode in a first road segment comprising reference elements and changes to a second operating mode when the vehicle travels on a second road segment not comprising reference elements.

[0005] DE 10 2019 215 658 B3 discloses a system for controlling automated driving functions for an automated vehicle, comprising several marking elements arranged in a road surface, several sensors that can be arranged on the vehicle and measure the field strengths of individual marking elements, and a computing unit that determines the position of the sensors based on the field strengths measured by the sensors and the position of the vehicle relative to the marking elements based on the arrangement of the sensors.wherein the system comprises a control unit for automated driving and the computing unit, during automated operation, controls a trajectory for the vehicle determined by the control unit depending on signals from the vehicle's environmental sensors, taking into account the vehicle's position relative to the marker elements and vehicle data, and in the event of a deviation, overrides or deactivates control signals from the control unit for actuators for longitudinal and / or lateral guidance of the vehicle.

[0006] The object of the invention was to provide a way to locate a vehicle in a functionally safe manner, in particular according to Automotive Safety Integrity Level D, abbreviated ASIL D.

[0007] The subject matter of the independent and subordinate claims each solves this problem. Advantageous embodiments of the invention emerge from the definitions, the subclaims, the drawings, and the description of preferred embodiments.

[0008] In one respect, the invention provides a computer-implemented method for the functionally safe localization of a vehicle on a roadway. Functional safety classifies safety requirements necessary to comply with the ISO 26262 standard. This standard identifies four Automotive Safety Integrity Levels, abbreviated ASILs. An ASIL is determined through a risk analysis of a potential hazard, taking into account the severity, exposure, and controllability of the vehicle operating scenario. The safety objective for this hazard, in turn, defines the ASIL requirements. ASIL D dictates the highest integrity requirements for the product. The preceding text regarding functional safety is based on the article "Automotive Safety Integrity Level" from the free encyclopedia Wikipedia and is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License (CC-BY-SA 4.0).

[0009] The vehicle can be, for example, an automated vehicle, such as a vehicle with a driving system for performing autonomous driving functions, i.e., an autonomous vehicle. The vehicle can be, for example, a passenger vehicle, a commercial vehicle, a people shuttle (e.g., an urban mobility shuttle), or a terminal tractor. The roadway can be, for example, a public roadway, a segregated lane, or a private roadway, such as on a factory site, at an airport, or in a business park.

[0010] Reference elements are arranged on, in, or along the roadway. For example, these reference elements may be embedded in the road surface, perhaps so that they are not visible from the outside. The reference elements can be, for example, optical markers, radio-frequency identification (RFID), transponders, or magnets. A reference element's field specifies, among other things, the spatial distribution of its properties. This field can be, for example, an electromagnetic field, such as a light field or a magnetic field. The vehicle is equipped with sensors that detect the fields of these reference elements.

[0011] In one step of the process, a reference map is provided, in which the positions of the reference elements are measured. For example, operating conditions for the automated vehicle are defined for a roadway. These conditions include, for example, the roadway geometry, roadway properties, and traffic characteristics. To fulfill these operating conditions, also known as the operational design domain (ODD), an arrangement of reference elements is determined and installed on, in, or on the roadway. The positions of the individual reference elements are measured with high precision, for example, in world coordinates (also known as Universal Transverse Mercator, UTM), and entered into the reference map. The reference map can have ASIL D integrity, meaning that the map is process-secured and its provision is secured by ASIL-D mechanisms.

[0012] The positions of the reference elements in the reference map represent a target line, also called a reference or ground truth, for the roadway to be driven. Within the detection range of the sensors, for example, the magnetic sensors, a tolerance can be selected, such as a lateral offset of + / - 50 cm from the magnetic line.

[0013] In a further step of the procedure, a starting position, also called the initial position, of the vehicle is obtained on the reference map. This starting position can be obtained, for example, using data from a global navigation satellite system, abbreviated GNSS. A safe, i.e., ASIL-D, starting position is obtained using GNSS in combination with the sequences disclosed below. In a further step, the safe starting position is obtained using GNSS in combination with data from additional vehicle environmental sensors, such as lidar and / or cameras.

[0014] During a journey of the vehicle along the roadway, the following steps of the procedure are carried out: Kinematic signals are read via an interface to the vehicle's electrical system. This system includes sensors such as acceleration sensors, wheel speed sensors, and steering angle sensors, as well as control units, actuators, and bus systems, such as the control area network (CAN bus). From these kinematic signals, the vehicle's change in position relative to its starting position is estimated. This estimation is called safety ego motion estimation (SafetyEME).

[0015] For example, a microcontroller, such as a microcontroller in an electronic control unit (ECU), reads the kinematic signals via corresponding interfaces to the respective sensors and executes the SafetyEME software component, which outputs the self-motion estimate. This allows the calculation of a change in the vehicle's position relative to its starting position.

[0016] The position change obtained by the SafetyEME represents a current path along the traversed roadway. For example, the vehicle's orientation can be estimated from the steering angle and acceleration, and its speed from the wheel rotations. Since the number of wheel rotations depends, among other things, on the tire pressure, which varies, the speed obtained in this way is variable and can only be used as an estimate in the calculation of the position change.

[0017] It should be noted that the kinematic signals are subject to further errors, which accumulate in the SafetyEME. As a result, the actual line obtained from the SafetyEME deviates from the target line of the roadway being traversed. This positional deviation is also known as ego motion drift.

[0018] The input kinematic signals are partially independent of each other and provide partially redundant information. This redundancy ensures that the SafetyEME output, i.e., the position change relative to the starting position, has ASIL D integrity.

[0019] After passing at least two consecutive reference elements, a positional deviation of the vehicle resulting from the estimated change in position is corrected in a further step of the method by adjusting the last two field measurements acquired by the vehicle's sensors to the respective measured positions of the reference elements in the reference map. In this process, a pose of the vehicle within the reference map is obtained after passing at least two consecutive reference elements. The pose includes position and orientation. If the reference map is in UTM format, for example, the position can be output in UTM coordinates. The localization concept according to the invention is applicable to any known map format. The orientation can be output as the yaw angle of the vehicle.Thanks to the known initial position and the SafetyEME, the sensory detections of the reference elements can be assigned to the correct reference elements in the reference map.

[0020] The step described above can be performed by a software component that reads the output of the SafetyEME—i.e., the estimated position change, the measured positions from the reference map, and the field detections by the sensors—and outputs the pose. The output of the SafetyEME and the positions from the reference map can each have ASIL D integrity. In the case of magnets as reference elements and magnetic sensors for sensor-based detection of magnetic fields, this software component is also called magnetic measurement sensor (MMS) localization, or MMS LOC for short. MMS localization can be particularly advantageous in correcting ego motion drift based on comparison with the reference map.

[0021] With each passage of successive reference elements, a new adjustment to the reference map, such as a new MMS localization, can be performed and a new pose obtained. After an aspect is passed, the most recently traversed reference elements, such as magnets, are stored in a buffer. This allows the adjustment to the reference map to be readjusted with each newly passed reference element, such as a single magnet.

[0022] In a further step of the method, the next sensory detection of a reference element is predicted based on the reference map, based on the respective pose, at least the last two sensory detections of the reference elements, and / or the current estimated position change. This step can be performed by a software component that reads the output of the SafetyEME, the field detections by the sensors, the pose (e.g., output by the MMS LOC), and the measured positions of the reference elements from the reference map. This ensures ASIL D integrity of the pose. This software component is called Safety Localization, abbreviated to SafetyLOC. The output of MMS LOC is validated by the continuous prediction of the next sensory detection using SafetyLOC.This ensures a functionally safe localization of the vehicle and guarantees safe operation of the vehicle within the applicable ODD.

[0023] Each prediction is added to a continuous list of reference elements based on an aspect and is only removed from the list when the sensory capture is received.

[0024] If two consecutive predicted detections are not detected by the sensors one after the other, in a further step of the procedure the position of the vehicle is marked as invalid and a stopping maneuver of the vehicle, for example an emergency stop, is initiated based on the last valid pose.

[0025] Due to potential external disturbances, certain robustness measures are implemented based on specific aspects. These robustness measures may include, for example, the following: • only driving for a certain threshold of x meters without sensory detection of a reference element is permitted; • only one reference element may be missed in succession; • the variance of the pose is estimated, only poses with a variance smaller than a given threshold are allowed.

[0026] Localization based on SafetyEME, MMS LOC, and SafetyLOC enables a separation between high-performance localization and functionally safe localization. Such a separation has several advantages and special features, for example: • High-performance localization can be run on any hardware without an ASIL rating. This allows the use of faster and more complex algorithms in high-performance localization. • Code reviews / audits. This allows plausibility checks to be shifted to the software / hardware path of functionally safe localization. • High-performance localization has comparatively higher requirements from a control perspective; for example, a continuous signal is needed. A modular approach can be implemented: The high-performance localization can be a black box with a QM localization output. For example, the signal from a simple, non-certified, environment sensor-based localization system can be used as an input for SafetyLOC.

[0027] Another aspect involves reading signals from an inertial measurement unit (IMU), such as vehicle speed, steering angle (e.g., via a steer encoder), and / or wheel speeds (e.g., via a wheel encoder). For example, the signals from the IMU, steer encoder, and wheel encoder are read into the SafetyEME, with each of these signals having, for example, ASIL B integrity. Due to the redundant information, the resulting output of the SafetyEME from these signals has ASIL D integrity.

[0028] Another aspect is that the positional deviation is corrected after passing at least three consecutive reference elements. It was found that three reference elements offer a comparatively good cost-benefit ratio.

[0029] From another perspective, the reference elements are magnets and the sensors are magnetic sensors.

[0030] The magnets are, for example, permanent magnets, such as anisotropic ferrite magnets, and can be arranged, for example, like needles in a road surface, such as asphalt. For instance, the magnets are positioned along road markings that vehicles must cross, such as longitudinal markings like guide lines or lane markings, surface markings like cycle lane markings, or transverse markings like stop signs or crosswalks. By utilizing a specific arrangement and / or polarity of the magnets, certain information, such as traffic regulations, is encoded into the road surface.

[0031] The magnetic sensors are integrated, for example, into a measuring carrier, such as a one- or two-dimensional array comprising multiple sensors, for example, 25 to 100 sensors. The sensors are arranged, for example, as a series of coils side by side. The magnetic sensors, such as the measuring carrier, can be located on the vehicle's underbody or bumper, for example, at a vertical distance of 30 cm from the road surface. Alternatively, the magnetic sensors can be arranged in a one- or two-dimensional array on the vehicle. In another configuration, the vehicle includes a first magnetic sensor array at the rear and a second magnetic sensor array at the opposite end in the direction of travel. The position of the magnets relative to the magnetic sensors can be measured with an accuracy of up to 5 mm.

[0032] According to another aspect, the arrangement of magnets comprises at least one predefined sequence of magnetic polarities, i.e., a sequence of north and south poles. The magnets belonging to this sequence are identified in the reference map, for example, by a corresponding sequence identifier. When the vehicle's position is determined after passing through this sequence, each sensor-detected magnet in this sequence is assigned to the corresponding measured magnet in the reference map by backward dead reckoning, also known as backward dead reckoning. According to another aspect, backward dead reckoning is performed for detections from the first and second magnetic sensor arrays. If no error is detected when comparing the detections from the first and second magnetic sensor arrays, the resulting pose has ASIL D integrity.

[0033] In a further aspect, regarding the procedural step of adapting the last two field measurements recorded by the vehicle's sensors to the respective measured positions of the reference elements in the reference map, the polarity of the magnets is recorded and compared with the polarities of the magnets stored in the reference map for verification.

[0034] Another aspect is that the vehicle's starting position is derived from its pose after completing this sequence. This provides an initial ASIL D localization.

[0035] Another aspect is that the magnets belonging to the sequence are spaced closer together than the other magnets arranged on, in, or along the roadway. For example, the magnets belonging to the sequence are spaced 1 meter apart, while the other magnets are spaced 4 to 5 meters apart. The spacing of the magnets is chosen such that the ego motion drift between adjacent magnets is small enough that the next magnet is not missed due to drift.

[0036] Another aspect is that the sequence comprises eight magnets. This number has proven particularly advantageous for SafetyLOC. For example, the sequence of magnetic north and south poles is as follows, with each pole oriented towards the magnetic sensor on the vehicle: South-North-North-North-South-North-North-North.

[0037] According to another aspect, the invention provides a computer program for the functionally safe localization of a vehicle on a roadway. The computer program comprises program instructions that cause a hardware component to execute the previously disclosed method when the computer program is loaded or executed by the hardware component. The program instructions can be machine instructions written in assembly language, an object-oriented programming language, for example, C++, a procedural programming language, for example, C, or a hardware programming language. The hardware component can comprise one or more CPUs, GPUs, ASICs, FPGAs, or other ICs, or a combination thereof. According to another aspect, the hardware component is part of a vehicle microcontroller, for example, a control unit. The program can be provided as interpretable or compiled code.

[0038] According to another aspect, the invention provides an automated vehicle, for example, a SAE J3016 Level 4 or 4 autonomous shuttle. The vehicle comprises an array of sensors. The sensors detect fields from reference elements arranged on, at, or within a roadway of the vehicle. The reference elements are, for example, magnets, and the sensors are, for example, magnetic sensors. Furthermore, the vehicle comprises at least one control unit. The control unit includes a memory into which a computer program is loaded or transferable. The computer program implements the method disclosed above. From the execution of the method, the control unit receives control and / or regulation signals and provides these to actuators for longitudinal and / or lateral control of the vehicle. This ensures the vehicle's functionally reliable localization.

[0039] The invention is explained using the following exemplary embodiments. They show: Fig. 1 an embodiment of information sources and software components of a method disclosed herein, Fig. 2 an example of localization, Fig. 3 an embodiment of a functionally safe localization using a sequence of magnetic polarities, Fig. 4 an example of a reference card, Fig. 5 an embodiment of an automated vehicle disclosed herein and Fig. 6 a flowchart of an embodiment of a method disclosed herein.

[0040] In the figures, identical reference symbols designate identical or functionally similar objects. To avoid repetition, only the relevant objects are identified by reference symbols.

[0041] Fig. Figure 1 shows the input information for the method disclosed herein. This can include kinematic signals S1, S2, and S3 of the vehicle 10, for example, a first signal S1 from an accelerometer or an inertial measurement unit, a second signal S2 from a steering angle, and a third signal S3 from a wheel revolution counter. These signals S1, S2, and S3 are read by a software component called SafetyEME. From this, SafetyEME estimates the vehicle 10's intrinsic motion, i.e., its change in position relative to a starting position A during travel along a track 1. The individual signals S1, S2, and S3 can each have ASIL B integrity. Due to the available redundant information, SafetyEME determines the vehicle's intrinsic motion with ASIL D integrity.

[0042] Vehicle 10 includes, for example, magnetic sensors 3. The magnetic sensors 3 are arranged, for example, in a front array and a rear array in the direction of travel, see also Fig. 5. The magnetic sensors detect the magnetic field strengths of reference elements 2 integrated into the road surface 1, for example, magnets. The sensory detection has, for example, QM integrity. QM stands for quality management. At the QM level, all assessed risks are tolerable from a safety perspective.

[0043] The positions of the magnets in lane 1 have been measured and can be retrieved from a reference map.

[0044] The software component MMS LOC reads the magnetic field strength measurements from sensors 3. From this, the position of the magnetic sensors relative to sensors 3 on vehicle 10 can be determined. Furthermore, the software component MMS LOC reads the precisely known position of the magnets from the reference map. Additionally, the software component MMS LOC reads the outputs from SafetyEME and a starting position A of vehicle 10. The starting position A can be determined, for example, using a global navigation satellite system (GNSS). From the known position of the magnets, the measured field strength, the vehicle's own motion, and the starting position A, the software component MMS LOC locates vehicle 10 on the roadway 1 and corrects any positional deviations resulting from the estimated motion. The software component MMS LOC outputs a pose of the vehicle with QM integrity.

[0045] The SafetyLOC software component reads the ASIL-D output from SafetyEME, the QM sensor readings of the magnets, the QM output from MMS LOC, and the ASIL D positions of the magnets from the reference map, and determines an ASIL D pose from this data. In other words, the QM pose determined by MMS LOC is upgraded to ASIL D integrity by SafetyLOC.

[0046] Fig. Figure 2 shows the output of MMS LOC, also using the north and south pole polarities of the magnets. A starting position A is initialized using GNSS. Three magnet acquisitions are performed. Through initialization and the SafetyEME, the magnet acquisitions can be assigned to the correct magnets in the reference map. After three magnet acquisitions, MMS LOC begins to correct the ego's movement based on the alignment with the reference map. The result is a new pose in UTM coordinates and a new yaw angle. With each new measurement, a new map adjustment is performed based on the last three magnet acquisitions.

[0047] Fig. Figure 3 illustrates the functionally safe localization using SafetyLOC. SafetyLOC continuously predicts the next magnet detection based on the output of MMS LOC, the last magnet detection, and the output of SafetyEME. Each prediction is added to a pending magnet detection list and is only removed when the detection is received. If the magnet is not detected, only one magnet is missing sequentially before the output becomes invalid. This triggers an emergency stop of vehicle 10 based on the last valid pose. The required inputs are the ASIL D-rated ego movement from SafetyEME, the ASIL D reference card, and the QM output from MMS LOC. An initial ASIL D localization pose must be provided. This is achieved by defining a unique sequence of polarities along a trajectory on track 1. For example, the sequence includes eight magnets.The vehicle pose at the magnet detection point is determined using a feedback calculation. The magnet assignment then results in a specific sequence identification that is the same for each measured magnet in the sequence.

[0048] For example, the process begins at starting position A and detects the eight magnets in the sequence. Using the sequence identification, MMS LOC can determine the end position B. Through backward feedback calculations starting from end position B, each detected magnet can be assigned to a magnet in the reference map. The magnet in the reference map then provides a sequence identification, which must be the same for all eight magnets in the sequence.

[0049] Fig. 4 shows an example of a reference map.

[0050] Fig. Figure 5 shows an embodiment of an automated vehicle, for example, a shuttle. The shuttle comprises sensors 3 in the form of magnetic sensors that detect fields of reference elements 2 in the form of magnets arranged in the roadway 1. A control unit 4 of the vehicle 10 comprises a memory 5 into which a computer program is loaded or transferable that implements the method disclosed here. The control unit 4 receives control and / or regulating signals from the execution of the method and provides them to actuators for longitudinal and / or lateral control of the vehicle 10.

[0051] Fig.Figure 6 schematically shows the method disclosed here. In a method step V1, the reference map Map is obtained, in which the positions of the reference elements 2 are measured. In a method step V2, the starting position A of the vehicle 10 is obtained in the reference map Map. While the vehicle 10 is traveling along the roadway 1, the following method steps are continuously performed:

[0052] In a method step V3, kinematic signals S1, S2, S3 are read in via an interface to the vehicle's electrical system, and a change in the position of the vehicle 10 relative to the starting position is estimated from these signals. This step is performed, for example, by the SafetyEME software component.

[0053] After passing at least two consecutive reference elements 2, a position deviation of the vehicle 10 resulting from the estimated position change is corrected in a method step V4 by adapting the last two field detections acquired by the sensors 3 of the vehicle 10 to the respective measured positions of the reference elements 2 in the reference map Map. The correction and adaptation is performed, for example, by the software component MMS LOC. A pose of the vehicle 10 within the reference map Map is obtained in each case. For example, MMS LOC outputs the pose.

[0054] In process step V5, the next sensory measurement is predicted based on the reference map, using the current pose, the last two sensory readings of reference elements 2, and / or the current estimated change in position. This step is performed, for example, by the SafetyLOC software component.

[0055] If two consecutive predicted detections are not detected by sensors one after the other, in a process step V6 the position of vehicle 10 is marked as invalid and a stop maneuver of vehicle 10 is initiated based on the last valid pose. Reference symbol 1 lane 2 Reference element 3 Sensor 4 Control unit 5 storage 10 vehicles Map Reference Map A starting position B End position North Pole S South Pole SafetyEME software component for estimating position changes MMS LOC Software component for localization based on the detection of Magnets through magnetic sensors SafetyLOC software component for functionally safe localization GNSS Global Navigation Satellite System ASIL B automotive safety integrity level B ASIL D automotive safety integrity level D QM quality management V1-V6 process steps

Claims

[1] Computer-implemented method for functionally reliable localization of a vehicle (10) on a roadway (1) at, on, or in which reference elements (2) are arranged, wherein the vehicle (10) comprises sensors (3) that detect fields of the reference elements (2), the method comprising the steps of: providing a reference map (V1) in which positions of the reference elements (2) are measured; obtaining a starting position (A) of the vehicle (10) in the reference map (V2); during a journey of the vehicle (10) along the roadway (1): • via an interface to an on-board network of the vehicle (10) reading in kinematic signals (S1, S2, S3) and estimating a change in position (SafetyEME) of the vehicle (10) from the kinematic signals (S1, S2, S3) relative to the starting position (A) (V3); • after passing at least two consecutive reference elements (2), correcting a position deviation of the vehicle (10) resulting from the estimated position change (SafetyEME) by adapting (MMS LOC) the last two detections of the fields recorded by the sensors (3) of the vehicle (10) to the respective measured positions of the reference elements (2) in the reference map (Map) and in each case obtaining a pose of the vehicle (10) within the reference map (Map) (V4); • from the respective pose (MMS LOC), the last at least two sensory detections of the reference elements (2) and / or the respective current estimated position change (SafetyEME), prediction (SafetyLOC) of the respective next sensory detection based on the reference map (Map) (V5); • if two consecutive predicted detections are not detected consecutively by sensors, marking the position of the vehicle (10) as invalid and initiating a stop maneuver of the vehicle (10) based on the last valid pose (V6). [2] Method according to claim 1, wherein the signals (S1, S2, S3) of an inertial measuring unit, speed, steering angle and / or wheel speeds of the vehicle are read in. [3] Method according to one of the preceding claims, wherein the position deviation is corrected after passing at least three consecutive reference elements (2). [4] Method according to one of the preceding claims, wherein the reference elements (2) are magnets and the sensors (3) are magnetic sensors. [5] Method according to one of claims 1, 2 or 3, wherein the reference elements (2) comprise magnets and the sensors (3) comprise magnetic sensors, the arrangement of the magnets comprises at least one predetermined sequence of magnetic polarities (N, S), the magnets belonging to this sequence are marked in the reference map (map), and in the pose of the vehicle (10) obtained after passing this sequence, each sensor-detected magnet of this sequence is assigned to the respective measured magnet in the reference map (map) by backward calculating the change in position of the vehicle (10). [6] Method according to claim 5, wherein the starting position (A) of the vehicle (10) is obtained from the pose of the vehicle (10) obtained after passing this sequence. [7] Method according to claim 5 or 6, wherein the magnets belonging to the sequence are each spaced at a smaller distance than the other magnets arranged on, on or in the roadway (1). [8] A method according to claim 5, 6 or 7, wherein the sequence comprises eight magnets. [9] Computer program for the functionally safe localization of a vehicle (10) on a roadway (1), the computer program comprising program instructions which cause a hardware component to execute the method according to one of the preceding claims when the computer program is loaded or executed by the hardware component. [10] An automated vehicle (10) comprising an array of sensors (3) which detect fields of reference elements (2) arranged on, on or in a roadway (1) of the vehicle (10), and at least one control unit (4) comprising a memory (5) into which a computer program is loaded or transferable which implements a method according to one of claims 1 to 8, wherein the control unit (4) receives control and / or regulating signals from the execution of the method, provides these actuators for longitudinal and / or lateral control of the vehicle (10) and localizes the vehicle (10) in a functionally reliable manner.

Images