Methods for operating a navigation system and information technology system

Abstract

The invention relates to a method for operating a navigation system (1) wherein, after entering a navigation destination (2), information (6) about business opportunities located in the vicinity of the navigation destination (2) is output to a user.The method according to the invention is characterized in that the navigation system (1) transmits at least one piece of information (5) characterizing the navigation destination (2) to an AI model (3) and the AI ​​model (3) determines the information (6) relating to the business opportunities from the Internet depending on the information (5) characterizing the navigation destination (2) and transmits it back to the navigation system (1) for output, wherein the AI ​​model (3) is at least trained to receive and interpret the information (5) characterizing the navigation destination (2) and, depending on the interpretation result, to carry out an Internet search and provide the results found as a response.

Description

The invention relates to a method for operating a navigation system of the type defined in more detail in the preamble of claim 1 and to an information technology system of the type defined in more detail in the preamble of claim 8.The use of navigation systems for orientation is well known. There are a wide variety of variants of execution, such as vehicle-integrated navigation systems, execution as a dedicated mobile terminal, or else a purely software-based solution, such as the execution of specific navigation apps on a smartphone. A navigation system can be used to calculate a travel route from a current location to a predefined destination. At this time, the current traffic volume, diversions, and the like can be taken into account to adjust the navigation route during travel.If a user searches for an unknown region, the user is usually not familiar with the circumstances on site and does not know where relevant infrastructure facilities such as supermarkets, gas stations, shopping centers, doctors, playgrounds and the like are present. Navigation systems therefore usually offer the functionality, known as points of interest, better known as point-of-interest (POI), for example, to be presented visually highlighted as a list or in a digital road map. The user of the navigation system can view these points of interest and thus plan further participations after arriving in the destination region. In particular in cloud-based solutions, the points of interest as well as information relating to the points of interest, for example opening times, contact information, a user assessment and the like, can be updated. However, the disadvantage here is that such accommodation options have to have been classified beforehand as a point of interest.A navigation system with improved guidance of arrival at a point of interest is known, for example, from DE 10 2013 106 671 A1. The document describes an assignment of points of interest according to a parent-child relationship. Thus, multiple points of interest that are in local proximity to each other may be related to each other. For example, there may be various stores in a shopping center. The shopping center then represents the parent point of interest and the individual stores the child points of interest. A parent point of interest could also be formed, for example, by a hospital and an associated child point of interest by a parking facility. As soon as the navigation system has approached the actual navigation destination to a fixed distance, the corresponding parent-child relationships are displayed via the navigation system.Furthermore, a personal GPS navigation system is known from U.S. Pat. No. 9,250,076 B2. The publication describes the integration of a navigation functionality into a mobile terminal by a software or application executed on the mobile terminal. The navigation system allows creation of its own points of interest and the exchange of the points of interest with other users. As a result, a particularly large number of points of interest can be communicated to the users of such navigation systems. However, the disadvantage here is that relevant locations must also have been defined beforehand as point of interest.The object of the present invention is to specify an improved method for operating a navigation system, with the aid of which users of the navigation system can be informed particularly extensively about possible participations which can be made in the vicinity of a navigation destination.According to the invention, this object is achieved by a method for operating a navigation system having the features of claim 1. Advantageous embodiments and refinements and an information technology system for executing the method are evident from the claims dependent thereon.A method of the generic type for operating a navigation system, wherein information about accommodation options located around the navigation target is output to a user after input of a navigation target, is further developed according to the invention in that the navigation system transmits at least one item of information characterizing the navigation target to an AI model and the AI model ascertains the information concerning the accommodation options from the Internet as a function of the information characterizing the navigation target and transmits it back to the navigation system for output, wherein the AI model is trained at least to receive, interpret, and perform an Internet search as a function of the interpretation result and thereby deliver discovered results as a response.The method according to the invention is applied to the application of AI models for finding relevant information. For this purpose, the AI model performs an Internet search orchestration of the Internet search and delivers the results found as a response. By using the required resources such as application programming interfaces, Internet search engines and the like, the AI model is able to independently obtain information from the Internet and provide it in a prepared manner as a result. Since the respective searches are newly performed at each search request and current information is accessed, it is possible to reduce the risk that legacy information is provided as a result. The AI model is also able, because of its intrinsic mode of operation, to configure or condition the search requests for one and the same term or the results to be delivered differently, so that it is made possible to deliver different, but always relevant results for the same navigation destination. The user can thus be informed particularly extensively about the possible accommodations that can be made in the vicinity of the navigation destination. The AI model automatically decides up to which perimeter around the navigation destination relevant contents should be output. In particular, this distance is dependent on the result of the Internet search.The AI model is also capable of outputting as proposal those accommodation possibilities that have not been dedicatedly classified as a point of interest or POI. Thus, an even wider variety of locations and accommodations are available that can be output to the user. However, both the navigation destination and the possible accommodations found by the AI model can generally be a point of interest.The navigation system and the AI model are in communication with one another via suitable communication paths. In particular, the AI model is implemented externally to the navigation system; for example, the AI model is implemented on a high-performance computer or server group. Data is exchanged between the navigation system and the AI model via corresponding application programming interfaces, also referred to as application programming interface (API). Further intervening hardware and / or software components may be interposed between the navigation system and the AI model, which hardware and / or software components take over orchestration of the AI functions. This simplifies the use of a wide variety of APIs for programmers, since the programmers thus do not have to know the individual APIs in detail. Thus, the exchange between the navigation system and the AI model may be defined in a general format and is appropriately translated by the orchestrating hardware and / or software component at corresponding locations in accordance with the current level of development of the respective programming interfaces. In particular, the navigation system and the AI model are in communication connection via the Internet. The navigation system can be wirelessly connected to the Internet, for example by mobile radio, Wi-Fi and the like. For this purpose, the navigation system can have internal communication means or be connected to a telecommunications unit external to the navigation system.The navigation destination may be any location in a digital road map. These can be, for example, geo-coordinates, an address, a point of interest or the like. The information characterizing the navigation destination may include the name, address, geoposition, or the like of the navigation destination, respectively. It is only relevant that a clear assignment of the navigation destination is possible with the aid of the characterizing information.The user of the navigation system can program the navigation destination in a variety of ways. The navigation system can have, for example, a touch-sensitive screen via which the user can interact with the navigation system. The navigation destination can also be defined by voice command.The output of the information supplied by the AI model can likewise take place in a variety of ways. The information can be output visually, for example. For this purpose, the result of the AI model can be displayed as text on the display of the navigation system. Depending on the type of information, the location at which the accommodation options can be carried out can also be displayed in the digital road map. Thus, the user of the navigation system can directly program the location thus found as the next navigation destination. The accommodation possibility can also be, for example, a migration. Instead of a single location, the corresponding traveling route can then be drawn in and highlighted in the digital road map, for example. The AI model can enrich the found information with additional information, such as photographs or videos, of the corresponding participation option. The photos and / or videos may be displayed on the display of the navigation system, respectively. This visual impression informs the user particularly comprehensively about the respective accommodation possibility and can therefore decide more reliably for the respectively preferred accommodation possibility. Additionally or alternatively, the results found by the AI model can be output acoustically. Thus, the result can be formulated as text and read by a computer voice. Acoustic information such as, for example, sounds recorded on site can also be output. If the accommodation possibility is, for example, a walking on the beach, sea noise can be emitted. If, for example, the visit is to a festival, concert, theater or the like, a middle cut of a corresponding piece of music can be output and the like.To output the result information, the navigation system can also use components external to the navigation system, such as external display devices and / or loudspeakers.The AI model can be offered in particular by a service provider such as a company specialized specifically for providing AI solutions. This service provider may have trained the AI model itself during development.The results found by the AI model can generally be output at different times. For example, the output can already take place during the programming-in of the navigation destination. The output can also take place automatically only when arriving at the navigation destination or at least when a fixed distance from the navigation destination is undershot. The corresponding result information can also be collected and transmitted to the navigation system, but is only displayed or output when this is confirmed by the user by proactively entering a manual operating action. In particular, when automatically outputting the result information, user convenience is improved because less manual effort is required.An advantageous further development of the method according to the invention provides that a large speech model is used as the AI model. Such large language models are also referred to as a large language model (LLM). This is so-called generative artificial intelligence, which is capable of generating text, images or other media in response to a request. Large language models are able to interpret requests formulated as text or language semantically correctly and to derive and execute the task correspondingly provided. Results found accordingly are then semantically correctly formulated as text or voice message. Because of their memory, large speech models are therefore particularly suitable for carrying out the Internet search and editing of the results mentioned above.As mentioned above, such large speech models can provide different results for one and the same query. The navigation system may have a function to request the AI model to output alternative responses for one and the same search. Thus, for example, when the navigation destination is reached, the result found by the AI model can be output and, in addition, a button can be displayed on a touch-sensitive display device, after which a further response is generated by the AI model after the user has operated. The user can actuate this control panel several times in succession in order to always obtain new results. Alternatively, this request can also be input by the user by voice command. For this purpose, a new Internet search can be carried out each time or the found result information can also be prepared in a different manner, for example by outputting a different subset of the found information.According to a further advantageous embodiment of the method according to the invention, at least the information characterizing the navigation destination is converted by means of a software development kit into a format processable by the AI model. As already mentioned above, further hardware and / or software components that orchestrating the AI functionalities can be connected between the navigation system and the AI model. A software development kit (SDK) is a collection of programming tools and libraries that facilitate the development of software components for programmers. The service provider of the AI model can have its own software development kits, which make it easier for the users of his AI model to interact with it. The corresponding software development kit facilitates the integration of the AI-controlled large language models into conventional programming languages. The SDK combines natural language semantic functions, traditional native code based functions, and the storage and embedding of information. This allows for the use of templates for prompts, function chains, vectorized storage functionalities, and smart scheduling capabilities from the beginning. Thus, existing program routines of the navigation system can be used further unchanged and the corresponding commands for carrying out the Internet search can nevertheless be transmitted to the AI model.Preferably, at least the information characterizing the navigation target is hereby incorporated into an input template for the AI model, wherein the input template is embodied as a text readable by the AI model, which contains, in addition to the information characterizing the navigation target, at least one command to the AI model to provide information about accommodation possibilities located in the vicinity of the navigation target. As described above, large speech models are capable of semantically correctly processing text inputs. Accordingly, the order for performing the Internet search is formulated as such a text. The use of a corresponding input template allows the AI model to be instructed to carry out the respective Internet search according to a specific pattern or to transfer the found results into a specific format. This allows the consistent quality of the search and the results to be ensured. For example, the input template contains the command that is always under the same formulation. If only individual words of the command are changed, the operation of the AI model can already be different. However, when the same formulation is used, the quality of the search and the result formulation can be further improved.The information characterizing the navigation destination can likewise be input into this input template in a specific format. Thus, for example, it can be assumed that, with each request, the geo-coordinates, the country, the state of the federal, the city, the circle and / or the like in which the navigation destination is located are transmitted as information. This makes it easier to unambiguously assign the navigation destination to the AI model. If only the geo-coordinates are transmitted, the AI model can independently determine which circle, federal country, state or the like the navigation destination is located in. However, this information can also be determined independently by the navigation system or an intermediately stored component and transmitted into the respective input template. Thus, potentially feasible errors can be reduced by the AI model.Preferably, at least one secondary rule and / or format rule to be fulfilled is additionally incorporated into the input template. With the aid of such secondary rules and format rules, the search performed by the AI model and the processing of the result information can be controlled even more specifically in response. As a secondary rule, the AI model can be informed, for example, that no raceistic, malicious, acceptable or otherwise discriminating contents are to be included in the search and the respective results. Because the large language model is able to process inputs semantically correctly, such a secondary rule could also be, for example, that the respective result is to be formulated to be short and flush or is to be formulated in the manner of a travel guide. Specific format rules can also be defined, such as outputting the response as complete but simple sets without enumerations. In addition, a maximum word number can be set. In addition, the AI model may be notified that should the AI model not be able to find suitable results, an empty text string should be returned. Thus, in particular, large speech models tend to "invent" results, which is also referred to as hallucinating. By including such a format rule, the risk of false information being output to the user can thus be reduced.A further advantageous embodiment of the method according to the invention further provides that the navigation system is integrated into a vehicle and the information about accommodation options located around the navigation destination is output in the vehicle acoustically and / or visually. In general, the navigation system can be a dedicated mobile terminal or else an application running on a smartphone, for example. However, it is particularly preferably a navigation system which is fixedly integrated into a vehicle. In particular, when travelling on vehicles, regions remote from a usual location environment of a user can be searched. In such a region, for example in the vacation, the respective user is particularly poorly familiar, which is why the method according to the invention can be particularly successful here. In order to output the results supplied by the AI model, other components of the vehicle can preferably also be used, such as various display devices and / or loudspeakers. The sound system of the vehicle can thus be controlled in a targeted manner in order to output voice messages or tones or melody. As the display device, for example, the display of the instrument cluster, the head unit, a dedicated passenger display, or the like can be used. Particularly preferably, the output of the result information is carried out on the display device, on which the digital road map of the navigation system is also displayed.It may be necessary here that the result information supplied by the AI model first has to be converted into a suitable format in order to be able to be output on the corresponding components of the navigation system or of the vehicle. This is carried out accordingly by the navigation system or a component connected between the navigation system and the AI model. It can thus be ensured that the response message generated by the AI model fits from its dimensions onto a corresponding display device and can be displayed. In this case, in particular, font, font size, font and background color and the like, i.e. formatting, can be adapted to the currently set parameters of a graphical user interface, also referred to as a graphical user interface (GUI). This firstly increases the recognizability of the corresponding results and ensures a particularly high immersion of the corresponding Kl-based search function into the vehicle. This can increase the acceptance for the user.According to a further advantageous embodiment of the method according to the invention, the information exchange between the navigation system and the AI model is carried out indirectly via a central computing device. The central computing device can be a server or server group. The central computing device can likewise be connected to the Internet and thus exchange data or information with the navigation system and the AI model. In particular, the central computing device is operated by the vehicle manufacturer. The indirect information exchange via the central computing device makes it possible to manage functions required for the communication between the navigation system and the AI model at a central location. In particular, the functionalities of the software development kits are executed on the central computing device and the requests to the AI model are correspondingly incorporated by the central computing device into suitable input templates and these are transmitted to the AI model. Accordingly, the central computing device formats the results received from the AI model into a data format and / or presentation format suitable for the navigation system or the vehicle. This can reduce the hardware requirements imposed on the navigation system.In an information technology system comprising a navigation system and an AI infrastructure, according to the invention the navigation system and the AI infrastructure are configured to execute a method described above. The AI infrastructure is used to execute the AI model. In particular, the AI infrastructure is a server farm, for example a high-performance computing cluster. The navigation system and the AI infrastructure are capable of exchanging information via corresponding interfaces, such as application programming interfaces. A conversion into respectively interpretable data formats can take place on the individual subcomponents of the information technology system.An advantageous embodiment of the information technology system provides that a central computing device is connected between the navigation system and the AI infrastructure in the information flow direction. The advantages of a central computing device connected between the navigation system and the AI model or the AI infrastructure have already been described above.The information technology system preferably further comprises a vehicle which in turn has a navigation system. The vehicle can be any road vehicle such as a car, truck, transporter, bus or the like. With the aid of such a vehicle, the user is able to plan his journeys spontaneously and nevertheless in a particularly experience-rich manner, since there are also made available accommodation possibilities via which the user originally does not know anything. Since not only points of interest are displayed as accommodation options, but also an even larger data base is used as information source by the AI model by carrying out an Internet search extensively, even more diverse accommodation options can be offered to the user.In this case, the totality of the results to be output in the vehicle is to be reduced to a suitable extent in order not to overload the user. Therefore, particularly preferably, texts kept short and configured in the manner of a travel guide are presented as a result. In addition, user preferences for accommodations may be taken into account. For example, the user may specify that proposals for accommodations to be performed naturally such as travel, bicycle travel, swimming, or the like should be made preferentially. Further alternatives would be museum supervision, the visit of hidden attractions such as viewpoints not mentioned in any travel guide or visited only by a low frequency of tourist, and the like. Corresponding preferences can be included as a respective secondary rule in the input template for the large speech model. Thus, the user can enter corresponding preferences in the vehicle or via the navigation system. Preferences can be formulated freely by a user and the respective text can be typed in or input by voice, or predefined preferences can also be displayed, for example as a touchable button or list, which can be selected accordingly by the user. The navigation system transmits the preferences to the central computing device, which correspondingly adapts the input template for the AI model.Further advantageous embodiments of the method according to the invention for operating the navigation system and the information technology system are evident from the exemplary embodiments, which are described in more detail below with reference to the figures.The following are shown: FIG. 1 shows a schematic illustration of the display of a navigation system when a user arrives at a destination; FIG. 2 shows a schematic illustration of an information technology system according to the invention for carrying out a method according to the invention for operating the navigation system shown in FIG. 1 ; and FIG. 3 is a schematic illustration of an input template used to issue commands to an AI model.FIG. 1 shows the representation on a display of a navigation system 1 upon arrival of a user of the navigation system 1 at a navigation destination 2. The navigation system 1 is preferably integrated into a vehicle.After the user arrives at the navigation destination 2, the latter may wish to make further interventions. The display of the points of interest 15 can be a first aid or incentive to the user when planning his further activities. Due to the requirement that relevant locations, in particular of a developer of the navigation system 1 or of the underlying map material, have first to have been defined as points of interest 15, the total amount of representable available participation possibilities is thus limited or only generally known and probably long-lasting activities and locations are proposed. A method according to the invention for operating the navigation system 1 starts here.The method according to the invention has to present the destination to the user a variety of information about possible accommodations located in the vicinity of the navigation destination 2, so that the user can experience, in particular spontaneously, a exciting stay in the destination region. For this purpose, the navigation system 1 transmits at least one piece of information 5 characterizing the navigation target 2 (see FIG. 3 ) to an AI model 3 shown in FIG. 2 The AI model 3 in turn determines information 6 relating to corresponding accommodation options from the Internet as a function of the information 5 characterizing the navigation target 2 and returns said information to the navigation system 1 for output. In this case, the AI model 3 is trained to receive, interpret the information 5 characterizing the navigation target 2 and to perform a corresponding Internet search as a function of the interpretation result and to deliver the results found in this case as a response.The AI model 3 is preferably a large speech model. This allows processing text messages and correct semantic interpretation of the content.FIG. 1 shows that the information 6 regarding the accommodation options is represented as a text message in a box 16. In the embodiment shown in Fig. 1, the user visits a city in South Korea. The text message could in this case be, for example: "Know Gut: Visit the lock muscle and the Seongju market for a cultural experience. Enjoy the nature of Jikso waterfall and the green tea fields of Cheongseong.". By using the AI model 3, in particular in the form of a large speech model, short text messages can be composed which are designed in the manner of a travel guide. In order not to overload the user, short sentences with the indication of a few possible accommodation options are preferably included in the corresponding information 6. In addition, a marker 17 can be displayed on the digital road map 13 or on the display of the navigation system 1, which marker illustrates the position of a suggested location for the user. Corresponding text locations in the text message shown in box 16 may be selectable by the user in order to present further information about the corresponding participation option and / or directly set a navigation route with the respective participation option as a further navigation destination. The ability to select the respective text passages is indicated in FIG. 1 by underlineing the respective text passages.Further information or objects can be displayed on the display of the navigation system 1, such as a search field 18 for inputting the actual navigation target 2, a touch surface 19.1 for selecting a favorite, a touch surface 19.2 for adding route points as an intermediate stop, a touch surface 19.3 for changing the navigation target 2 and / or a touch surface 19.4 for searching points of interest in the vicinity of the navigation target 2. When the touch surface 19.4 is touched again, a new search query can then be made, for example, and a further response possibility can be obtained and displayed.FIG. 2 shows a schematic illustration of an information technology system 11 according to the invention, which comprises at least the navigation system 1 and a AI infrastructure 12 used for executing the AI model 3, in particular a high-performance computing cluster. In the embodiment shown in FIG. 2, a central computing device 10 is connected between the navigation system 1 and the AI infrastructure 12.Indicated by an arrow 201, the navigation system 1 transmits at least the information 5 characterizing the navigation target 2, for example a name, an address, geo-coordinates or the like, to the central computing device 10. For this purpose, the central computing device 10 can, for example, keep available or execute further hardware and / or software components, such as a system development kit suitable for this purpose. Thus, the central computing device 10 can fill in an input template 4 shown in FIG. 3 and transmit it as a request, indicated by an arrow 202, to the AI infrastructure 12.The AI infrastructure 12 processes this request by means of the AI model 3 and performs a corresponding Internet search. The results found are channelized to form a response and, indicated by an arrow 203, transmitted back to the central computing device 10. The central computing device 10 then converts the result obtained in this way into a native data format that can be processed by the navigation system 1 or the vehicle and, if required, into a format that can be output via the output means of the navigation system 1 or the vehicle, such as a specific size, font type and / or text color for visual information to be output on the display of the navigation system 1 or a display device of the vehicle. The response generated in this way is then transmitted, indicated by an arrow 204, to the navigation system 1 or the vehicle and output there.FIG. 3 shows an exemplary embodiment of the input template 4 that can be used for making requests to the AI model 3. the input template 4 contains, in addition to the actual (search) command 7 for the AI model 3, at least the information 5 characterizing the navigation target 2. Particularly preferably, the information 5 characterizing the navigation target 2 is subdivided into a plurality of parts, such as a name, a district, a state of a federal and a country or the like. This makes it easier for the AI model 3 to find relevant information. Optionally, further contents can be inserted into the input template 4, such as one or more secondary rules 8 and / or format rules 9.Furthermore, it is possible (not shown) for the user himself to make further changes in the input template 4, wherein these changes are carried out in the navigation system 1 or the vehicle and can be transmitted to the central computing device 10 for maintenance in the input template 4. For example, the user may define preferred accommodation options such as visit of culture events or perform activities in nature.References included in the specificationThis list of documents cited by the applicant has been produced in an automated manner and is only included for the better information of the reader. The list is not part of the German patent application or utility model application. The DPMA does not take any adhesion for any faults or omissions.Patent Literature citedDE 10 2013 106 671 A1

[0004] U.S. Pat. No. 9,250,076 B2

[0005]

Claims

Method for operating a navigation system (1), wherein, after input of a navigation destination (2), information (6) about accommodation possibilities located around the navigation destination (2) is output to a user, characterized in that the navigation system (1) transmits at least one item of information (5) characterizing the navigation destination (2) to an AI model (3) and the AI model (3) ascertains the information (6) relating to the accommodation possibilities from the Internet as a function of the information (5) characterizing the navigation destination (2) and transmits it back to the navigation system (1) for output, wherein the AI model (3) is trained at least to receive the information (5) characterizing the navigation destination (2), to interpret and perform an Internet search as a function of the interpretation result and thereby deliver discovered results as a response.Method according to Claim 1, characterized in that a large speech model is used as the AI model (3).Method according to Claim 1 or 2, characterized in that at least the information item (5) characterizing the navigation target (2) is converted by means of a software development kit into a format which can be processed by the AI model (3).Method according to one of Claims 2 to 3, characterized in that at least the information item (5) characterizing the navigation target (2) is placed in an input template (4) for the AI model (3), the input template (4) being embodied as a text which can be read in by the AI model (3) and which, in addition to the information item (5) characterizing the navigation target (2), contains at least one command (7) to the AI model (3) for providing information item (6) about possible participations located around the navigation target (2).Method according to Claim 4, characterized in that at least one secondary rule (8) and / or format rule (9) to be fulfilled is additionally incorporated into the input template (4).Method according to one of Claims 1 to 5, characterized in that the navigation system (1) is integrated into a vehicle and the information (6) about accommodation options located in the vicinity of the navigation destination (2) is output in the vehicle acoustically and / or visually.Method according to one of Claims 1 to 6, characterized in that the information exchange between the navigation system (1) and the AI model (3) is carried out indirectly via a central computing device (10).Information technology system (11), comprising a navigation system (1) and an AI infrastructure (12), characterized in that the navigation system (1) and the AI infrastructure (12) are configured to carry out a method according to one of Claims 1 to 6.Information technology system (11) according to Claim 8, characterized bya central computing device (10) which is connected in the information flow direction between the navigation system (1) and the AI infrastructure (12) and is set up to carry out a method according to Claim 7.Information technology system (11) according to Claim 8 or 9, characterized bya vehicle comprising the navigation system (1).

Images