METHOD AND SYSTEM FOR MANIPULATING A CHARGING-PORT COVER OF AN ELECTRIC VEHICLE

Abstract

A method for manipulating a charging—port cover of an electric vehicle by an autonomous charging device (ACD), the method comprising, by a control unit controlling an operation of the ACD: obtaining at least one parameter of the charging—port cover; executing a cover manipulation strategy for manipulating the cover into an open state, the strategy being based at least on the obtained parameter of the charging—port cover, wherein the cover—manipulation strategy comprises: performing, by a cover manipulator coupled to the ACD, at least one manipulation of the cover, wherein the cover manipulation strategy further comprises at least one step of: obtaining sensor data for determining a position and / or orientation of the cover, or a feature thereof; or verifying if the cover is in an open state.

Description

ABSTRACT The present invention relates to a method for manipulating a chargingport cover of an electric vehicle (EV) by an autonomous charging device (ACD). TECHNICAL FIELD The present invention relates to the field of autonomous charging systems for electric vehicles. Specifically, it pertains to methods, systems, and computer programs for optimizing the operation ofautonomous charging devices (ACDs). BACKGROUND OFTHE INVENTION The surge in electric vehicle (EV) adoption has propelled a substantial demand for autonomous charging solutions, often referred to as autonomous charging systems or devices. Traditional approaches to vehicle charging often involve manual intervention, requiring drivers to physically connect charging cables to their vehicles. Autonomous charging systems typically feature actuated robotic systems, mechanisms to support and manipulate a charging connector, and computer vision technologies. These technologies, often supported by neural networks and algorithms, enable the system to recognize the vehicle's charging inlet and facilitate the connection and disconnection of the charger to the vehicle inlet in an automated manner. Devices and related methods for this purpose are known in the state of the art, for instance from the international patent applications W02020222640, W02021167462, W02022005281, W02022045881, W02022086320, W02022234059, W02023131577, W02023227412, W02024068806, PCT / EP2023 / 086489, PCT / EP2023 / 058170, NL2035240, NL2035932, NL2036719, NL 2037994, and NL 2037995, all from the same applicant, all of which are herein incorporated by reference. An important challenge lies in reliably and efficiently manipulating the chargingport cover to access the vehicle's charging inlet, especially given the wide variability in chargingport cover designs, geometries, structural features, vehicle parking uncertainties, and sitespecific variabilities. As the task of manipulating the chargingport cover needs to be automated and reliable, the system must avoid causing damage to the cover or the vehicle while ensuring that the manipulation is performed in a precise, fast, and dependable manner. The present disclosure addresses the challenges described herein. SUMMARY OFTHE INVENTION 1 An aspect of the present disclosure provides a method for manipulating a chargingport cover of an electric vehicle by an autonomous charging device (ACD), the method comprising, by a control unit controlling an operation of the ACD: obtaining at least one parameter of the chargingport cover; executing a covermanipulation strategy for manipulating the cover into an open state, the strategy being based at least on the obtained parameter of the chargingport cover, wherein the covermanipulation strategy comprises: (i) performing, by a cover manipulator coupled to the ACD, at least one manipulation of the cover, wherein the cover manipulation strategy further comprises at least one step of: (ii) Obtaining sensor data for determining a position and / or orientation ofthe cover, or a feature thereof; or (iii) Verifying if the cover is in an open state. In some implementations, the sensor data is obtained via one or more sensors coupled to the cover manipulator. In some implementations, the covermanipulation strategy comprises steps (ii) and (iii). In some implementations, an output ofanyone of steps (ii) or (iii) is used to determine whether a subsequent manipulation of the cover is required. In some implementations, an output ofanyone of steps (ii) or (iii) is used to determine whether an adjusted cover manipulation strategy of the cover is required. In some implementations, the control unit is configured to select the cover manipulation strategy based on the at least one parameter of the chargingport cover, wherein the selection is a rulebased selection. In some implementations, control unit is configured to select the cover manipulation strategy based on the at least one parameter of the chargingport cover, wherein the selection is a trainingbased selection. In some implementations, the open state comprises a fullyopened state and a partiallyopened state. In some implementations, the sensor data comprises data characterizing a spatial relationship or a physical interaction between the cover manipulator and the cover, or a feature the cover, wherein the sensor data comprises at least one of tactile, haptic, force, proximity, or torque data, generated by one or more sensors mechanically coupled to the cover manipulator. In some implementations, the at least one parameter of the chargingport cover includes one or more of: a cover pose estimation, a cover opening direction, a cover type, a cover state, a cover locking mechanism, a cover hinge pose. 2 In some implementations, the steps of manipulating of the cover the obtaining of sensor data for determining a position and / or orientation of the cover, are performed substantially simultaneously In some implementations, the manipulation of the cover is a means for determining a position and / or orientation of the cover. In some implementations, the manipulation of the cover comprises applying, by the cover manipulator, a motion within an area of the outer surface of the cover to bring the cover into the open state, wherein controller is configured to use sensor data obtained while executing said manipulation for determining a position and / or orientation of the cover. In some implementations, the area ofthe outer surface ofthe cover on which the motion is applied is determined based on the parameter of the cover, wherein said parameter comprises a pose estimation ofthe cover determined using data from an imaging sensor in communication with the control unit. In some implementations, the cover parameter comprises a subsequent pose estimation of the cover determined upon repositioning the cover manipulator or the imaging sensor to obtain such subsequent pose estimation using data from the imaging sensor. In some implementations, the manipulation of the cover comprises applying, by the cover manipulator, at least one of the following motions to the cover, or the vehicle: a pushing motion, a tapping motion, a touching motion, a pressing motion, a pulling motion, a swiping motion, a rolling motion, a retracting motion, or any combination thereof. In some implementations, the manipulation of the cover comprises a retracting motion comprising moving the cover manipulator away from the cover after performing the pushing motion to prevent interference with a covers opening motion. In some implementations, the manipulation strategy comprises obtaining sensor data from a region outside of the cover, which in combination with sensor data obtained from an area within the outer surface of the cover allows providing a more accurate determination of a position and / or orientation of the cover, or a feature the cover. In some implementations, manipulating a side edge of the cover in a partiallyopened state for bringing the cover into a fullyopened state while obtaining sensor data for providing a more accurate determination of a position and / or orientation of the cover, and if necessary, performing a further manipulation of the cover based on such more accurate determination, for manipulating the cover into an open state. 3 In some implementations, manipulating the cover for bringing the cover into a fullyopened state comprises executing a motion substantially about a cover hinge axis, wherein such cover hinge axis is determined based on any previously obtained determination of a position and / or orientation of the cover, and the at least one parameter of the chargingport cover. In some implementations, the motion for bringing the cover into the fullyopened state is applied only after verifying if the cover is in the partiallyopened state. In some implementations, verifying if the cover is in the fullyopen state comprises determining whether a charging port inlet is detectable by means ofthe imaging sensor. In some implementations, verifying if the cover is in the fullyopen state or partiallyopened state comprises applying, by the cover manipulator, a motion at a region on which a fullyopen cover or a partiallyopen cover is expected, and using sensor data obtained thereby to determine ifthe cover is in the fullyopen state or partiallyopened state. In some implementations, verifying if the cover is in the partiallyopened state comprises using sensor data obtained while applying, by the cover manipulator, a motion on a side edge of the partiallyopened cover, or a motion on an outer surface of the cover. In some implementations, verifying if the cover is in the fullyopen state or partiallyopened state comprises or obtaining information from the imaging sensor In some implementations, the at least one parameter is the state of the cover, and wherein the covermanipulation strategy comprises manipulating the cover to bring the cover to the fully opened state if the state of the cover indicates that the cover is in the partiallyopened state. In some implementations, the at least one parameter is the state of the cover, and wherein the covermanipulation strategy comprises manipulating the cover to bring the cover to a partially opened state if the state of the cover indicates that the cover is in a closed state. A second aspect of the present disclosure provides an autonomous charging system comprising: a cover manipulator device configured to execute a covermanipulation strategy for manipulating the cover into an open state; a sensor system configured to generate a sensor data feedback, wherein the autonomous charging device comprises, or is in communication with, a control unit configured to control the operation of the autonomous charging system, by at least (i) performing, by the cover manipulator coupled to the ACD, at least one manipulation of the cover, (ii) obtaining, by the cover manipulator, information of a position and / or orientation of the cover, or a feature the cover, by using sensor data; and (iii) Verifying ifthe cover is in an open state. 4 In some implementations, the control unit is configured to control the operation of the autonomous charging system and to execute any of the actions described in any of the implementation described herein. In some implementations, the autonomous charging system comprises a connector manipulator configured to engage with a charging connector to insert it into, and extract it from, the vehicle inlet. BRIEF DESCRIPTION OFTHE DRAWINGS FIG. 1A is a perspective view of an example autonomous charging device 10 capable of manipulating a chargingport cover. FIG. 13 is a perspective view of an example autonomous charging device 10 capable of manipulating a chargingport cover. FIG ZA is a perspective view of an example autonomous charging device 10 capable of manipulating a chargingport cover. FIG ZB is a perspective view of an example cover manipulating motion of the autonomous charging device 10. FIG 3A is a perspective view of an example cover manipulating motion of the autonomous charging device 10. FIG 3B is a perspective view of an example cover manipulating motion of the autonomous charging device 10. FIG 4A is a perspective view of an example cover manipulating motion of the autonomous charging device 10. FIG B is a perspective view of an example autonomous charging device 10 suspended from a linear rail 21. FIG. 5 is a simplified structural block diagram of an autonomous charging system 400 and a charging device 10, a station management system 200 according to an embodiment of the disclosure. FIG. 6 is a simplified structural block diagram of an autonomous charging system 400 according to an embodiment of the disclosure. FIG. 7 is a flowchart of an example method for manipulating a chargingport cover. FIG. 8 is a flowchart of an example method for manipulating a chargingport cover. 5 FIG. 9 shows a simplified structural block diagram of a computing environment 101. BRIEF DESCRIPTION OF THE PRIORART U52021094431A1 describes an apparatus and a method for automatically charging a vehicle include a charging plug to connect to a charging port of the vehicle, a manipulator to move the charging plug to the charging port of the vehicle, a power supply connected with the charging plug to supply charging power to the vehicle, and a processor. This reference further describes that the method includes requesting the vehicle to open the chargingport cover and recognizing a position and a connection angle of the charging port of the vehicle when the chargingport cover is opened. DETAILED DESCRIPTION OFTHE INVENTION The present invention is described with reference to flowcharts and block diagrams illustrating the method, system, and apparatus according to embodiments. Each block or combination of blocks may be implemented using computer program instructions or corresponding hardware. These instructions can be executed by a generalpurpose or specialpurpose computer, embedded processor, or other programmable device to perform the specified functions or operations. The computer program instructions may be stored in a computerreadable memory to instruct a computer or programmable device to function in a particular way, thereby enabling the implementation ofthe specified functions or operations. These instructions may also be loaded onto a computer or programmable device to execute a series of steps, resulting in computerimplemented processing. The order of functions or operations in the flowcharts may vary, with some blocks being performed simultaneously or in reverse order, depending on the specific process. An autonomous charging system 400 according to an implementation ofthe present disclosure will be described below in conjunction with FIG 1A and FIG 13. The autonomous charging system 400 is configured to carry out one or more functions related to the autonomous charging of an electric vehicle, including but not limited to manipulation ofthe chargingport cover and manipulation of the charging connector for inserting it into, and extracting it from, the charging inlet. Referring to FIG. 5, a schematic block diagram of an autonomous charging system 400 is shown, which includes an autonomous charging device (ACD) 10 and an ACD control unit 100. The system 400 may, but necessarily has to, be managed by a charging station management system 6 200, responsible for overseeing the operation of the autonomous charging device and other connected systems. Alternatively, the system 400 may be in direct communication with the control unit 100 for control operation. In further reference to FIG. 1A or FIG. 13, the autonomous charging device 10 includes several components, including a motion assembly 20 providing mechanical movement to position and manipulate the device components; an end effector 30 coupled to the motion assembly 20; a chargingportcover manipulator 40 configured for manipulating the chargingport cover; a connector manipulator 50 configured for engaging with the charging connector 41 to insert it into, and extract it from, the vehicle inlet; and one or more sensors which provide feedback or data to assist in the positioning and operation of the manipulator systems. In particular, FIG. lB depicts an imaging sensor 61. The chargingportcover manipulator 40 can be any device able to engage with the charging port cover to manipulate it, such as into an open position or closed position. In preferred examples, the chargingportcover manipulator may be selected from a range of suitable devices, including a poker configured to apply to apply a localized force to the charging port cover, either to push, tap, or nudge it at specific points. A poker may also be configured to hook the cover for pulling or lifting motions, or a poker with a suction cup comprising suctionbased manipulating features, or a gripper equipped with a haptic or a forcefeedback sensor. In further reference to FIG 43 an autonomous charging device 10 is depicted, which comprises a motion assembly 20 comprising an actuated mechanical system mounted on a linear rail 21. The motion assembly 20 may may include at least two actuators, preferably at least three actuators, and more preferably at least six actuators configured to provide control over the position, orientation, and movement of a manipulator coupled to the device 10. These actuators may enable the motion assembly to perform motions, such as translational, rotational, tilting, or lifting movements, required for interacting with the charging port cover or other components of the vehicle. FIG. 43 further depicts a parking bay 330 and a vehicle 300 positioned in the vicinity of charging device 10. The actuated mechanical system mounted on the linear rail 21 may facilitate the motion assembly to be positioned through different parking bays. In some cases, the motion assembly can be a generalpurpose robot arm, a purposebuilt actuated mechanical system, or a combination thereof. In further reference to the structural block diagram of FIG. 5, the autonomous charging system 400 further comprises, or is in communication with an ACD control unit 100 which is configured to execute instructions in accordance with one or more of systems, flowcharts, methods, and / or 7 processes described herein. The control unit 100 includes a controller 110 for managing the operations of the ACD 10 and related components; a motion control unit 120 controlling the movement and actions of the motion assembly 20, end effector 30, cover manipulator 40 and / or connector manipulator 50; a computervision unit 130 for processing visual data from an imaging sensor 61; a communication module 140 using any feasible communication means for facilitating a data exchange between any of the management system 200, autonomous charging system 400, ACD 10, control unit 100, and the vehicle 300. The control unit is further configured to obtain and process sensor data obtained by the sensor 60. Communication means include but are not limited to wireless communication protocols, data exchange interfaces, or network connections. Wireless communication protocols include wireless protocols, such as WiFi, UWB, Bluetooth, Bluetooth Low Energy, 36, 4G, SG, and / or near field communication (NFC) protocols, as some nonlimiting examples. The ACD control unit 100 further comprises, or is in communication with a computing environment 101 as further detailed in FIG. 9. In reference to FIG. 6, a schematic block diagram of vehicle 300 is shown, which in an exemplary realization interacts with the charging system 400 (shown in dashed lines). In further reference to FIG. 1B, the ACD 10 comprises, or is in communication with an imaging sensor, also referred to as a camera sensor 61 or as camera. In reference to FIG. 1B, the camera 61 is mechanically coupled to the end effector 30. In other cases, the camera may be mechanically couped to the cover manipulator 40 or to the connector manipulator 50 and can thus be coupled to the movement of any of such components. ln some embodiments the camera 61 can be coupled to a base component of the motion assembly 20 and thus its position remains substantially unchanged in relation to the movement of the end effector 30, the cover manipulator 40 and / or the connector manipulator 50. In reference to FIG. 5 and FIG. 6, the control unit 100 includes a computervision unit 130 configured to determine a pose of the chargingport cover 310 and / or a pose of the charging inlet 320 utilizing one or more images captured by the camera 61. The ACD 10, by means of the computervision unit 130, is configured to determine the pose of the chargingport cover 310, and / or the charging inlet 320 by utilizing recognizable features thereof visible in an image captured by the camera 61. The computervision unit 130 is preferably provided with an algorithm to determine the pose (i.e. position and / or orientation) of an object through analysis ofone or more images, where the algorithm comprises a suitable neural network trained to identify and locate features, or trained to directly provide the pose of target object. Suitable neural networks for image processing and object recognition include convolutional neural networks (CNNs), which are configured to analyzing images to recognize specific features, and 8 recurrent neural networks (RNNs), which are useful for understanding sequential data patterns. Additionally, graph neural networks (GNNs) may be employed to understand the relationships between different parts of the features to be determined. Other neural networks and algorithms such as deep learning algorithms and reinforcement learning techniques may also be utilized. Recognizable features of the chargingport cover 310 include, but are not limited to a geometric shape of the cover, edges, contours, special relationships of a set of adjacent or nonadjacent image elements, dimensions, color, and combinations thereof. Recognizable features of the charging inlet 320 comprise inlet pins, inlet holes, inlet housing, inlet locking mechanisms, and combinations thereof. Other recognizable features comprise a referential marker, such as a QR code, whose position or relative position to the chargingport cover 310 or charging inlet 320 is known. The autonomous charging system is further configured to obtain pose information of the vehicle 300, chargingport cover 310, or a feature thereof, as well as of the charging inlet 320 by means of sensors configured to measure an interaction of the charging device 10 with its surroundings. In particular, the motion assembly 20 can provide feedback on physical interactions of its end effector 30, or any component coupled thereof (such as the cover manipulator 40 or connector manipulator 50), with objects, including measurements of interaction forces or torques, detection of changes in these forces or torques in specific directions, and identification of thresholds being exceeded. It may also capture the exact pose of the motion assembly during such interactions. In some implementations, the motion assembly is supplemented by a measurement assembly to enhance feedback capabilities. Preferably, the control unit is configured for obtaining sensor data via one or more sensors coupled, or in communication with, the ACD, which facilitate the determination of a position and / or orientation of the cover, or a feature the cover. Such sensors 60 include haptic, force, torque or touch sensors, as well as other sensors configured to determine distance to an object, such as laser distance measurement sensors or ultrasonic sensors, or sensors to determine proximity to an object, like capacitive or inductive sensors. Such sensors can be mechanically coupled to the covermanipulator and are disposed to provide data as the manipulator interacts with the chargingport cover or its surroundings. For example, FIG. ZB depicts a cover manipulator 40 provided with a haptic sensor 62 which is configured to provide additional data points for determining the pose of the chargingport cover or charging inlet, such as the pose of the end effector when measuring a set resistance or contact force, which can complement visual pose estimation. 9 In some cases, these sensors 60 may be coupled to the end effector, the motion assembly, or other components of the motion assembly. The sensors need not be directly coupled to the cover manipulator but may be coupled directly or indirectly to the motion assemblyjoints or other structural elements. Sensor data from these positions can be processed to infer the position, orientation, or applied force on the cover during manipulation. Alternatively, these data points can serve as a single source of information for determining the pose of the charging port cover or charging inlet. By obtaining feedback from different regions or features of the cover, including areas near the edge or hinge, the system can more accurately determine the covers position and orientation. Advantageously, manipulation of the chargingport cover itself serves as a means to obtain (further) information about the position and / or orientation ofthe cover, or a feature thereof. Such information may serve as a confirmation of an assumed position and / or orientation of the cover, in particular, for verifying that the cover is in an open state. For example, when the ACD comprises, or is in communication with, haptic sensors, force sensors, or touch sensors as described herein, the control unit can obtain feedback from such sensors during the manipulation or physical contact with the cover. This feedback facilitates the system to determine the pose of the object being manipulatedwhether it is the chargingport cover, the vehicle, or a feature thereof. Thereby, a more precise determination of the pose ofthe charging port cover, or a feature thereofmay be achieved, particularly in scenarios where visual data from the camera sensor alone may be insufficient. In some cases, the system 400 may determine a position and / or orientation of the vehicle 300, chargingport cover 310, charging inlet 320, or any feature thereof, using information received from the vehicle 300 itself, or from a charging station management system 200 as depicted in FIG. 6. Such information may be transmitted through the communications module 140. Information about the pose of the vehicle, the pose of the chargingport cover 310 or the pose ofthe charging inlet 320 comprises at least one of accurate pose coordinates, estimated pose coordinates, and expected pose coordinates, a range of pose coordinates indicating a volume where the chargingport cover and / or charging inlet can be found, at any time or at a predetermined time. Thereby, the ACD 10 may determine the pose not only by direct analysis of image data or sensor data, but by incorporating previously obtained pose information. This allows the image data and / or sensor data to be used for final or more accurate pose estimation. A method for manipulating a chargingport cover of an electric vehicle by an autonomous charging device, will be described below with further references to the anyone of FIG. 1 to FIG. 9. 10 FIG. 7 is a brief flowchart of the method 500 according to an embodiment of the present disclosure. In further reference to any of FIG. 1A and FIG 1B, the method 500 comprises, by a control unit controlling an operation of the autonomous charging device: Step 510: obtaining at least one parameter of the chargingport cover; Step 520: executing a covermanipulation strategy for manipulating the cover into an open state, the strategy being based at least on the obtained parameter of the chargingport cover, wherein the covermanipulation strategy comprises Step 530: performing, by a cover manipulator coupled to the ACD, at least one manipulation of the cover, wherein the cover manipulation strategy further comprises at least one step of: Step 540: Obtaining sensor data for determining a position and / or orientation ofthe cover, or a feature thereof; or Step 550: Verifying if the cover is in an open state. A cover manipulation strategy, as used herein, comprises a combination of motions and iterative determinations of the position and / or orientation of the chargingport cover, or a feature thereof in order to bring the cover to an open position. In some cases, one single cover manipulation strategy may be sufficient to achieve that goal. In some cases, more than one cover manipulation strategy may be required to bring the cover to an open position, for example when the cover has not been successfully manipulated, orwhen the cover is stuck or its position has changed during the execution of a first manipulation strategy. A cover manipulation strategy can operate as an iterative process, wherein the manipulator executes one or more motions to manipulate the cover while simultaneously or subsequently obtaining additional data about the cover's position and / or orientation. Each new manipulation and each new determination of the position and / or orientation may provide more accurate information about the cover's spatial state and may facilitate the control system to adjust, or determine the need of, a subsequent manipulation step. Thereby, the system may adjust the cover manipulation strategy, such as by performing an additional manipulation steps or by performing an additional pose determination where it is necessary in order to bring the chargingport cover to an open state. In some cases, one single In some cases, the cover manipulation strategy comprises the execution of a predefined sequence of motions for manipulating the cover into an open state without the obtaining of further information of a position and / or orientation of the cover, or a feature thereof, by using sensor data. Thereby, where the control unit is provided with one or more parameters of the chargingport cover which provide sufficient accurate data of the chargingport cover and its position and orientation, the strategy may not require subsequent pose determinations but only a final verification step that the cover has reached an open state. 11 In some cases, the cover manipulation strategy further comprises, at any moment of the execution of such strategy, verifying if the cover is an open state. Preferably, the determination of a position and / or orientation of the cover serves both for updating the pose, and verifying that the cover has reached an open position. Thereby, the system 400 may adapt to unexpected conditions, such as variations in the covers resistance, misalignments, or deviations from the expected pose, or the cover not reaching the expected open position. In an exemplary realization, the method initiates by the control unit 100 obtaining information on at least one parameter of the chargingport cover, which can be an estimated position and / or orientation or range of positions and / or orientations of the chargingport cover. This can be done using information provided by the vehicle 300, the charging station management system 200, the charging device 10 or the system 400. For example, if a known vehicle type is parked in a specified direction at a parking bay for which coordinates and measurements are stored in advance, it is possible to obtain at least an estimate information ofwhere the chargingport cover is likely to be located and how it is approximately oriented. The control unit 100 then can gather further information about another parameter of the chargingport cover. Given a known vehicle type, or even vehicle identification, such parameter may include the covers shape, size, color, and how it opens (e.g., swings outward to the left or right, or slides up / down / inside). Such additional parameter may be previously stored, for example, when the charging system 400 or the management system 200 have prior knowledge of the parameters of a specific vehicle brand and model, the system can automatically gather such information. Based on the estimated pose ofthe cover and the covers parameters as detailed above, the control unit 100 can determine a strategy for manipulating the cover into an open position. For example, ifthe cover swings outward and to the left after a first push on the cover to trigger a release mechanism, the strategy can include a first pushing of the cover at a specific point or region on a right portion of the cover surface and then retracting the manipulator to avoid interference as the cover opens. Furthermore, the strategy can include a pose determination of the cover while performing such first pushing motion in order to obtain more accurate pose determination of the cover. In the case where such first pushing motion enables the cover to be partially opened rather than fully opened, the strategy may further include verifying if the cover has achieved that partially opened state. This verification can be performed by obtaining visual information from the camera or by collecting sensor data from a haptic sensor coupled to the cover manipulator. For 12 instance, after the initial first pushing motion, the cover manipulator may execute a motion to touch an outer side of the chargingport cover with the purpose to detect if, and where, it touches the cover. If the cover is partiallyopened state, resistance or an opposing force may be recorded at a specific location on the outer side of the cover where such resistance is expected. Conversely, if no resistance is detected at the expected location, the control unit may infer that the cover has not yet reached the partiallyopened state and can adjust a next step in the manipulation strategy accordingly to achieve the desired state. For example, the control unit may repeatedly reattempt the first pushing motion until the verification of reaching the expected open position is positive, and / or revert to obtaining a new image of the chargingport cover, and a new estimate of the pose of the chargingport cover using the computervision unit. Based on a verification that the cover is partially open, the control system may then initiate a subsequent motion to bring the cover into a fullyopened state. According to the present disclosure, manipulating the cover, and verifying ifthe cover is in an open state, are steps which can occur simultaneously, sequentially, or iteratively and in any order. Subsequently, using the previously obtained parameter of the cover, its pose determination and the additional feedback received during a first manipulation, the control unit can calculate a suitable trajectory for the manipulator to follow in order to bring the cover into the fullyopened state. This trajectory can consider factors such as the hinge position, opening direction, and cover orientation to enable a transition from partiallyopened state to fullyopened state. Once the cover manipulator has executed a motion to bring the cover to a fullyopened state, the control unit may verify if the charging cover has reached the fullyopened state. In such a case, the control unit may obtain image data from an imaging sensor or haptic feedback from the sensor coupled to the cover manipulator to verify the covers position and orientation. For example, the imaging sensor may verify that the charging inlet is fully visible, indicating that the cover is in its fully open state. Alternatively, a haptic sensor may detect an absence of resistance or confirm contact with a predefined stopping point, as an indication that the manipulator has completed the motion to fully open the cover. Specific details of each ofthe process steps depicted in FIG. 7 and preferred realizations thereof are described in the following. Step 510 comprises obtaining at least one parameter of the chargingport cover. Such parameter may be obtained through various means, either directly from the vehicle 300, the charging station management system 200, the charging device 10 or the system 400, in each case utilizing suitable communication means. In some embodiments, parameters about the vehicle 300 may 13 be prestored in the autonomous charging system 400 or may be obtained from the charging station management system 200 and in each case, information about such parameter may be determined based on historical data, prior knowledge of vehicle behavior, prior knowledge of vehicle properties, or utilizing a combination of sources of information as described herein. For instance, the system may obtain information of the vehicle which is parked in the corresponding bay, and based on such data, the control unit 100 may infer any relevant parameters of the chargingport cover, such as position relative to the vehicle, size, opening direction, among others. Preferably, the at least one parameter of the chargingport cover comprises an estimated pose information of the chargingport cover, which may facilitate the subsequent covermanipulation strategy by providing an approximate pose thereof. This estimated pose may include information about the spatial orientation, such as the angle, tilt, or rotation of the cover relative to the vehicle body. The pose estimation can be based on data obtained from visual sensors, such as cameras, or from other sensors, such as haptic or proximity sensors, integrated into the autonomous charging system. Obtaining an estimated pose of the chargingport cover does not necessarily require a communication from the vehicle at each instance, as the charging system 400 may already possess or access relevant data. For instance, if the vehicle is parked within a defined range from the charging system 400, the control unit 100 can predict the estimated position of the charging port cover based on the known type of vehicle, without requiring realtime communication from the vehicle 400. Similarly, if the vehicle is expected to park near the device 10 for a scheduled charging session, the vehicle 300 may provide parking information to the system 400 in advance, or the vehicle may be instructed to park in a specified manner, allowing to anticipate the vehicle's position or chargingport cover estimated position and prepare for the charging process. The estimated position of the chargingport cover can also be obtained, confirmed or improved based on feedback from sensors, including cameras, force sensors, touch sensors, haptic sensors, laser sensors, or any combination thereof. Preferably, camera data may provide visual information of the cover's position by identifying recognizable features such as edges, contours, or markers on the cover. Force or haptic sensors may detect contact and resistance during interaction with the cover, providing additional data points about the cover's position, orientation, or movement characteristics. Laser sensors may be positioned on the device 10 or in any other relevant location, such as a charging station infrastructure, to capture distance measurements or spatial data. This information can contribute to determining the pose of the 14 vehicle or the chargingport cover relative to the manipulator. For example, a laser sensor may scan the vehicle's surface or the area surrounding the charging port to identify the pose of the cover, even in lowlight or obstructed conditions Other parameters of the chargingport cover include the geometric shape, size, type dimension, color, material, texture, orientation, and opening mechanism of the chargingport cover. Parameters also include information about the hinge and its opening directionwhether the cover swings upward, downward, sideways, or outward, and the angle of rotation required to partially or fully open it. Likewise, the system can obtain this information through a combination of means, such as camerabased feedback, proximity sensors, or data transmitted from the vehicle or a management system. In particular, hinge direction data may enable the control unit to determine the manipulator's motion, so that the manipulation strategy (such as pushing and pulling) is applied in the correct direction to open the cover without causing damage or interference. Additionally, the control unit may account for other related factors, such as the hinges position and stiffness, which could influence the required force and strategy for opening the cover. A parameter may also include a current state of the inlet cover (such as closed, partiallyopened or fullyopened) and a workspace surrounding the inlet cover for manipulation of the charging connector; a dimension of the inlet cover, a volume surrounding the inlet cover amongst others, and any combination thereof. In an exemplary embodiment, the parameter of the chargingport cover is the opening angle of the inlet cover, specified as an angle in degrees. For example, the inlet cover may open at an angle of about 90 degrees with an opening direction to the left. Another parameter includes the inlet orientation relative to the vehicle, such as the inlet being positioned at a certain angle in relation to the fixed world. For example, the inlet could be oriented at five degrees relative to a vertical plane of the vehicle. The type of chargingport cover may influence the approach angle ofthe manipulatorwhether it comprises push, pull, or apply combined forces. The material of the cover might require adjustment of the force applied to avoid damage, particularly with fragile or lightweight composite materials. Step 520 comprises executing a covermanipulation strategy for manipulating the cover into an open state, the strategy being based at least on the obtained parameter of the chargingport cover. The covermanipulation strategy comprises Step 530 performing, by a cover manipulator coupled to the ACD, at least one manipulation of the cover, and wherein the cover manipulation strategy further comprises at least one step of: 540 (ii) Obtaining sensor data for determining a 15 position and / or orientation of the cover, or a feature thereof; or 550 (iii) Verifying if the cover is in an open state. A cover manipulation strategy preferably comprises a sequence of motions for the manipulator to manipulate the chargingport cover into an open position. The strategy preferably involves performing additional steps to obtain more accurate information of a position and / or orientation of the cover, or a feature thereof, by using sensor data, as an initial estimated position may not provide sufficient accuracy for a reliable manipulation sequence. The control unit 400 is configured to select a suitable covermanipulation strategy to be applied to each specific case in consideration of the at least one parameter of the chargingport cover. The selection of such suitable covermanipulation strategy can be a rulebased selection, a trainingbased selection or a combination thereof. For example, a rulebased selection indicates a strategy that is predefined based on a situation, and the trainingbased selection indicates a strategy pattern that is changeable based on a situation through training. The selection of the manipulation strategy may be also be made through a neural network model, algorithm, or look up table. . A cover manipulation strategy may be stored locally in the control system 100, the communication module 140 or the computing environment 101 or accessed remotely via cloud computing services. The model processes input data, such as the at least one parameter of the charging cover, and outputs a manipulation strategy. This predetermined model may already contain predefined strategies for specific types of vehicles or chargingport covers. In some cases, the model can also be updated or adjusted based on realtime data encountered during operation. In some implementations, the control unit 400 may utilize advanced machine learning models, such as supervised, unsupervised, or reinforcement learning models, to refine the manipulation strategy. These models may include, but are not limited to, convolutional neural networks (CNN), recurrent neural networks (RNN) with long shortterm memory (LSTM) units or gated recurrent units (GRU), deep belief networks (DBN), or statespace dynamic neural networks (SSDNN). Such models can be trained on datasets comprising various vehicle types, cover designs, and environmental conditions to provide adaptive and accurate manipulation strategies. Additionally, alternative artificial intelligence techniques, such as decision trees, Bayesian networks, and principal component analysis, may complement these neural networks, providing ensemble learning approaches or hybrid systems. 16 Further, the control system may be configured to adapt the behavior of the cover manipulator based on different operational modes or scenarios. For example, the system may switch between predefined manipulation behaviors based on the current state of the vehicle, the charging port cover, or external environmental conditions. In a "safe mode," the manipulator may prioritize conservative motions to minimize the risk ofdamage to the cover or vehicle, relying on rulebased algorithms for enhanced reliability. In contrast, a "normal mode" may balance safety with operational efficiency, enabling more swift manipulations when the system is confident in its pose estimations. For advanced scenarios, such as handling unexpected resistance or obstructions, a "trainingbased mode" may prioritize reinforcement learning to dynamically optimize the manipulation strategy in real time. A cover manipulation strategy can combine vehiclespecific information with environmental factors. For example, if the ACD has previously interacted with similar vehicles, the model may be able to predict the cover's pose, opening direction, or potential obstructions. This reduces the risk of errors, ensures efficient manipulation of the cover, and minimizes potential damage to the vehicle or the charging system. The model may also help optimize the ACD's operational performance. As a result, the ACD can anticipate the vehicles chargingport cover properties and determine a motion strategy suitable to the vehicle's specific requirements. In accordance with the present disclosure, the covermanipulation strategy comprises step 530 of performing, by a cover manipulator coupled to the ACD, at least one manipulation of the cover, which may include a combination of a pushing motion, a pressing motion, a pulling motion, a swiping motion, a rolling motion, a retracting motion, or any combination thereof by the cover manipulator to manipulate the cover, depending on the cover's type, dimensions, hinge mechanism, and orientation. Manipulating the cover may also include vibrational or tapping motions to assist in overcoming friction or dislodging a stuck cover. In accordance with the present disclosure, the covermanipulation strategy may further comprise step 540 comprising obtaining sensor data for determining a position and / or orientation of the cover, or a feature thereof. Thereby accuracy and reliability of the manipulation process may be improved. As used herein, sensor data may include any sensor in as described in the present disclosure. In some cases, obtaining haptic, touch or force data for determining a position and / or orientation of the cover, or a feature thereof, while, either simultaneously or independently from the step of manipulating the cover, can allow determining a cover hinge axis orientation. Such data may be obtained, for example, from a point or region on, or near the edge of the 17 charging cover where the manipulator interacts with the cover. By analyzing the resistance and positional data from this edge, together with a geometric parameter of the cover, the control unit 100 can determine the pose of the cover hinge axis. Such information can facilitate the control system 100 to adjust the manipulation strategy so that subsequent motions, such as pulling, rotating, or retracting the charging cover to bring it to fullyopened state, are in accordance with a hinge's orientation. Step 550, which comprises verifying if the cover has reached an open position may involve determining whether the cover has reached a fullyopened state or a partiallyopened state. The verification process may include checking intermediate states, such as confirming that the cover has been moved from a closed to a partiallyopened state. Verifying if the cover is partially open may be done by obtaining feedback from any of the sensors described in the present disclosure. Such sensor feedback, such as haptic or force data, can provide realtime information about the cover's position and resistance, providing information that the cover has reached the desired partially open state. In each case, the camera feedback may also be utilized to verify if the cover has reached an open position. Verifying if the cover is in a fullyopened state can be done by identifying whether the vehicle's charging inlet has become fully visible. The visibility of the inlet not only serves as an indication that the cover has been fully opened but also allows the system to determine the pose of the inlet. This pose data can be used for further mating of the charging connector into the inlet. In accordance with the present disclosure, steps 530 (manipulating the cover), 540 (determining a position and / or orientation of the cover), and 550 (verifying ifthe cover has reached an open position) may be performed in any order or sequence, depending on the operational requirements conditions encountered during the cover manipulation process. For instance, in some embodiments, determining the position and / or orientation ofthe cover (step 540) may precede manipulation (step 530) to ensure accurate targeting, while in other embodiments, manipulation (step 530) may be initiated first, with pose information subsequently refined during or after manipulation. Similarly, verification (step 550) may occur intermittently during manipulation or as a final step to confirm that the desired open state has been achieved. An exemplary embodiment of the present disclosure is described below in further reference with anyone of FIG. 2A, FIG ZB, FIG. 3A, FIG. 3B, FIG. 4 and FIG 8. In reference to FIG. ZA, the ACD 10 records, via the camera 61, a first image ofwhere it expects the chargingport cover 310 to be positioned and / or oriented, based on the at least one parameter of the chargingport cover, and determines a more accurate pose of the cover using 18 the computervision unit 130. The ACD 10 may subsequently adjust camera parameters and / or reposition the camera 61 based on the more accurate pose of the cover, and record a second image of the cover 310, and further refining the pose ofthe cover 310 using the computervision unit based on that second image of the cover. Subsequently, if, based on the obtained cover parameter, the cover 310 swings outward, the strategy comprises applying a first pushing motion comprising moving the cover manipulator 40 forward until a first interaction is detected, after which the cover manipulator is used to exert force to trigger the covers release mechanism as depicted in FIG. ZB. Subsequently, the cover manipulator is retracted to avoid interference with the opening motion of the cover. Subsequently, as depicted in FIG. 3A, to verify that the cover 310 has reached the partially opened state, the control unit may instruct the motion control unit for the cover manipulator to partially repeat the first pushing motion until it detects a second interaction. When the second interaction takes place earlier along a motion path of the first pushing motion, the system may infer that the cover has indeed reached a partiallyopened position. When the second interaction happens substantially at the same position as the first interaction, the system may infer that the cover 310 has not opened. Alternatively, the first pushing motion may be followed by a swiping motion of the cover manipulator along the body work of the vehicle starting from a position adjacent to the cover towards the edge 311 of the cover 310, such that the cover manipulator 310 catches the edge 311 of the cover in the case that the cover has indeed reached the partiallyopened position as depicted in FIG. 3B. By analyzing the resistance and positional data from this edge 311, together with a previously obtained geometric parameter of the cover, the control unit 100 can determine the pose of the cover hinge axis 312 as depicted in FIG. 4A. Such information can facilitate the control system 100 to adjust the manipulation strategy so that subsequent motions, such as pulling, rotating, or retracting the charging cover to bring it to fullyopened state, are in accordance with a hinge's orientation. Alternatively, after the first pushing motion, the camera 61 and computervision unit may be used to evaluate the state of the chargingport cover 310. Alternatively, for a cover manipulator comprising a feature to latch onto covers using magnetism or negative pressure, the retracting motion after the first pushing motion may experience a higher resistance when the cover is latched versus when the cover is not latched, indicating that the cover reached (or is reaching) the partiallyopened position. 19 Sensor data can be further gathered from a point on, or near the chargingport cover, which the control system can utilize to obtain information about the position and / or orientation of the cover, or a feature thereof. For example, once the cover has reached a partiallyopened state, the manipulator may apply a second touch motion outside of the cover surface or adjust its position relative to the hinge axis to refine the pose estimation further as depicted in FIG 3A. Combining sensor data from multiple interactions may allow the control system to improve its knowledge ofthe cover's position and orientation. Once it has been verified that the cover has reached a partiallyopened state, a motion is applied to bring the cover into a fullyopened state. Such motion is determined based on the pose information obtained from previous manipulations, including data regarding the hinge position, opening direction, and the spatial relationship between the cover and the manipulator. The control unit may calculate a trajectory for the manipulator to follow, ensuring that the motion is aligned with the hinge's orientation and the cover's opening mechanism. For instance, the manipulator may apply a pulling motion at a determined region on the covers edge to rotate it about the hinge axis. Finally, the control unit may verify that the cover 310 has reached the fullyopened state as seen in FIG. 4A by using one or more methods, such as visual confirmation via the camera, resistance feedback indicating the end of the covers travel, or a geometric alignment check with the charging port. Verification that the cover is fully opened may also involve determining whether the charging port inlet is completely visible, thereby ensuring unobstructed access for the charging connector. If the verification indicates that the cover is not fully opened, the strategy may include obtaining additional pose information of the cover, using sensors such as the camera or haptic sensors, and initiating an additional motion to bring the cover to the fully opened state. Another exemplary embodiment of the present disclosure is illustrated in FIG. 8. The process begins with Step 600, where the system initiates the method for charging a vehicle. Step 610 begins by the positioning ofthe vehicle in a designated charging area. In Step 620 the controller acquires an initial pose estimation of the charging port cover using a camera sensor. Such estimation provides approximate information about the covers position and orientation, for subsequent steps. In Step 630, based on the obtained pose estimation and other cover parameters, the controller selects an appropriate cover manipulation strategy. In Step 640 the cover manipulator executes the selected strategy to manipulate the cover into a partially opened position. During Step 640, the system gathers sensor data, such as haptic or force feedback, to refine the pose determination of the cover. In Step 650, once the cover is in a 20 partially opened state, the cover manipulator uses the refined pose information to execute a subsequent motion to bring the cover to a fully opened position. In Step 660 the system verifies whether the cover has reached the fully opened position, which may involve visual feedback from the camera sensor, haptic feedback from the manipulator, or a combination of both. If the verification is successful, the process ends, as the cover is fully open and the charging port is accessible for further proceeding with the charging process. If the verification indicates the cover is not fully open, the system proceeds to Step 670 where the controller gathers additional sensor data using the sensors coupled to the cover manipulator and such refined pose information is used to adjust the manipulation strategy and reattempt opening the cover fully. In Step 680, ifthe system detects a failure in the initial strategy to open the cover, the controller may acquire an updated cover parameter and adjust the manipulation strategy accordingly. FIG. 9 depicts a generalized example of a suitable computing environment 101 in which the described innovations may be implemented. The computing environment 101 is not intended to suggest any limitation as to scope of use or functionality, as the innovations may be implemented in diverse generalpurpose or specialpurpose computing systems. For example, the computing environment 101 can be any of a variety of computing devices (e.g., desktop computer, laptop computer, server computer, tablet computer, etc.) With reference to FIG. 9, the computing environment 101 includes one or more processing units 110, 115 and memory 120, 125. In FIG. 8, this basic configuration 1030 is included within a dashed line. The processing units 110, 115 execute computerexecutable instructions. A processing unit can be a generalpurpose central processing unit (CPU), processor in an applicationspecific integrated circuit (ASIC) or any other type of processor. In a multi processing system, multiple processing units execute computerexecutable instructions to increase processing power. For example, FIG. 9 shows a central processing unit 110 as well as a graphics processing unit or coprocessing unit 115. The tangible memory 120, 125 may be volatile memory (e.g., registers, cache, RAM), nonvolatile memory (e.g., ROM, EEPROM, flash memory, etc.), or some combination of the two, accessible by the processing unit(s). The memory 120, 125 stores software 180 implementing one or more innovations described herein, in the form of computerexecutable instructions suitable for execution by the processing unit(s). A computing system may have additional optional features. For example, the computing environment 101 includes storage 140, one or more input devices 150, one or more output devices 160, and one or more communication connections 170. An interconnection mechanism (not shown) such as a bus, controller, or network interconnects the components of the computing environment 101. Typically, operating system software (not shown) provides an 21 operating environment for other software executing in the computing environment 101, and coordinates activities of the components ofthe computing environment 100. The tangible storage 140 may be removable or nonremovable, and includes magnetic disks, magnetic tapes or cassettes, CDROlVls, DVDs, or any other medium which can be used to store information in a nontransitory way and which can be accessed within the computing environment 101. The storage 140 stores instructions for the software 1080 implementing one or more innovations described herein. The input device(s) 150 may be a touch input device such as a keyboard, mouse, pen, or trackball, a voice input device, a scanning device, or another device that provides input to the computing environment 100. The output device(s) 160 may be a display, printer, speaker, CD writer, or another device that provides output from the computing environment 100. As used herein, chargingport cover features comprise, but are not limited to one or more geometric shapes, edges, contours, hinge position, hinge orientation, color, texture, surface patterns, locking mechanism, opening direction, relative positions to the charging inlet, relative position to the vehicle body, and combinations thereof. As used in the present disclosure, the term "pose" refers to a spatial position and / or orientation in any one of a three, four, five, or sixdimensional space of an object. Additionally, or alternatively, the pose may involve dimensions of information representing the object's position and / or orientation relative to the camera. As used herein, "manipulating" refers not only to the act of moving the chargingport cover to achieve a desired position or orientation but also to the physical interaction between the cover manipulator and a part ofthe vehicle. Such interaction may serve to obtain information about the pose of the chargingport cover or other related features. As used herein, chargingport cover parameters include, but are not limited to, a pose estimation of the cover determined using data from an imaging sensor, as well as subsequent pose estimations obtained by repositioning either the cover manipulator or the imaging sensor to refine the pose estimation. Additional parameters may include the type of vehicle, the vehicles parking position relative to the autonomous charging device (ACD), communication of the covers pose, determination of the covers pose, state of the cover, cover opening direction, an opening sequence initiated by the electric vehicle (EV), or the position of the cover hinge. As used herein, the state of the chargingport cover comprises at least three distinct conditions: a closed state, a partiallyopened state, and a fullyopened state. In the closed state, the 22 chargingport cover remains positioned over the charging port, fully obstructing both visual and physical access to the charging port. In this state, the cover provides protection for the charging port from external elements such as dirt, moisture, or physical impacts. In the partiallyopened state, the chargingport cover is no longer the closed state but remains in a position where unobstructed access to the charging port is not yet provided. This state is typically achieved by applying an initial push or other manipulation to the cover, transitioning it from a position that is flush with the vehicle's bodywork into a state where it is standing ajar. In this position, the cover becomes grippable or accessible for further manipulation, as the manipulator can engage with the edge of the cover or get behind it to continue the opening process. In the fullyopened state, the chargingport cover is positioned such that unobstructed access to the charging port is provided for the charging connector. This access may be visual, allowing the control system to confirm alignment and readiness, or physical, enabling the charging connector to mate directly with the charging port. Although the steps of the disclosed methods are presented in a particular order for convenience, they may be rearranged unless a specific order is required. Operations may also be performed concurrently. For simplicity, the figures do not illustrate all possible combinations with other methods. In the present disclosure, the term robot" or robotic may, but not necessarily must, conform to the traditional industry standards definition of a complete autonomous robot with sensory perception, decisionmaking capabilities, and mobility. Instead, in the context of the present disclosure, the term robot" primarily refers to a controlled actuated mechanism or arm designed for specific tasks, such as the ones in accordance with the present disclosure. The term "robot" emphasizes the mechanized and programmable nature of the system rather than the strict dependance on a fully autonomous robot with generalpurpose capabilities. The disclosed embodiments can be implemented as computerexecutable instructions stored on various computerreadable media (e.g., optical discs, memory, or hard drives) and executed on a computer, including mobile devices. The instructions can be part of a dedicated application or accessed via a web browser and can operate on local computers or in a network environment (e.g., via the Internet or a cloud computing network). For clarity, only selected aspects of software implementations are described, and wellknown details are omitted. The disclosed technology is not limited to specific programming languages or hardware. 23 The functionality described can also be implemented via hardware logic components such as FPGAs, ASICs, ASSPs, SoCs, or CPLDs. Softwarebased embodiments may be uploaded, downloaded, or accessed remotely through communication means such as the Internet, cable, or wireless technologies (e.g., RF, microwave, or infrared). Terms like configured to or able to encompass components in active, inactive, or standby states, unless otherwise specified. The terms comprising," including," and having" are used inclusively and do not exclude additional elements. Or" is inclusive, meaning one, some, or all items in a list. The articles a," an," and the" mean one or more" unless specified otherwise. At least one of" or and / or" refers to any combination of listed items. The term allow" includes permitting, instructing, enabling, or facilitating a specified action. The detailed description is illustrative, and variations that do not depart from the essence of the claimed invention are within its scope. 24

Claims

1 Procedure for manipulating a tailgate of an electric vehicle by a autonomous ACD, where the mode of operation comprises, by a control unit that a controls the operation of the ACD: obtaining at least one parameter of the Iaadpoort valve; executing a valve manipulation strategy for manipulating the valve in an open position condition, where the strategy is based at least on the parameter obtained from the Loading gate valve, where the valve manipulation strategy comprises: the execution, by a valve manipulator connected to the ACD, of at least one manipulation of the valve, where the valve manipulation strategy further comprises at least one step by: (ii) obtaining sensor data to determine a position and / or orientation of the valve, or a characteristic thereof; or (iii) verifying whether the valve is in an open position. 2 Method according to claim 1, whereby the sensor data are obtained via one or more sensors that are connected to the valve manipulator. 3 Method according to one of claims 1 or 2, where the valve manipulation strategy steps (ii) and (iii) includes. 4 Method in accordance with one of the preceding claims, involving the execution of steps (ii) or (iii) is used to determine whether subsequent manipulation of the valve is required. 5 Method of operation in accordance with one of the preceding claims, involving the execution of steps (ii) or (iii) is used to determine whether a modified valve manipulation strategy is required. 6 Method of working in accordance with one of the preceding conclusions, where the control unit is configured to select the valve manipulation strategy based on at least one parameter of the Iaaadpoortklep, where the selection is a rule-based selection. 7 Method in accordance with one of the preceding conclusions, where the control unit is configured to select the strategy for manipulating the valve based on the ten at least one parameter of the valve of the Iaadpoort, where the selection is training-based selection is 25 8 Method according to one of the preceding conclusions, where the open state is a ful includes an open state and a partially open state. 9 Method according to one of the preceding conclusions, whereby the sensor data include those that involve a spatial reaction or a physical interaction between the valve manipulator and the valve characteristics, or a characteristic of the valve, where the sensor data at least one of the tactile, include haptic, force, proximity, or torsion data, generated by one or more sensors which are mechanically coupled to the valve manipulator. 10 Method in accordance with one of the preceding claims, whereby at least one parameter of the flap of the loading gate comprises one or more of the following: an estimate of the flap position, a opening direction of the valve, a type of valve, a valve status, a locking mechanism of the flap, a hinge position of the flap. 11 Method according to one of the preceding conclusions, whereby the steps of manipulating of the valve and obtaining sensor data for determining a position and / or orientation of the valve, be carried out substantially simultaneously. 12 Method of operation in accordance with one of the preceding conclusions, whereby the manipulation of the valve a midde| is for determining a position and / or orientation of the valve. 13 Method of operation in accordance with one of the preceding conclusions, whereby the manipulation of the valve the applying involves, by the valve manipulator, of a movement within an area of ​​the outer surface of the valve to bring the valve to the open position, whereby the controller is configured to use sensor data obtained during the execution of said manipulation for determining a position and / or orientation of the valve. 14 Method according to claim 13, where the area of ​​the outer surface of the valve on which the movement applied is determined based on the parameter of the valve, where The said parameter comprises a position estimate of the valve that is determined using data of an image sensor in communication with the control unit. 15 Method according to one of the preceding conclusions, where the valve parameter a subsequent position estimation of the valve includes, determined upon repositioning of the valve manipulator or the image sensor to obtain such a subsequent pose estimation using data from the image sensor. 16 Method of operation in accordance with one of the preceding conclusions, whereby the manipulation of the valve the applying involves, by the valve manipulator, of at least one of the following movements on the 26 flap or the vehicle: a push movement, a tap movement, a touch movement, a press movement, a pulling movement, a sweeping movement, a rolling movement, a retraction movement or a combination of them. 17 Method of operation in accordance with one of the preceding claims, whereby the manipulation of the valve a involves a retraction movement, whereby the valve manipulator, after performing the pushing movement of the The valve is moved away to prevent interference with the opening movement of a valve. 18 Method of working according to one of the preceding conclusions, whereby the manipulation strategy the obtaining sensor data from an area outside the valve involves, which in combination with sensor data obtained from an area within the outer surface of the valve made it possible to make a more accurate determination of a position and / or orientation of the valve, or a characteristic of the valve, to be provided. 19 Method of operation in accordance with one of the preceding conclusions, whereby the method involves the manipulation of encompasses a side edge of the valve in a partially open position to hold the valve in a fully to bring to the open state while sensor data is obtained to determine a more accurate to provide determination of a position and / or orientation of the valve, and if necessary, to carry out of further manipulation of the valve based on such a more precise determination, in order to the to manipulate the valve into an open position. 20 Method according to conclusion 19, whereby manipulating the valve to the valve in a conduit to bring to an open state, performing a movement that substantially to a flap hinge shaft rotates, where such a flap hinge shaft is determined based on a previously obtained determination of a position and / or orientation of the valve, and at least one parameter of the loading gate valve. 21 Method according to conclusion 19, whereby the movement around the valve in the fully opened to bring to condition, a||a is applied after verifying whether the valve is in the partial is in an open state. 22 Procedure according to claim 8, whereby verifying whether the valve is in the fully open condition, determining whether a charging port inlet is detectable by means of the image sensor. 23 Procedure according to claim 8, whereby verifying whether the valve is in the fully open or is in a partially open state, the application involves, by the valve manipulator, of a movement in an area where a fully open valve or a partially open valve is 27 expected, and using sensor data obtained thereby to determine whether the valve is in the fully open or partially open position. 24 Procedure according to claim 8, whereby verifying whether the valve is in the partially open state is in, the use of sensor data obtained during application, by the tilting mechanism, of a movement on a side edge of the partially opened flap, or a movement on an outer surface of the valve. 25 Procedure according to claim 8, whereby verifying whether the valve is in the fully open or is in a partially open state, involving the acquisition of information from the image sensor. 26 Method of working in accordance with one of the preceding claims, where at least one parameter the condition of the valve is, and where the valve manipulation strategy involves manipulating the valve to bring the valve to the fully open position if the state of the valve indicates that the valve is in the partially open position. 27 Method of working in accordance with one of the preceding claims, where at least one parameter the is the condition of the valve, and where the tilt manipulation strategy involves manipulating the valve to bring the valve to a partially open position if the condition of the valve indicates that the valve is in a closed position. 28 Autonomous charging device, comprising: a tilt manipulator device configured to execute a valve manipulation strategy for manipulating the valve in an open position; a sensor system that is configured to generate sensor data feedback, where the autonomous charging device comprises a control unit, or is in communication with, which is configured to control the operation of the autonomous steering system, by at least (i) by the valve manipulator connected to the ACD, at least one manipulation of the valve to carry out, (ii) to obtain sensor data to determine a position and / or orientation of the valve, or a characteristic thereof; and (iii) to verify whether the valve is in an open position. 29 The autonomous charging device within the meaning of claim 28, where the device is configured to a to carry out the steps of one of the conclusions 1 through 27: 28 3O A computer-readable storage medium containing instructions which, when executed by a processor, induce the processor to execute a process according to one of the conclusions 1 through 27. 5 29

Images