Inventory system for a vehicle, procedures for providing an inventory system, and procedures for operating an inventory system

Abstract

The invention relates to an inventory system (100) for an environment (U) in a vehicle (F), a method for providing a corresponding inventory system (100) for an environment (U) in a vehicle (F) and a method for operating a corresponding inventory system (100) for an environment (U) in a vehicle (F).

Description

[0001] The invention relates to an inventory system for an environment within a vehicle. Furthermore, the invention relates to a method for providing a corresponding inventory system for an environment within a vehicle. The invention further relates to a method for operating a corresponding inventory system for an environment within a vehicle.

[0002] It is usually difficult for humans to keep track of all objects in their immediate surroundings, especially if these objects are small or there are a large number of them. However, there are use cases where confirmation is needed when an object of interest is located within a limited area, for example: - A family wants to go on holiday in their own car and wants to make sure that all relevant items are in the car, e.g. certain suitcases, medicine box, glasses cases, mobile devices, etc. - A traveler wants to receive confirmation that all essential items for a business trip are in the car, e.g., a laptop, a smartphone, a briefcase, etc. - It must be ensured that a list of items is available in a room, e.g. a first aid kit, a warning triangle, etc. in the trunk of the vehicle.

[0003] It is therefore an object of the present invention to overcome at least one of the disadvantages described above, at least partially. The object addressed in this invention is how such a list of relevant objects can be automatically learned by an intelligent inventory system for different users without requiring manual specification of the objects on the lists. In particular, it is an object of the invention to provide an inventory system for an environment within a vehicle. Furthermore, it is an object of the invention to provide a method for providing a corresponding inventory system for an environment within a vehicle. Finally, it is an object of the invention to provide a method for operating a corresponding inventory system for an environment within a vehicle.

[0004] The object of the invention is achieved by an inventory system for an environment in a vehicle with the features of the independent system claim. Furthermore, the object of the invention is achieved by a method for providing a corresponding inventory system for an environment in a vehicle. Finally, the object of the invention is achieved by a method for operating a corresponding inventory system for an environment in a vehicle. Features and details described in connection with the different embodiments and / or aspects of the invention naturally also apply in connection with the other embodiments and / or aspects, and vice versa, so that the disclosure relating to the individual embodiments and / or aspects always makes, or can make, reciprocal references.

[0005] The preceding problem is solved by: an inventory system for an environment in a vehicle, comprising: a sensor system, e.g., permanently installed in the environment or mobile, in particular an RFID-based sensor system, a storage device, e.g., permanently installed in the environment, mobile (e.g., in a user's mobile device), or external (e.g., in a cloud), in which a list of objects is stored, a computing device, e.g., permanently installed in the environment, mobile (e.g., in a user-owned mobile device), or external (e.g., in a cloud), wherein the computing device is designed to register multiple objects in the inventory system, which are equipped with tags that have unique identifiers of the objects, and wherein the sensor system is designed to detect the tags in order to identify the objects.

[0006] The inventory system according to the invention thus includes a sensor system, e.g. an RFID-based sensor system.

[0007] Objects that are recorded in the inventory system can be equipped with tags (e.g., RFID tags) that are detectable, especially by the sensor system (e.g., scannable).

[0008] The tags can also be described as information providers (and can, for example, be passively read) and / or signal providers (and can, for example, actively output information, e.g., only when the sensor system proves its authorization to the tags).

[0009] The unique identifiers include, for example, names and / or identifications of the objects. An identification can be advantageous for clearly distinguishing between objects with the same name. For example, if several "eyeglass cases" are listed in the inventory system, the identification can include a specific number (1, 2, 3, etc.), a specific identification code (a sequence of characters), or something similar to differentiate between the "eyeglass cases."

[0010] The tags can be integrated directly into objects or attached afterwards.

[0011] Integrated tags can enable pre-installation and / or pre-programming, especially before the objects reach the user, for example to assign names and / or identifications to objects (e.g. "medication box").

[0012] When tags are attached to objects by the user, they can advantageously be programmed afterwards to assign names and / or identifications to the objects. This can be done, for example, via a user-owned mobile device, such as a smartwatch or smartphone, equipped with NFC.

[0013] Furthermore, the sensor system can be configured to authenticate the tags, for example, with an identification code, to assign objects to an authorized user or group of users (e.g., a family). This authentication ensures that only the user or group of users can read information from the tag. Authentication can be pre-installed and / or programmed subsequently. For example, an identification code can be stored in the memory unit, which is assigned to an authorized user or a list of users.

[0014] Advantageously, users can also have a user-specific tag and be included in the inventory system (advantageously using a special identifier that, for example, allows a distinction between an object and a user).

[0015] Furthermore, the user-specific tag can be used in the learning process to link recognized tags or identified objects to specific users. The user-specific tag can list user-specific objects, preferably with corresponding probabilities for the presence of the objects. User-specific inventory lists of objects can be assigned to each user.

[0016] The inventory system can first scan an environment.

[0017] The environment can represent a defined scan volume, preferably an enclosed space (e.g., vehicle interior and / or trunk, etc.).

[0018] The environment can be assigned to a user or a specific group of users.

[0019] The inventory system can scan the environment for tags (and thus for objects).

[0020] During scanning, tags are detected that uniquely belong to specific objects.

[0021] An existing object can be represented by a 1 as a variable value.

[0022] A missing object can be represented by a 0 as the variable value.

[0023] A user who is present can be represented by a variable value of 1.

[0024] An absent user can be represented by a 0 as a variable value.

[0025] The inventory system can contain a neural network, which can be stored, for example, in the storage unit.

[0026] The neural network can be trained to estimate a probability (e.g., 0 or 1) for each registered object in a given list of users.

[0027] For training purposes, the following information (so-called basic truth) can be provided to the neural network, with user intervention (so-called supervised learning) or without user intervention (so-called autonomous learning): If an object o i is present (represented by a variable value o i = 1, so-called basic truth), the probability p(o i ) for a recognized day also p(o i ) = 1. If an object o i is not present (represented by a variable value o i = 0, so-called fundamental truth), the probability p(o i ) for an unrecognized day also p(o i ) = 0. Probabilities can also vary between 0 and 1 or between 0% and 100%. For example, a probability p(o i ) ≤ 50% indicates an unrecognized day. For example, a probability p(o i ) > 50% for a recognized day.

[0028] The presence of objects in an environment can depend in particular on the user(s) present.

[0029] Further training steps can be performed to learn a correlation between existing objects and present users of a given list.

[0030] The presence of objects in an environment can further vary depending on the time of day (e.g., described by a time of day and / or a day of the week, e.g., with numbers from 1 to 7 (i.e., Monday to Sunday)) and / or planned journey (e.g., described by a starting point and / or destination).

[0031] Further training steps can be performed to learn a correlation between existing objects, users present in a particular list, and specific times and / or locations.

[0032] The neural network can be pre-trained. This training can be used to create user-specific inventory lists of objects.

[0033] The neural network can also be trained during operation, e.g., event-specifically with each scan, and / or periodically, e.g., daily, weekly, monthly, etc. This training can be used to adapt user-specific inventory lists to objects over time, for example, as user habits change.

[0034] Furthermore, the user can be warned if (depending on a previously learned, user-defined inventory list) the following is detected: - missing object o i , if p(o i ) - o i> 0.5. - additional object o i , which is normally missing when p(o i ) - o i < -0.5.

[0035] Assigning tags to objects o i or, for example, to object numbers i, this can be done as follows: When a user wants to add a new object (to an inventory list) to the inventory system, it can be configured (e.g., during the initial scan) that the object is registered so that it is assigned a name and / or identification. This can be done via a user-owned mobile device, such as a smartwatch or smartphone equipped with NFC (e.g., retrospectively), a dedicated reader (e.g., in advance), or a suitable interface, such as directly in the vehicle via the vehicle's user interface. If the inventory system recognizes this object during subsequent operation, it learns the object's name and / or identification. The inventory system can then add the new object to an inventory list.

[0036] If the user wants to delete an object, the user can simply remove the object from the inventory list.

[0037] If an object is no longer recognized for a certain period of time, the inventory system can ask the user, e.g. via a suitable user interface, whether the object needs to be removed from the inventory list.

[0038] If an object is no longer recognized for a certain period of time, then the object can be removed from the inventory list, e.g. automatically, even without having to ask the user.

[0039] Transferring learned inventory lists can be done, for example, as follows: Complete inventory lists of registered objects and / or model parameters of a fully trained neural network can be saved and transferred as part of a personal profile to another vehicle (or to the same, but updated and / or serviced, vehicle). The storage device can be used for this purpose.

[0040] However, it is also possible to quantize learned inventory lists by quantizing the probabilities p(o i ) to save and easily activate them when the user is present (u j = 1): p(oi):={0if p(oi)≤0.51if p(oi)>0.5

[0041] Several advantages can be achieved with the help of the invention: - Inventory lists can be automatically learned for a specific context (time, place and co-users). - Inventory lists adapt to the changing usage patterns of the user. Inventory lists can also be seen as a representation of user identity (especially if they include personal items) and serve as a, potentially additional, means of user authentication. This user identity can also be learned. - Location context can be represented by the start and end positions rather than the complete movement path, since objects accompany the user, so the object list should not change during the journey. - Time context can be represented by the time of day and / or the time of week, as daily and weekly usage patterns show repetitions.

[0042] Furthermore, the computing device may be configured to register at least one user in the inventory system who carries a user-specific tag. This user-specific tag can uniquely authenticate the user. First, the user-specific tag can serve to register a user in the inventory system. Advantageously, the user-specific tag can also serve to distinguish a user from an object. Furthermore, the user-specific tag can establish a link between a user and an inventory list of objects. Finally, the user-specific tag can be used to create a user-specific inventory list of objects.

[0043] The proposed inventory system allows for multiple user-specific inventory lists to be maintained for different users or for different groups of users.

[0044] Furthermore, the storage device may contain an identification code for at least one user or a user list. This identification code can be used to authorize the sensor system to read the tags.

[0045] In principle, it is conceivable that a specific user or group of authorized users could be assigned to the environment. This can be advantageous for a private vehicle.

[0046] Furthermore, it is conceivable that the environment could be assigned to several specific users or several specific groups of authorized users. This can be advantageous for a company vehicle or a rental vehicle. The boarding user can conveniently transfer their user-defined inventory list of objects and / or model parameters from a pre-trained neural network to the corresponding vehicle as part of their personal profile.

[0047] Furthermore, it may be provided that an artificial neural network is stored in the storage device, which has been specifically trained to create user-specific inventory lists of objects and to adapt them if necessary, e.g. over time, for example, if user habits have changed.

[0048] In principle, it is conceivable that the artificial neural network could be designed as a feed-forward neural network, a convolutional neural network, or a recurrent neural network.

[0049] Furthermore, it is conceivable that the sensor system can be provided as a permanently installed system in the environment or as a mobile sensor system.

[0050] Furthermore, it is conceivable that the storage device can be provided as a permanently installed storage device in the environment, as a mobile device (e.g., in a user-owned mobile device), or as an external storage device (e.g., in a cloud).

[0051] Furthermore, it is conceivable that the computing device can be provided as a permanently installed device in the environment, as a mobile device (e.g., in a user-owned mobile device), or as an external device (e.g., in a cloud).

[0052] Furthermore, it is conceivable that the sensor system, the storage device, and, if applicable, the computing device, could be designed as a single, integrated module. This module could, for example, be permanently installed in the vehicle, such as in the dashboard. Alternatively, the module could be easily brought along and placed in the vehicle, for example, on the dashboard.

[0053] Furthermore, it is conceivable that the sensor system is spatially separated from the storage device and / or the computing device. For example, the sensor system could be an integral part of the vehicle, whereas the storage device and / or the computing device could be implemented elsewhere, e.g., on a user-owned mobile device such as a smartwatch or smartphone, or on an external device such as a cloud.

[0054] The preceding problem is solved by: a method for providing an inventory system, which may be implemented as described above, for an environment in a vehicle, comprising: - Providing a sensor system, in particular an RFID-based sensor system, - Providing a storage device in which a list of objects is stored, - Providing a computing device, wherein the computing device is used to register multiple objects in the inventory system, which are equipped with tags that have unique identifiers of the objects, and where the sensor system is used to scan the tags in order to detect the objects.

[0055] This achieves the same advantages described above, to which full reference is made here.

[0056] Furthermore, the procedure for providing an inventory system may include: - Providing or equipping objects with tags that have unique identifiers for the objects.

[0057] The identifiers can advantageously include names and / or identifications of the objects. This allows the objects to be uniquely identified, even if they have the same name, for example, if two or more objects with the same name are recorded in the inventory system.

[0058] As mentioned above, tags can be used to register objects in the inventory system.

[0059] Furthermore, it is conceivable that the tags can be integrated directly into objects or subsequently attached to objects. If the tags are integrated directly into objects, they can have a pre-installation and / or pre-programming to assign names and / or identifications to the objects. If the tags are attached to objects subsequently, they can be programmed to assign names and / or identifications to the objects, for example, via a user-owned mobile device, preferably equipped with NFC, such as a smartwatch or smartphone.

[0060] Furthermore, the procedure for providing an inventory system may include: - Providing or equipping at least one user with a user-specific tag that uniquely authenticates that at least one user.

[0061] As mentioned above, the user-specific tag can be used to register a user in the inventory system. Advantageously, the user-specific tag can be used to distinguish a user from an object, to establish a link between a user and an inventory list of objects, and / or to create a user-specific inventory list of objects.

[0062] Furthermore, the procedure for providing an inventory system may include: - Storing an identification code for at least one user or a user list in the storage device.

[0063] As mentioned above, the identification code can be used to identify the sensor system as authorized (in the case of the tags) to read the tags.

[0064] As mentioned above, an environment can be assigned to a specific group of authorized users, e.g., a family or similar.

[0065] Furthermore, the procedure for providing an inventory system may include: - Providing an artificial neural network that is specifically trained to create and, if necessary, adapt user-specific inventory lists of objects.

[0066] The artificial neural network can enable the proposed inventory system to achieve its improved functionality.

[0067] Furthermore, the procedure for providing an inventory system may include: - Scanning the vehicle's environment for tags.

[0068] Advantageously, the environment can be defined as a defined scan volume, preferably a closed compartment, such as a vehicle interior and / or a trunk.

[0069] Furthermore, the procedure for providing an inventory system may include: - Assigning scanned tags to registered objects and / or users.

[0070] As mentioned above, an existing object and / or user can be represented by a 1 as a variable value.

[0071] As mentioned above, a non-existent object and / or a non-existent user can be represented by a 0 as a variable value.

[0072] Furthermore, the procedure for providing an inventory system may include: - Training a neural network to provide a probability, e.g. between 0 and 1, for each registered object and / or each logged-in user.

[0073] As mentioned above, the probability of an existing object and / or user can be represented by a 1.

[0074] As mentioned above, the probability of a non-existent object and / or a non-existent user can be represented by a 0.

[0075] The training can be conducted as supervised training (at least in advance, e.g. before the inventory system is put into operation) or as independent training (at least during the vehicle's operation, e.g. after the inventory system has been put into operation).

[0076] Furthermore, the procedure for providing an inventory system may include: - Training a neural network to learn a correlation between existing objects and present users of a given list.

[0077] It is conceivable that existing objects in the environment may depend on the user(s) present.

[0078] Furthermore, the procedure for providing an inventory system may include: - Training a neural network to learn a correlation between existing objects, users present in a given list, and specific times and / or locations.

[0079] Furthermore, the procedure for providing an inventory system may include: - Storing user-specific inventory lists of objects and / or model parameters of a trained neural network according to one of claims 13 to 15 as part of a personal profile.

[0080] This can be advantageous for activating user-specific inventory lists, for example, when a particular user is present. It can also be advantageous for enabling the transfer of user-specific inventory lists.

[0081] The neural network can be trained in advance (e.g., before the inventory system is put into operation).

[0082] Furthermore, the neural network can be trained during operation (e.g., after commissioning of the inventory system), e.g., event-specifically, e.g., with each scan, and / or periodically, e.g., every day, every week, every year, etc.

[0083] The preceding problem is solved by: a method for operating an inventory system, which may be implemented as described above, for an environment in a vehicle that has been trained as described above, comprising: - Warn a user if an object is missing from a user-defined inventory list, and / or - Warn a user when an additional object is detected.

[0084] This achieves the same advantages described above, to which full reference is made here.

[0085] Furthermore, the procedure for operating an inventory system may stipulate that when a user wants to add a new object to a user-defined inventory list: - Registering the new object so that the object is assigned a name and / or an identification, e.g. during an initial scan of the new object.

[0086] This can be done, for example, via: - a user-side mobile device, preferably equipped with NFC, e.g. a smartwatch or a smartphone, e.g. subsequently, - a special reading device, e.g. in advance, and / or - a suitable interface, e.g. directly in the vehicle environment, preferably via a vehicle user interface.

[0087] Furthermore, as part of the procedure for operating an inventory system, it can be provided that if an object is no longer recognized for a certain period of time, the user is asked whether the object should be removed from a user-generated inventory list.

[0088] Alternatively, it can be provided that if an object is no longer recognized for a certain period of time, the object is automatically removed from a user-defined inventory list, e.g. automatically, preferably without asking the user.

[0089] Advantageously, the following can be provided for in the procedure for operating an inventory system: - Adapting learned inventory lists to changing user usage patterns.

[0090] Furthermore, the following may be provided for in the procedure for operating an inventory system: - Using user-specific inventory lists of objects and / or model parameters of a trained neural network as part of a personal profile.

[0091] This can be advantageous to enable the activation of user-specific inventory lists when a particular user is present, and / or to enable the transfer of user-specific inventory lists.

[0092] Furthermore, the following may be provided for within the framework of the procedure for operating an inventory system: - Using learned inventory lists of registered objects as a means, possibly an additional means, for user authentication and / or for enabling vehicle functions.

[0093] The invention and its advantages are explained in more detail below with reference to a drawing. It schematically shows: Fig. 1. An exemplary procedure for checking whether an inventory list is complete.

[0094] The Fig. Section 1 serves to explain the invention idea, which proposes: an inventory system 100 for an environment U in a vehicle F, a method for providing a corresponding inventory system 100 for an environment U in a vehicle F and a method for operating a corresponding inventory system 100 for an environment U in a vehicle F.

[0095] The inventory system 100 includes a sensor system 10, e.g. an RFID-based sensor system.

[0096] objects o i Items that are listed in inventory system 100 can be equipped with tags (e.g., RFID tags) that are recognizable, especially by sensor system 10 (e.g., scannable).

[0097] The tags can be understood as information providers (and can, for example, be passively read) and / or signal providers (and can, for example, actively output information, e.g., only when the sensor system 10 proves its authorization to the tags).

[0098] The unique identifiers include, for example, names and / or identifications of the objects. i . Identification can be advantageous for distinguishing between objects o i to be able to clearly distinguish between items with the same name. For example, if several "eyeglass cases" are listed in inventory system 100, then the identification can include a specific number (1, 2, 3, etc.), a specific identification code (a sequence of characters), or similar, in order to differentiate between different "eyeglass cases".

[0099] The tags can be placed directly in objects or i be integrated or added later.

[0100] Integrated tags can enable pre-installation and / or pre-programming, especially before the objects are used. i the user u j to reach, for example, objects or i To assign names and / or identifications (e.g. “medicine box”, “warning triangle”, “laptop”, etc.).

[0101] If tags are from the user u j on objects o i If they are to be attached, then they can advantageously be programmed afterwards to control objects or i Assigning names and / or identifications, e.g. via a user-side mobile device, e.g. equipped with NFC, such as a smartwatch or a smartphone.

[0102] Furthermore, authentication can be provided for the tags of sensor system 10, e.g. with an identification code, to identify objects. i an authorized user u j or a group of users u j(e.g., a family). Authentication can ensure that only the user u j or a group of users u j Information can be read from the tag. Authentication can be pre-installed and / or programmed subsequently. For this purpose, for example, an identification code can be stored in storage device 20, which is assigned to an authorized user. j or a list of users u j is assigned.

[0103] The users also u j They can have a user-specific tag and also be included or registered in inventory system 100 (advantageously using a special identifier that, for example, distinguishes between an object or i and a user u j can enable).

[0104] Furthermore, the user-specific tag can be used in the learning process to identify recognized tags or identified objects.i with specific users u j to be able to link them. The user-specific tag can link user-specific objects, preferably with corresponding probabilities p(o). i ) for the existence of the objects o i , list. To each user u j Can user-specific inventory lists be created for objects? i be assigned.

[0105] The inventory system 100 can first scan an environment U.

[0106] The environment U can be designed as a defined scan volume, preferably a closed compartment (e.g. vehicle interior and / or trunk, etc.).

[0107] A user can access the environment U. j or a specific group of users u j be assigned.

[0108] The inventory system 100 can search the environment U by tags (and thus by objects). i scan.

[0109] During scanning, tags are detected that identify specific objects. i clearly belong.

[0110] An existing object o i can be represented by a 1 as the variable value.

[0111] A missing object o i can be represented by a 0 as the variable value.

[0112] A user present u j can be represented by a 1 as the variable value.

[0113] An absent user u j can be represented by a 0 as the variable value.

[0114] The inventory system 100 can have a neural network KNN, which can be stored, for example, in the storage device 20.

[0115] The neural network KNN can be trained to predict a probability p(o i ) (e.g., 0 or 1 and / or between 0 and 1 and / or between 0% and 100%) for each registered object o ia specific list of users u j to appreciate.

[0116] For training purposes, the following information (so-called basic truth) can be provided to the neural network ANN, with user intervention (so-called supervised learning) or without user intervention (so-called autonomous learning): If an object o i is present (represented by a variable value o i = 1, so-called basic truth), the probability p(o i ) for a recognized day also p(o i ) = 1.

[0117] If an object o i is not present (represented by a variable value o i = 0, so-called fundamental truth), the probability p(o i ) for an unrecognized day also p(o i ) = 0.

[0118] Probabilities p(o i ) can also vary between 0 and 1 or between 0% and 100%.

[0119] For example, a probability p(o i ) ≤ 50% indicates an unrecognized day.

[0120] For example, a probability p(o i ) > 50% for a recognized day.

[0121] The presence of the objects o i In an environment U, the user(s) present may, in particular, j depend.

[0122] Further training steps can be performed to establish a correlation between existing objects or i and present users u j to learn a specific list.

[0123] The presence of the objects o i In an environment U, the following can also occur depending on the time of day (e.g., described by a time of day t). d and / or a day of the week w , e.g. with numbers from 1 to 7 (i.e. Monday to Sunday)) and / or planned trip (e.g. described by a starting point x) s and / or destination x d ) vary.

[0124] Further training steps can be performed to establish a correlation between existing objects or i , present users u j to learn a specific list and at specific times and / or places.

[0125] The ANN neural network can be pre-trained. This training can be used to create user-specific inventory lists of objects. i to create.

[0126] The ANN neural network can also be trained during operation, e.g., event-specifically with each scan, and / or periodically, e.g., daily, weekly, monthly, etc. This training can be used to create user-specific inventory lists of objects. i to adapt over time, for example when user habits change.

[0127] The user u j A warning can be issued if (depending on a user-defined inventory list) the following is detected: - missing object o i , if p(o i ) - o i > 0.5. - additional object o i , which is normally missing when p(o i ) - o i < -0.5.

[0128] Assigning tags to objects o i or, for example, to object numbers i, this can be done as follows: If the user u j a new object o i (to an inventory list) that is to be managed in inventory system 100, it can be specified (e.g., during the first scan) that the object o i is registered so that the object o iA name and / or identification can be assigned. This can be done via a user-side mobile device, e.g., an NFC-enabled smartwatch or smartphone (e.g., retroactively), and / or via a special reader (e.g., in advance), and / or via a suitable interface, e.g., directly in the vehicle F via the vehicle user interface. If the inventory system 100 detects this object during subsequent operation... i Once it recognizes the object, it learns its name and / or identification. i The inventory system can then add the new object 100. i Add to an inventory list.

[0129] If the user u j an object o i If the user wishes to delete, they can u j the object o i Simply remove it from the inventory list.

[0130] If an object o i If the user is no longer recognized after a certain period of time, then the user can... jThe inventory system will ask 100 people, e.g., via a suitable user interface, whether the object o i must be removed from the inventory list.

[0131] If an object o i If the object is no longer recognized after a certain period of time, then the object may be... i to be removed from the inventory list, e.g. automatically, even without the user's intervention. j to ask.

[0132] Transferring learned inventory lists can be done, for example, as follows: Complete inventory lists of registered objects o i and / or model parameters of a fully trained neural network (KNN) can be saved and transferred as part of a personal profile to another vehicle F (or to the same, but updated and / or serviced vehicle F). Storage device 20 can be used for this purpose.

[0133] However, it is also possible to quantize learned inventory lists by quantizing the probabilities p(o i ) to save and easily activate them when the user u j is present (u j = 1): p(oi):={0if p(oi)≤0.51if p(oi)>0.5

[0134] The preceding explanation of the embodiments describes the present invention solely by way of examples. Naturally, individual features of the embodiments can be freely combined with one another, provided this is technically feasible, without departing from the scope of the present invention. Reference symbol list 100 inventory system 10 Sensor system 20 Storage device 30 calculating device O i object u j user F vehicle U surroundings KNN artificial neural network t d Time of day tw weekday x s Starting point x d Destination

Claims

[1] Inventory system (100) for an environment (U) in a vehicle (F), comprising: a sensor system (10), in particular an RFID-based sensor system, a storage device (20) in which a list of objects (o i ) is stored, a computing device (30), wherein the computing device (30) is designed to process multiple objects (o i ) to register in the inventory system (100) which are equipped with tags that provide unique identifiers of the objects (o i ) exhibit, and wherein the sensor system (10) is designed to detect the tags in order to identify the objects (o i ) to identify. [2] Inventory system (100) according to claim 1, wherein the computing device (30) is designed to support at least one user (u i ) to register in the inventory system (100) that carries a user-specific tag, which requires unique authentication of at least one user (u j ) shows, and / or which is designed to create a user (u j ) to register in the inventory system (100), and / or which is designed to enable a user (u j ) of an object (o i to distinguish, and / or which is designed to create a link between a user (u j ) and an inventory list of objects (o i ) to produce, and / or which is executed to create a user-specific inventory list of objects (o i to create. [3] Inventory system (100) according to claim 1 or 2, wherein the storage device (20) contains an identification code (ID) for at least one user (u j ) or a user group is assigned to it, which is used to identify the sensor system (10) as authorized to read the tags, where preferably the environment (U) includes a specific group of authorized users (u j ) is assigned. [4] Inventory system (100) according to one of the preceding claims, wherein an artificial neural network (ANN) is stored in the storage device (20), which has been specifically trained to generate user-specific inventory lists of objects (o i to create and, if necessary, adapt. [5] Inventory system (100) according to any one of the preceding claims, wherein the sensor system (10) can be provided as a permanently installed sensor system in the environment (U) or as a mobile sensor system, and / or wherein the storage device (20) can be provided as a storage device permanently installed in the environment (U), as a mobile device, e.g. in a user-owned mobile device, or as an external storage device, e.g. in a cloud, and / or wherein the computing device (30) can be provided as a fixed computing device in the environment (U), as a mobile computing device, e.g. in a user-owned mobile device, or as an external computing device, e.g. in a cloud. [6] Inventory system (100) according to one of the preceding claims, wherein the sensor system (10) and the storage device (20), and optionally the computing device (30), are designed as a connected module unit, or wherein the sensor system (10) is spatially separated from the storage device (20) and / or the computing device (30). [7] Method for providing an inventory system (100), in particular according to one of the preceding system claims, for an environment (U) in a vehicle (F), comprising: - Providing a sensor system (10), in particular an RFID-based sensor system, - Providing a storage device (20) in which a list of objects (o i ) is stored, - Providing a computing device (30), wherein the computing device (30) serves to process several objects (o i ) to register in the inventory system (100) which are equipped with tags that provide unique identifiers for the objects (o i ) exhibit, and wherein the sensor system (10) serves to scan the tags in order to identify the objects (o i ) to recognize. [8] Method according to the preceding claim, further showing: - Providing or equipping objects (o i) with tags that uniquely identify the objects (o i ) exhibit, where the labels are names and / or identifications of the objects (o i ) include, and / or where the tags serve to identify objects (o i ) to register in the inventory system (100), and / or where the tags are directly embedded in objects (o i ) integrated or subsequently added to objects (o i ) be attached, and / or wherein the tags enable pre-installation and / or pre-programming if the tags are directly embedded in objects (o i ) are integrated to handle objects (o i ) To assign names and / or identifications, and / or where the tags are programmed subsequently if the tags are subsequently attached to objects (o i ) are attached to objects (o i) To assign names and / or identifications, e.g. via a user-side mobile device, preferably equipped with NFC, such as a smartwatch or smartphone. [9] Method according to any one of the preceding claims, further showing: - Providing or equipping at least one user (u j ) with a user-specific tag, which uniquely authenticates at least one user (u j ) shows, and / or which serves to provide a user (u j ) to register in the inventory system (100), and / or which serves to provide a user (u j ) of an object (o i to distinguish, and / or which serves to establish a link between a user (u j ) and an inventory list of objects (o i ) to produce, and / or which serves to create a user-specific inventory list of objects (oi to create. [10] Method according to any one of the preceding claims, further showing: - Storing an identification code (ID) for at least one user (u j ) or a user list in the storage device (20), to identify the sensor system (10) as authorized to read the tags, where preferably the environment (U) includes a specific group of authorized users (u j ) is assigned. [11] Method according to any one of the preceding claims, further showing: - Providing an artificial neural network (ANN) specifically trained to generate user-specific inventory lists of objects (o i to create and, if necessary, adapt. [12] Method according to any one of the preceding claims, further showing: - Scanning an environment (U) inside the vehicle (F) for tags, wherein in particular the environment (U) is designed as a defined scan volume, preferably a closed compartment, such as a vehicle interior and / or trunk, and / or - Assigning scanned tags to registered objects (o i ) and / or users (uj), where in particular an existing object (o i ) and / or an existing user (u j ) is represented by a 1 as the variable value, preferably a non-existent object (o i ) and / or a non-existent user (uj) is represented by a 0 as a variable value. [13] Method according to any one of the preceding claims, further showing: - Training a neural network (ANN) to achieve a probability (p(o i )), e.g. between 0 and 1, for each registered object (o i ) and / or any registered user (u j ) to provide, where in particular a probability (p(o i )) for an existing object (o i ) and / or can be represented by a 1 for an existing user (uj), where preferably a probability (p(o i )) for a non-existent object (o i ) and / or for a non-existent user (u j ) can be represented by a 0. [14] Method according to any one of the preceding claims, further showing: - Training a neural network (ANN) to establish a correlation between existing objects (o i ) and users present (u j ) to learn a specific list, in particular existing objects (o i ) in the environment (U) of the user(s) present (uj). [15] Method according to any of the preceding claims, further comprising: - Training a neural network (ANN) to establish a correlation between existing objects (o i ), present users (u j ) to learn a specific list and specific times (t) and / or places (x). [16] Method according to any one of the preceding claims, further showing: - Saving user-specific inventory lists to objects (o i ) and / or model parameters of a trained neural network (ANN) according to one of claims 13 to 15 as part of a personal profile, to activate user-specific inventory lists when a specific user (u j is present, and / or to enable the transfer of user-specific inventory lists. [17] Method according to any one of the preceding claims, wherein the neural network (ANN) is pre-trained according to one of the preceding claims, and / or wherein the neural network (ANN) according to one of the preceding claims is trained during operation, e.g. with each scan. [18] Method for operating an inventory system (100), in particular according to one of the preceding system claims, for an environment (U) in a vehicle (F) which has been trained according to one of the preceding method claims, comprising: - Warning a user (u j ), if an object (o i ) is missing from a user-defined inventory list and / or - Warning a user (u j ), if an additional object (o i ) is recognized. [19] Method according to the preceding claim, where if a user (u j ) a new object (o i ) to add to a user-side inventory list, furthermore showing: - Registering the new object (o i ), so that the object (o i) a name and / or an identifier is assigned, e.g. during an initial scan of the new object (o i ), especially about: - a user-side mobile device, preferably equipped with NFC, e.g. a smartwatch or smartphone, e.g. subsequently, - a special reading device, e.g. in advance, and / or - a suitable interface, e.g. directly in the environment (U) in the vehicle (F), preferably via a vehicle user interface. [20] Method according to any one of the preceding claims, where if an object (o i ) is no longer recognized for a certain period of time, the user (u j ) is asked whether the object (o i ) is to be removed from a user-side inventory list, or where, if an object (o i ) is no longer recognized for a certain period of time, the object (o i) is automatically removed from a user-defined inventory list, preferably without the user (u j to ask. [21] A method according to any of the preceding claims, further comprising: - Adapting learned inventory lists to changing user usage patterns (u j ). [22] A method according to any of the preceding claims, further comprising: - Using user-defined inventory lists on objects (o i ) and / or model parameters of a trained neural network (ANN) as part of a personal profile, to enable the activation of user-specific inventory lists when a specific user (u j is present, and / or to enable the transfer of user-specific inventory lists, and / or - Using learned inventory lists of registered objects (o i) as a, possibly additional, means for user authentication and / or for enabling vehicle functions.

Images