METHOD FOR DETERMINING A VALUE REPRESENTATIVE OF A CONTEXT PARAMETER WITH RESPECT TO A TARGET PROPOSAL OF A NAVIGATION SYSTEM FOR A MOTORIZED TERRESTRIAL VEHICLE

Description

[0001] The present invention relates to the field of navigation systems for motor vehicles. In particular, the invention relates to a method for determining at least one representative value of at least one contextual parameter in relation to a destination suggestion provided by a navigation system for a motor vehicle. The invention is particularly applicable to motor vehicles.

[0002] Today, navigation systems for motor vehicles include destination prediction features, based in particular on machine learning mechanisms that, among other things, allow the determination of the most likely destinations given, for example, a known departure location and a given date (and time). These predictions are generally based on the history of known journeys and search for destinations most frequently reached under the same departure conditions.

[0003] However, the success rate of prediction algorithms is not 100%. Errors can occur, particularly at the beginning of the learning phase, even when measures are in place to ensure, for example, that a prediction is not made if the calculated probability is too low.

[0004] However, a key element for user acceptance of any artificial intelligence-based system is the user's ability to follow and understand what the machine has learned. Therefore, there is a need to present the proposals and results provided by such systems alongside explanatory or contextual elements that allow users to understand the reasoning behind them and, where applicable, the reasons for any errors. These additional pieces of information help encourage users to be more understanding of errors and attentive to progress. It is through such features that a higher rate of acceptance of artificial intelligence mechanisms by end users can be achieved.Thus, current navigation systems that do not take such aspects into account and provide destination predictions without adding informative elements that allow the user to understand the framework of the reasoning implemented present certain disadvantages in terms of human-machine interaction.

[0005] The invention aims to overcome these drawbacks. In particular, the invention aims to provide a method capable of linking a destination prediction with contextual information presented to a user through a motor vehicle's navigation system. More specifically, the invention aims to provide a method that ensures any suggestion made based on the implementation of artificial intelligence mechanisms is accompanied by at least one relevant contextual clue, enabling the user to understand the reasoning behind that suggestion. The invention also aims to provide a method that can extract certain influential distinctive parameters from the results provided using machine learning mechanisms and process these distinctive parameters to provide useful explanatory information for the user.

[0006] To this end, the invention relates to a method for determining, by a computer system, at least one representative value of at least one contextual parameter with respect to a destination suggestion provided by a navigation system for a motor vehicle, the method comprising the steps of: to constitute operating data, said operating data being generated following the emission of at least one signal by at least one first sensor arranged in the vehicle and configured to emit a signal according to a vehicle operating regime, to receive location data, said location data being transmitted by a satellite positioning system and / or a remote data source, to constitute identification data relating to at least one passenger of the vehicle, said identification data being partly generated following the emission of a signal by at least one second sensor arranged in the vehicle and configured to emit a signal according to at least one component of a physical condition of said passenger, to transmit said operating data, said location data and said identification data to a machine learning module which,Based on the aforementioned operational data, location data, and identification data, establishes a predictive model capable of processing at least one input parameter, uses the predictive model to determine the suggestion, determines at least one distinctive parameter related to the suggestion, and determines the representative value using the distinctive parameter.

[0007] According to one variant, the process may include a step of using the representative value to determine a contextual indication.

[0008] According to another variant, the process may include a step of interaction with an information dissemination interface capable of supporting contextual indication to produce a sound message and / or an animation.

[0009] According to another variant, the said information dissemination interface can be an integral part of said navigation system.

[0010] According to another variant, the step of determining the distinctive parameter may include a step of varying the input parameter.

[0011] According to another variant, the step of using the predictive model may include a step of using location data to update said suggestion.

[0012] The invention further relates to a computer system for determining at least one representative value of at least one contextual parameter with respect to a destination suggestion provided by a navigation system for a motor vehicle, in which are arranged at least one first sensor configured to emit a signal according to a vehicle operating regime, at least one second sensor configured to emit a signal according to at least one component of a passenger's physical condition, a receiver for receiving location data provided by a satellite positioning system and allowing the time-stamped determination of the vehicle's geographic position, and a means for transmitting and receiving radio frequency signals to communicate with a remote server and / or a portable electronic device, the system comprising means for implementing a method as defined above.

[0013] According to one variant, the system may include a computing unit, a machine learning module and storage means in which at least one program is stored for the execution of steps of the determination process as defined above.

[0014] The invention further relates to a computer program comprising instructions for executing the steps of a process as defined above.

[0015] The invention further relates to a vehicle comprising a system as defined above.

[0016] Other features and advantages of the invention will become apparent upon examination of the detailed description below, and the accompanying drawings, in which: there figure 1 is a block diagram of a determination system according to the invention, and the figure 2 is a flowchart illustrating certain steps of a determination process according to the invention.

[0017] The 100 determination system according to the present invention is illustrated in the figure 1 .The system comprises an information processing unit 101 including one or more processors, data storage means 102, input and output means 103, and a machine learning module 104. In some embodiments, the system 100 is embedded in a motor vehicle and distributed among one or more computers. In other embodiments of the invention, the system 100 comprises one or more computers, one or more servers, one or more supercomputers, and / or any combination thereof. Other embodiments may also be envisaged in which some elements of the system 100 are hosted partly on board a motor vehicle, on one or more computers, while other elements are distributed across one or more remote servers.

[0018] Regardless of the configuration chosen, System 100 is capable of interacting with a vehicle, specifically to extract and / or receive data generated by one or more sensors installed in the vehicle. Advantageously, System 100 interacts with any system integrated into the vehicle, including systems and sensors that identify the driver and any additional passengers, as well as analyze their physical condition. For certain tasks, System 100 interacts, for example via its input and output means 103, with a receiver of the vehicle's satellite positioning system, with the vehicle's onboard navigation system, and / or with a radio frequency signal transmitter / receiver in the vehicle, particularly to extract data from one or more remote servers.

[0019] In addition, all these elements contribute to enabling system 100 to implement a process for determining at least one representative value of at least one contextual parameter with respect to a destination suggestion provided by a navigation system for a motor vehicle.

[0020] As illustrated in the figure 2According to step 201, the system 100 constitutes, i.e., extracts, records, and / or establishes, operating data. Operating data is generated when one of the sensors arranged in the vehicle emits a signal according to a vehicle operating regime. For example, during each journey, operating data is generated according to the signals received from sensors that are capable of emitting signals according to the fuel level in the tank, the battery charge level, the ambient light level, the presence of raindrops on the windshield, or the operating regime of a lighting component or a windshield wiper of the vehicle.

[0021] According to step 202, the 100 system also receives location data. As is generally the case, this location data is transmitted via a satellite positioning system and / or a data source hosted on a remote server. This location data may therefore contain map elements, traffic information, or other information such as points of interest or road signs. Advantageously, the location data received by the 100 system does not exceed the information elements currently available and used by existing navigation systems.

[0022] According to step 203, location data is processed with regard to a prediction task to be performed, in particular to determine certain input parameters which feed into the machine learning module 104. During this step, the system 100 determines in particular a current date, a current time and / or a current location.

[0023] According to step 204, system 100 generates identification data relating to at least one passenger in the vehicle. This identification data concerns both the driver and, where applicable, other passengers in the vehicle.

[0024] In the first instance, identification data is generated when System 100 interacts with a portable identification device that the driver uses to unlock and start the vehicle (e.g., remote control, electronic key, etc.). Alternatively, or cumulatively, System 100 generates another portion of the identification data by interacting directly with a smartphone, for example, via a Bluetooth connection provided by the vehicle's radio frequency signal transmission system. Advantageously, System 100 uses this system to extract identifiers from certain portable electronic devices located in the vehicle, including smartphones.

[0025] Based on this data, system 100 is able to determine certain information about the vehicle's passengers, in particular the number of passengers transported. Advantageously, this identification data also allows system 100 to retrieve and extract a user profile, for example, stored on data storage devices 102. In this user profile, some information defines a passenger's identity, while other information relates to their age or driving experience.

[0026] In another part, identification data is generated following the emission of a signal by at least one second sensor located in the vehicle and configured to emit a signal based on at least one component of a passenger's physical condition. For example, during each trip, identification data is generated based on signals received from sensors capable of emitting signals based on a passenger's pulse, body temperature, weight, level of intoxication, and / or gaze position, particularly that of the driver. In step 205, the operating data, location data, and identification data are transmitted to the machine learning module 104. Initially, this data is used by the machine learning module 104 to establish a training database.Gradually, the machine learning module 104 builds a predictive model that becomes more refined over time. To do this, the system 100 implements, for each journey, an analysis step of the input parameters, for example the current date, the current time or the current location, in relation to a destination actually reached at the end of the journey.

[0027] In step 206, the system 100 uses the predictive model established by the machine learning module 104 to determine a destination suggestion. Based on one or more common input parameters, the machine learning module 104 returns a destination suggestion.

[0028] In a more specific embodiment, system 100 simultaneously performs an additional step in which the suggestion returned by the machine learning module 104 is updated using location data. Through this mechanism, system 100 is able to take into account that, in some cases, location data takes precedence over the suggestions provided by the machine learning module 104. This is the case, for example, when, over time, the machine learning module 104 has learned that certain departure locations inevitably imply certain destination locations (for example: "departure = work" implies "destination = home"; "departure = children's school" implies "destination = work").

[0029] During step 207, system 100 determines at least one distinctive parameter related to the suggestion determined during step 206. To do this, system 100 performs a step that involves varying one or more input parameters. Specifically, by checking whether the prediction returned by the predictive model of the machine learning module 104—i.e., the suggestion—is impacted by relative variations in certain input parameters, system 100 can determine at least one distinctive parameter that influences the suggestion. For example, the system determines, with respect to a destination suggestion, whether it remains the same by varying the input parameter that corresponds to the current date. If, for example, a suggestion remains the same every day of the week except the weekend, or only every Monday morning, system 100 determines that the distinctive parameter is related to the current date.

[0030] During step 207, System 100 determines the representative value using the distinguishing parameter. For example, when the distinguishing parameter relates to the current date, as above, System 100 determines that a representative value for this distinguishing parameter corresponds to a verbal formulation such as "every day except the weekend" or, in another example, "every Monday morning." Other representative values ​​are possible, however. Thus, when the distinguishing parameter relates to the current location, System 100 could determine a representative value that corresponds to a verbal formulation such as "whenever this location is the starting location." Similarly, if the distinguishing parameter relates to the fuel level in the tank, System 100 determines a representative value that corresponds to a verbal formulation such as "whenever your tank is almost empty."As we have understood, the representative value therefore defines a context parameter related to the suggestion.

[0031] In one particular embodiment, System 100 uses the representative value determined in step 207 to derive a contextual clue. For example, the contextual clue determined by System 100 in this step might correspond to a sequence of words similar to the verbal formulations mentioned above. In other cases, the contextual clue corresponds to a set of numerical data that define an image, a video, and / or an audio message.

[0032] In another step, once the contextual information is determined, it is transmitted to an information display interface. The information display interface is configured to use the contextual information and produce a signaling element, such as an audible message or a visual animation. Preferably, the information display interface is an integral part of the vehicle's navigation system. Alternatively, the information display interface is a separate electronic device located in the vehicle's console, comprising a screen and / or a speaker. In other cases, it is conceivable that one of the information display interfaces of a smartphone (e.g., screen, speakers) located in the vehicle is used at this stage by the system.

[0033] The system according to the invention thus significantly improves its potential for user acceptance, as it becomes possible to provide a navigation system that not only learns habits and suggests relevant destinations but is also able to justify the suggestions it provides. Users, who better understand the processes and reasoning implemented by the artificial intelligence functionalities, are then more understanding of potential errors and more receptive to progress. This advancement in human-machine interfaces for automotive navigation relies on the method and system according to the invention, which, as described above, establish a specific selection of parameters and a particular way of processing these parameters in order to obtain actionable results related to the desired improvement goal.

Claims

1. Method for determining by a computer system at least one value representative of at least one contextual parameter facing a destination suggestion provided by a navigation system for an engine land vehicle, comprising the steps of: - constituting operating data, said operating data being generated following the emission of at least one signal by at least one first sensor arranged in the vehicle and configured to emit a signal as a function of an operating speed of the vehicle, - receiving location data, said location data being provided by a satellite positioning system, enabling the time-stamped determination of the geographical position of the vehicle, - constituting identification data relating to at least one passenger of the vehicle, said identification data being generated in part following the emission of a signal by at least one second sensor arranged in the vehicle and configured to emit a signal as a function of at least one component of a physical condition of said passenger, - transmitting said operating data, said location data and said identification data to a machine learning module which, on the basis of said operating data, said location data and said identification data, establishes a predictive pattern capable of establishing at least one entry parameter, - using the predictive model to determine said suggestion, on the basis of at least one entry parameter, - determining at least one parameter influencing the suggestion, called distinctive parameter, in connection with said suggestion, - determining the representative value using the distinctive parameter, wherein the method comprises a step of using the representative value to determine a contextual indication, wherein the method comprises a step of interaction with an information broadcasting interface capable of taking into load the contextual indication to produce a sound message and / or an animation, and in which the step of determining the distinctive parameter comprises a step consisting in varying the parameter of entry.

2. Method as claimed in claim 1, wherein said information broadcasting interface forms an integral part of said navigation system.

3. Method as claimed in claim 1, wherein the step of using the predictive location comprises a step of using the data of the predictive model to update said suggestion.

4. Method according to claim 1, wherein the at least one entry parameter is selected from the previous group: a current date, a current time, or a current location.

5. Computer system for determining at least one value representative of at least one contextual parameter facing a destination suggestion provided by a navigation system for an engine land vehicle in which are arranged at least one first sensor configured to emit a signal as a function of an operating regime of the vehicle, at least one second sensor configured to emit a signal as a function of at least one component of a physical condition of a passenger, a receiver for receiving location data provided by a satellite positioning system and allowing the time-stamped determination of the geographical position of the vehicle and a means for emitting and receipt of radiofrequency signals for communicating with a remote server and / or a mobile phone electronic apparatus, characterised in that it comprises means for implementing a method according to any one of the previous claims.

6. System as claimed in claim 1, previous in that it comprises a compute unit, an automatic learning module and storage means in which are stored at least one plan for the execution of steps of the determination method implemented by the system.

7. Land vehicle with an engine, comprising a system according to one of claims 5 or 6.

8. Computer plan comprising instructions for executing the steps of a method according to any one of claims 1 to 4.