WO2022017737A1

WO2022017737A1 - Interactive system and associated interaction method

Info

Publication number: WO2022017737A1
Application number: PCT/EP2021/067804
Authority: WO
Inventors: Georges De Pelsemaeker
Original assignee: Valeo Systemes Thermiques
Priority date: 2020-07-21
Filing date: 2021-06-29
Publication date: 2022-01-27
Also published as: FR3112747A1

Abstract

The invention relates to an interactive system (1) including a measuring device (5) designed to acquire at least one physiological parameter of a user (U) during at least one sequence of an interactive exercise, and a processing unit (7) designed to be connected to the measuring device (5), comprising an interpretation module designed to receive the physiological parameter and to define, based on the physiological parameter, an item of data representative of the physiological state of the user (U) by means of a corresponding evaluation model. According to the invention, the processing unit (7) includes an analysis module designed to identify, depending on the representative item of data, an adjustment of at least one parameter of at least one sequence of the exercise to be applied, depending on at least one correlation law, and a correction module designed to modify the parameter of the sequence of the interactive exercise according to the identified adjustment. The invention also relates to an interaction method at least partially implemented by such an interactive system.

Description

INTERACTIVE SYSTEM AND ASSOCIATED INTERACTION METHOD

The invention relates to an interactive system with at least one particular user occupying a motor vehicle. The invention also relates to a method of interaction with a user occupying a motor vehicle.

[0002] The transition to an autonomous or semi-autonomous vehicle presents more and more issues, challenges, as regards the creation of an intelligent, intuitive vehicle, making it possible to make driving and/or traveling more pleasant for the occupant such as the driver or passengers. In particular, improvements can be focused on safety, comfort and entertainment or infotainment. In such autonomous or semi-autonomous vehicles, passengers and drivers have more free time. There is an interest in using this free time in the vehicle in a positive way.

[0003] Certain interactive systems make it possible to evaluate the physiological or emotional state of an occupant of a motor vehicle. It is in fact known to carry out measurements making it possible to determine various parameters concerning an occupant, and thus to deduce therefrom data representative of his physiological or emotional state.

[0004] In general, the measurements can be carried out using portable devices making it possible to collect data such as the conductivity of the skin, the temperature, or even the heart rate. The system used generally comprises a camera, for example of the GoPro® type.

[0005] In known systems, the parameters obtained are then used to determine the emotional state of the occupant. A known model makes it possible to obtain, from the different parameters, a map of the emotional state of the passenger. Thus, the latter can be represented by a point in a space, for example in two dimensions, formed of an axis of abscissas corresponding to the valence (intrinsically pleasant or unpleasant quality of a stimulus or of a situation) and an axis of ordinates corresponding to the arousal (strength of the emotional stimulus).

The invention aims to improve the known systems by proposing an interactive system making it possible not only to model the physiological, emotional and cognitive state of a user occupying a vehicle, for example, but also to interpret it in such a way as to use in order to personalize the interaction with the user, in particular to improve his free time in the motor vehicle. [0007] To this end, the subject of the invention is an interactive system with at least one user comprising a measuring device configured to acquire at least one physiological parameter of G user during at least one sequence of an interactive exercise performed by the user through the interactive system, and a processing unit configured to be connected to the measurement device, comprising an interpretation module configured to receive said physiological parameter and to define, from said physiological parameter, data representative of the state of the user by means of a corresponding evaluation model.

[0008] According to the invention, the processing unit comprises an analysis module configured to identify, as a function of the data representative of the user's state defined, an adjustment of at least one parameter of at least at least one sequence of the exercise to be applied, according to at least one correlation law making it possible to link said representative datum to an adjustment, and a correction module configured to modify the parameter of the sequence of the interactive exercise according to the adjustment identified.

[0009] The adjustment of one or more sequences of the interactive exercise makes it possible to specifically personalize this interactive exercise for the user according to his state.

[0010] In the present, the term "state" means a physiological state, an emotional state, a cognitive state, or even a combination of several states. Said at least one datum representative of the user's state can be datum representative of the physiological state, datum representative of the emotional state, datum representative of the cognitive state, an aggregate or a combination of the preceding. Said representative datum can be obtained from one or more state variables which can be determined from direct measurements by the measuring device.

[0011] Said interactive system may also include one or more of the following characteristics, taken separately or in combination.

The data representative of the state of the user corresponds for example to a point in a characterization space of the state of G user.

[0013] In particular, the space for characterizing the emotional state can be formed of at least a first axis corresponding to the valence, and a second axis corresponding to the arousal. [0014] According to a variant embodiment, the space for characterizing the emotional state is a three-dimensional space formed by a first axis corresponding to the valence, a second axis corresponding to the arousal and a third axis corresponding to dominance.

[0015] The correlation law is defined from at least one element among a matrix or a correlation table, a decision tree, a transfer function. The correlation law can be recorded on a memory medium of the processing unit and/or be accessible from a server.

[0016] The correlation law is predefined. Alternatively, the correlation law is adaptive, and is intended to be adapted by learning.

The correction module is configured to apply a corrective action during at least one sequence of the interactive exercise, when the application is executed.

The measuring device is configured to acquire at least one physiological parameter of G user during the execution of the corrected sequence according to the corrective action.

The analysis module is configured to analyze the reaction of the user following the corrective action, in particular according to said at least one physiological parameter of the user acquired during the execution of the corrected sequence and at least one reference datum. The correction module is configured to adapt the corrective action according to the analysis results.

[0020] Advantageously, the interpretation module is configured to receive said at least one physiological parameter acquired during the execution of the sequence corrected according to the corrective action and to define data representative of the state of the user during the execution of the sequence corrected according to the corrective action.

[0021] The analysis module is configured to compare the data representative of the state of the user during the execution of the corrected sequence according to the corrective action with said at least one reference datum. The analysis module can be configured to compare the evolution of a state variable obtained from said at least one physiological parameter acquired during the execution of the sequence corrected according to the corrective action with said at least one datum of reference. [0022] The correction module is configured to adapt the corrective action according to the comparison results. This makes it possible to define or adapt the adjustment defined by the correlation law by learning.

The correction module is configured to select the corrective action randomly. Alternatively, the correction module is configured to select in a controlled manner the corrective action to be applied.

[0024] Said at least one reference datum is chosen from datum representative of a target state, datum representative of the previous state of the user during the execution of a previous or current sequence, a threshold value data representative of a state, or a threshold value of at least one state variable of the user as a function of at least one physiological parameter, a target value of at least one state variable, or still a previous value of the state variable during the execution of a previous or current sequence.

[0025] The analysis module is configured to determine, based on the comparison results, whether the corrective action improves the state of the user during the execution of the corrected sequence.

[0026] The correction module is configured to increase the corrective action if the user's condition is improved. Otherwise, the correction module is configured to reduce or stop the corrective action, during the current sequence or a subsequent execution of the sequence.

[0027] The analysis module is configured to identify at least one piece of time information and at least one piece of data relating to a deviation amplitude of the data representing the state of the user (or of a state variable of the user) in relation to the reference data.

The interactive system comprises at least one actuator configured to activate at least one stimulus to interact with the user, said stimulus making it possible to modify the state of the user.

The processing unit comprises a data communication module with a server in which is stored at least one reference data library, in particular data evolution models representative of target states for at least one sequence of an interactive exercise. [0030] The correction module is configured to adjust at least one parameter of at least one sequence of the interactive exercise from among the type, duration, rhythm, intensity.

The interactive exercise is preferably a remedial exercise to be carried out by the user to modify his state.

[0032] The interactive exercise comprises a predefined number of sequences from among a sequence of choices in an application menu, a relaxation or preparation sequence, a sequence for carrying out the remediation exercise by the user, a explanation sequence of the remediation exercise, and a user feedback sequence on the remediation exercise carried out and possibly on the sequence adjustment applied.

[0033] The correction module is configured to adjust the parameter of at least one sequence from among the relaxation or preparation sequence, the sequence for performing the remediation exercise by the user, the sequence for explaining the remediation exercise, and the user feedback sequence on the remediation exercise performed.

The measuring device comprises at least one sensor configured to acquire at least one physiological parameter of the user. Said sensor can be a sensor without contact with the user, chosen from at least one ultra-wideband radar, an infrared camera, a sensor in the visible range, a microphone.

The interactive system is configured to interact with a user occupying a motor vehicle, the interactive system being at least partly embedded in the motor vehicle. The measuring device can be embedded in the motor vehicle. The processing unit can be embedded in the motor vehicle. The interface can be embedded in the motor vehicle or alternatively be integrated into a mobile terminal configured to be connected to the processing unit.

The interactive system comprises at least one actuator to activate at least one stimulus so as to interact with the user's multi-sensory environment.

The invention also relates to a method for interacting with at least one user. Such a method is intended to be implemented at least in part by an interactive system as described previously. Said method comprising a phase of execution of an application of an interactive exercise to be carried out by G user according to at least one sequence.

Said method includes a measurement phase of at least one physiological parameter of G user by a measurement device of the interactive system.

[0040] Said method comprises a phase of reception and interpretation of said at least one physiological parameter by an interpretation module of a processing unit of the interactive system, to define, from said at least one physiological parameter, a datum representative of the user's state by means of a corresponding evaluation model.

[0041] Said method comprises an analysis and learning phase by an analysis module of the processing unit to identify, depending on the data representative of the defined user's state, an adjustment of at least at least one parameter of at least one sequence of the interactive exercise to be applied, as a function of at least one correlation law making it possible to link said representative datum to an adjustment.

Said method comprises an adjustment phase by a correction module of the processing unit, to modify the parameter of the sequence of the interactive exercise according to the identified adjustment.

Said method may include a step for applying a corrective action during at least one sequence of the interactive exercise, when the application is executed, by the correction module. Corrective action can be random. Alternatively, the corrective action is selected in a controlled manner according to a graph or a decision tree. The objective of the corrective action is to allow an analysis of the reaction of G user and to play on this corrective action in an iterative way to define by learning the adjustment to be applied according to the state of the user.

Said method may comprise a step for acquiring at least one physiological parameter of the user during the execution of the corrected sequence according to the corrective action, by the measuring device.

[0045] Said method may comprise steps for analyzing the reaction of the user following the corrective action, as a function of said at least one physiological parameter of the user acquired during the execution of the corrected sequence and of at least a reference datum, and to adapt the corrective action according to the analysis results. Said method may comprise a step for receiving said at least one physiological parameter during the execution of the corrected sequence according to the corrective action and defining data representative of the state of the user during the execution of the sequence corrected according to the corrective action, by an interpretation module of the processing unit.

[0047] Said method may comprise a step for comparing the data representative of the state of G user during the execution of the corrected sequence according to the corrective action with at least one reference data, by the analysis module of the 'processing unit.

Said method may include a step for adapting the corrective action to be applied during a subsequent execution of the sequence of the interactive exercise, depending on the comparison results, by the correction module of the processing unit .

The method comprises a sequence of choices in a menu of the application, a sequence of relaxation or preparation, a sequence of carrying out the interactive exercise, for example of remediation, by the user, a sequence of explanation of the remediation exercise, a user feedback sequence on the remediation exercise carried out and possibly on the adjustment applied.

Other characteristics and advantages of the invention will appear more clearly on reading the following description, given by way of illustrative and non-limiting example, and the appended drawings, including:

[0051] [Fig. 1] Figure 1 very schematically represents the functional blocks of an interactive system with a user in a motor vehicle and a remote server.

[0052] [Fig. 2] Figure 2 shows another schematic representation of the interactive system of Figure 1.

[0053] [Fig. 3] Figure 3 is a schematic representation of successive sequences of a remediation exercise.

[0054] [Fig. 4] Figure 4 is a schematic representation of successive sequences of a remediation exercise adapted for respective users.

[0055] [Fig. 5] Figure 5 is a schematic diagram of at least certain phases of an interaction method intended to be implemented at least partially by the interactive system of Figure 1 or 2. [0056] [Fig. 6] FIG. 6 is a schematic representation of a three-dimensional space for characterizing the emotional state of a user.

[0057] [Fig. 7] Figure 7 is a schematic representation of a mapping of a user's emotional state in the three-dimensional emotional state characterization space.

[0058] [Fig. 8] Figure 8 is a graphical representation showing a user state evolution curve and an evolution curve representing a target state over time during a sequence.

[0059] [Fig. 9] Figure 9 schematically represents a decision graph matching a user state to a corrective action to be applied to one or more of the sequences.

In these figures, identical elements bear the same references.

The following embodiments are examples. Although the description refers to one or more embodiments, this does not necessarily mean that each reference is to the same embodiment, or that the features apply only to a single embodiment. Simple features of different embodiments can also be combined or interchanged to provide other embodiments.

In the description, certain elements can be indexed, such as for example first element or second element. In this case, it is a simple indexing to differentiate and name elements that are close but not identical. This indexing does not imply a priority of one element over another and such denominations can be easily interchanged without departing from the scope of the present description. Nor does this indexing imply an order in time.

[0063] INTERACTIVE SYSTEM

Referring to Figure 1, the invention relates to an interactive system 1 with one or more users U. In particular, these are users occupying a motor vehicle 100, such as the driver or one or more passengers. In this case, the interactive system 1 is advantageously at least partly embedded in the motor vehicle 100. Such an interactive system 1 makes it possible to implement, at least in part, a method of interaction with the user U, described in more detail below.

The interactive system 1 makes it possible to supervise one or more sessions of an interactive exercise, for example a remediation exercise, which can be carried out by a user U in the motor vehicle 100 to improve his condition. As used herein, "state" means a physiological state, an emotional state, a cognitive state, or a combination of these states.

The remediation exercise can be a meditation exercise or a relaxation exercise. Consideration could be given to recreational exercise in an area of interest to the user U. It may alternatively be exercise in a primarily therapeutic setting. In particular, therapeutic exercise may be exercise of repetitive movements or repetitive actions that does not require the physical presence of a health specialist, therapist, nurse or doctor. For example, without limitation, a therapeutic exercise can be chosen from one of the following areas: speech therapy, optometry, orthoptics, rehabilitation, or even physiotherapy, concentration, breathing. In general, the interactive exercise can be performed according to a predefined number of sequences.

As described in more detail below, this interactive system 1 makes it possible to adapt, to personalize, the interactive exercise, in particular remediation, in particular at least one sequence of the exercise, to the state of the user U. This adaptation can be done by applying one or more corrective actions to at least one sequence of the interactive exercise (as explained below).

Furthermore, the interactive system 1 can be configured to select and automatically propose an application of a remediation exercise to the user U which corresponds to his condition.

One or more elements of the interactive system 1 can be configured to communicate, exchange data with a server 11. This is in particular a remote server 11.

In this server 11, one or more libraries can be stored. Mention may be made, for example, of at least one library of models for the evolution of target values, reference values, for parameters linked to the user, in particular physiological parameters or state variables, or data representative of target states , for at least one sequence of an interactive exercise and/or according to stimuli applied during this exercise.

In this server 11 can also be stored at least one library of applications, for example for the management of interactive exercises, in particular remedial exercises, at least one piece of data representative of a physiological and/or emotional and/or cognitive state. being for example associated with each application. The applications are intended to be downloaded on demand into the motor vehicle 100.

[0072] The interactive system 1 can still allow G user U to be in communication with a medical or health team 13.

For these purposes, the interactive system 1 may comprise one or more interfaces 3 for interacting with a user U. The interactive system 1 comprises in particular a measurement device 5, and a processing unit 7.

[0074] In addition, the interactive system 1 can advantageously play on the multi-sensory environment of the user U, in particular so that he has a better perception of the remediation exercise. To this end, the interactive system 1 may further comprise at least one actuator 4 configured to activate at least one multi-sensory stimulus to interact with G user U, so as to modify the state of G user U.

One or more elements of the interactive system 1 can be embedded in the motor vehicle.

One or more elements of this interactive system 1 are connected or intended to be connected together. “Connected” means the fact that a computer link is established between at least two elements, so as to be able to exchange data. This can be a wired or wireless connection. The exchange of data can for example be done by a CAN network for "Controller Area Network" in English, LIN BUS, Wifi, Bluetooth, or others.

The interface 3 or at least one of the interfaces can be embedded in the motor vehicle 100. Such an interface 3 is generally called a man-machine interface, known by the acronym HMI for "human-machine interface" in English. .

[0078] As an alternative or in addition, the interactive system 1 can include a mobile terminal 9, such as a smart phone known by the English name “smartphone”. The mobile terminal 9 can be configured to be connected to the unit of processing 7. Such a mobile terminal 9 may comprise an interface 3 for interacting with the user U. The interaction may take place by sound and/or by image and/or by tactile feedback.

According to another alternative or in addition, the interactive system 1 can have one or more interfaces 3 which can be, for example, in the form of one or more so-called smart or connected portable devices, such as a watch and/or glasses and/or a bracelet and/or a belt and/or a shirt, intended to be worn by the user U.

[0080] In a non-exhaustive manner, at least one command input interface and one output interface may be provided. The output interface may or may not be the command input interface.

In particular, the interactive system 1 may include at least one display interface 3. A display interface allows for example the display of a drop-down menu through which the user can navigate and make a choice among several options. A display interface 3 in the form of a screen which may or may not be touch-sensitive, may also be provided. The touch screen makes it possible to act both as a command input interface by pressing, or user contact U, and as an output interface by display.

The interface 3, or at least one of the interfaces provided, makes it possible to receive one or more notifications for the attention of the user U.

The actuator(s) 4 are advantageously embedded in the motor vehicle 100. In order to interact with the user U by multi-sensory stimuli, one or more actuators 4 can activate at least one item of equipment, a physical element, in the vehicle automobile 100. The multi-sensory stimuli are for example based on the five senses: sight, smell, touch, hearing, taste. The multi-sensory stimuli used can, for example, be: a. for the view: the use of interior lighting with lights of different colors and/or variable intensities; b. for the sense of smell: the use of a nebulizer or a diffuser for the diffusion of perfume, essential oils or even pheromones; vs. for touch: the activation of massage or vibration of equipment in contact with the user U, or even the use of radiant panels, water nebulization, use of hot and/or cold air jets; d. for hearing: the selection of sounds, or music or background noises; e. for taste: the diffusion of perfume or nebulization that can create pheromones combining the five main flavors, namely salty, acid, bitter, sweet, umami. Other types of stimuli can be considered.

As shown schematically in Figure 2 for illustrative and non-limiting purposes only, several actuators 4 are provided, such as an actuator 41 to control a display on a display interface such as a screen, an actuator 42 to control a sound system, loudspeakers, in order to modify the sound environment, an actuator 43 to control a lighting device so as to modify the lighting and the light environment for example in the passenger compartment, a actuator 44 for controlling the adjustment of at least one aeraulic parameter of an air flow intended to pass through the passenger compartment, such as the temperature and/or the air flow, or the adjustment of thermal parameters of at least least equipment or physical element in the vehicle such as the steering wheel or a seat, so as to play on the thermal environment, an actuator 45 to control the diffusion of perfume, of fragrance, so as to modify the surrounding odor. It is also possible to provide one or more actuators to control the actuation of a massage and/or of a vibration, for example via equipment of the vehicle such as the steering wheel or a seat.

One or more actuators 4 make it possible to activate one or more stimuli according to a command from the user via the interface 3. For example, if the user validates via the interface 3 a command to execute a remediation exercise, one or more actuators 4 can be activated.

[0086] Concerning the measuring device 5, it can be embedded in the motor vehicle. It comprises one or more 50, 51, 52, 53, 54 sensors. They are preferably sensors embedded in the motor vehicle. Their presence in the vehicle may be linked to infotainment in the vehicle, to safety and comfort factors.

The sensors 50, 51, 52, 53, 54 of the measuring device 5 are in particular configured to implement one or more measurement phases of the interaction method described afterwards. The sensors 50-54 of the measuring device 5 are configured to acquire, record, at least one parameter linked to the user or several users.

The user-related parameters may relate to parameters obtained by direct measurements, such parameters are also called primary variables. In this case, the parameters linked to the user include in particular at least one physiological parameter of the user. In a non-exhaustive and non-limiting way, the physiological parameters can be vital signs, such as heart rate, heart rate, pulse, breathing, respiratory rate and/or amplitude, the ratio between inspirations and expirations, or the coloring of the skin on the face to detect a blood flow. Other physiological parameters may be the level of perspiration of the user, the conductivity of the skin, the position of the limbs and/or of the head, the orientation of the head, the orientation of the gaze, the position of the pupils , pupil diameter, skin temperature, voice, tremor.

[0089] The measuring device 5 can also record other parameters linked to G user, such as clothing, clothing temperature, gender, age.

The acquisition or determination of parameters related to the user can be done at any time, in particular before, during or after the execution of an application, by means of the interactive system 1. It can be done before , during or following a corrective action so as to adapt by learning at least one sequence of the interactive exercise (as described below). It can still be done depending on whether the multi-sensory environment is modified or not, i.e. for example before and/or during and/or after the application of a stimulus.

One or more of the sensors 50-54 can be sensors without contact with the user. Alternatively, they may be contact sensors, i.e. intended to be in contact with the user to perform a measurement.

[0092] The sensors can include vital sign sensors 50, one or more microphones 51 for recording the voice of the user or users in the motor vehicle, one or more cameras 52, 53, 54, for example in the field of the seen.

The sensors, in particular of vital signs 50, can for example take the form of contactless sensors such as for example a radar or a camera. One or more cameras can be provided to identify facial expressions or postures of G user U. At least one camera can be a time of flight camera 54 or "time of flight" in English. One or more cameras are for example directed towards the expected positions of the various occupants of the vehicle: driver's seat, passenger seat, rear seat. In particular, one or more very wide-angle cameras (for example of the “fish-eye” type known under the name “fish-eye” in English) can cover several positions simultaneously.

In a preferred embodiment, one or more cameras may in particular be infrared cameras 52, 53 configured to take images in the infrared domain. Infrared cameras preferentially detect wavelengths between 0.7 pm and 100 pm, preferentially 25 pm and 100 pm. These cameras advantageously comprise near infrared cameras 52 “NIR” for “Near Infra Red” in English, and/or far infrared cameras 53 “FIR” for “Far Infra Red” in English. The images from NIR cameras, near infrared, can for example be used to delimit the position, dimensions and movements of different parts of the body of a vehicle user. The images from FIR, far infrared, cameras can for example be used to identify the parts of the user's body that exchange the most heat with the passenger compartment, for example the head and hands, which are not covered by clothing and appear warmer.

According to a particular embodiment, the interactive system 1 includes an ultra-wideband short-range radar. The frequency of the radar can for example be between 10 GHz and ITHz, preferably between 50 GHz and 160 GHz.

[0097] Alternatively or in addition, the sensors can be provided in portable elements (watch, shirt, bracelet, etc.), worn by the user. In particular, vital signs sensors 50 intended to be in contact with the user can also be used. They come, for example, in the form of conductive elements installed in parts of the vehicle (armrest, steering wheel, seat, etc.). These vital signs sensors 50 can be configured to measure parameters such as rhythm and/or heart rate, respiratory rate and/or amplitude, skin conductivity, brain waves, or others.

As a variant or in addition, other types of parameters can also be recovered by means of other sensors. The measuring device 5 advantageously makes it possible to determine the environment of the passenger(s), this being able to influence the state of the passenger(s). Thus, parameters describing the environment, such as temperature, light intensity, noise, vehicle speed, can also be recorded.

The measuring device 5 can also include biosensors that can detect parameters such as organic compounds, ions, bacteria, or others.

[0101] In addition, at least one sensor of the measuring device 5 can be configured to follow the evolution of the position and/or of the movements performed by the user, for example of his eyes, of a limb, of a part of the body, to monitor your breathing during a session, or other.

The processing unit 7 is advantageously on board the motor vehicle 100. The processing unit 7 is configured to be connected to the interface or interfaces 3. The processing unit 7 is configured to be connected to the measuring device 5.

[0103] Furthermore, the processing unit 7 can comprise one or more processing means. These processing means are in particular configured to implement at least in part the interaction method described below. In general, the processing means may comprise one or more of the means from among one or more telecommunications or telematics means, a comparator, at least one computer, a processor and/or any other hardware making it possible to execute software or no, an application-specific integrated circuit known as ASIC, read only memory known as ROM, random access memory known as RAM, or other memory, or other conventional or custom hardware.

[0104] In particular, the processing unit 7 is configured to determine, by analyzing the state of the user, how to adjust at least one parameter of one or more sequences of the exercise. For this, the processing unit 7 notably comprises one or more modules from among: a. an interpretation module for defining at least one datum representative of the state of the user, b. an analysis module to identify, based on the defined representative data(s), an adjustment to be applied, and vs. a correction module that can apply the identified adjustment.

The modules or systems of the processing unit 7 may respectively comprise one or more of the aforementioned processing means. They are differentiated to highlight certain functions of the processing unit 7. Of course processing means may or may not be common between different modules.

Said at least one piece of data representative of the state of the user can be data representative of the physiological state, data representative of the emotional state, data representative of the cognitive state, data representative of a combination of these states, an aggregate or a combination of the previous ones.

The interpretation module is configured to receive one or more parameters linked to the user U recorded by one or more sensors of the measuring device 5 and from the parameter or parameters received, define the data representative of the state of the user U, using a corresponding evaluation model. Such representative data may correspond to a physiological, emotional or cognitive state variable, or preferably an aggregate or a combination of several state variables.

[0108] The state variables linked to the user can be determined from the direct measurements recorded by the measuring device 5. For example, physiological state variables can relate to metabolism, fever, or even comfort. user heat. Variables of the user's emotional state are, for example, valence, arousal, dominance. Cognitive state variables are for example a level of attention, a level of distraction.

In addition, the interactive system 1 can accommodate applications available on the server 11. To this end, the interactive system 1, and in particular the processing unit 7, can include a download module for downloading, from the server 11, one or more of the elements among, an application selected and validated by the user U, at least one model for evaluating the state of individuals, at least one model for changing target values of parameters linked to the user, for example physiological, or target values of physiological state variables, or emotional such as valence or arousal, or cognitive, or even at least one model of evolution of data representative of target states. Templates can be downloaded with an app. Finally, the interactive system 1, and in particular the processing unit 7, comprises at least one module or operating system making it possible to execute an application, for example for managing at least one session of a interactive exercise to be carried out by the user, in particular a remediation exercise, intended to take place according to a predefined number of sequences S1 to S5 (see FIGS. 3 and 4). The operating system makes it possible to execute an application downloaded on demand from the server 11.

Furthermore, the processing unit 7 (shown schematically in Figures 1 and 2) may include a memory medium on which one or more data is intended to be recorded. In particular, the data relating to parameters linked to the user U, data representative of the state of the user U, state variables, performance data, can also be recorded on this memory medium.

Finally, the processing unit 7 comprises at least one processing means for detecting a command, such as a command to execute an application or to validate a session of a management application. a remedial exercise, by the user via the interface 3, for example by pressing on a touch screen, or by selecting via a wheel or a touchpad, a notified session or application.

Furthermore, the processing unit 7 may include a data communication module. This data communication module can exchange data with the server 11.

Advantageously, the processing unit 7 can comprise a data processing or selection module configured to automatically select, among the applications on the server 11, at least one application for managing at least one remedial exercise to be carried out by the user U to modify, preferably improve his condition. The selection of a remediation exercise can be done advantageously as a function of one or more defined data representative of the state of the user U and of the data representative of a state associated with each application.

The processing unit 7 may in particular comprise a control module for at least one element of the interactive system 1, such as an actuator 4 for activating a stimulus, or even for executing an application.

The processing unit 7 may include an information module configured to notify the user, via one or more interfaces 3, of one or more information, for example concerning the adjustment of a sequence, or even an applied stimulus, or a progression, or even at least information concerning an application selected automatically and/or downloaded.

In addition, the interactive system 1, in particular the processing unit 7, may include a telephony and/or messaging and/or video telecommunications module, so as to allow the user U to interact, for example with at least one specialist or a team of specialists in the field of the application executed, in particular a medical or health team 13. The telecommunications module can be embedded in the motor vehicle 100 or alternatively be integrated into the mobile terminal 9.

[0118] METHOD OF INTERACTION

The method of interaction with a user U is described below. It can be implemented at least in part by the interactive system 1 as described above.

The following description refers to a user. Of course, the interaction method can be applied, simultaneously or not, to several occupants of the motor vehicle.

[0120] With reference to FIGS. 1 to 5, the interaction method globally comprises one or more of the following phases: a measurement phase M, a reception and interpretation phase I, an execution phase of an application for managing an interactive exercise which can in particular be a remediation phase R when the application is an application for managing a remediation exercise, an analysis and learning phase A, an adjustment phase .

One or more phases of the interaction process can be implemented continuously. Moreover, these phases can be implemented in parallel.

[0122] In addition, one or more stimuli can be activated during one or other of these phases.

During the measurement phase M, one or more parameters linked to the user U, in particular physiological parameters of the user U, can be recorded, acquired. This measurement phase M can be performed by means of at least one sensor 50-54, of the measurement device 5 of the interactive system 1 previously described.

The measurements can be performed simultaneously or one after the other for different users in the motor vehicle 100. Parameters describing the environment in which G user U is located can also be noted, collected so as to deduce therefrom their possible effect on the state of user U.

The measurement phase M can be implemented prior to the execution of the application and also when the application is executed. The measurement phase M can be done during at least one of the sequences SI to S5 of an interactive exercise, in particular remediation. In particular, the measurement phase M is implemented before, during or after the application of a corrective action making it possible to modify a characteristic of one or more sequences and/or to activate a stimulus or a group of stimuli.

The data obtained during the measurement phase M can be transmitted to the memory medium which can be internal to the motor vehicle 100 or downloaded onto a data storage medium external to the motor vehicle 100.

During the reception and interpretation phase I, the parameter(s) noted during the measurement phase M can be received and analyzed or interpreted in order to define data representative of the state of G user U by means of of an evaluation model. This reception and interpretation phase I can be performed by the processing unit 7 of the interactive system 1, in particular the interpretation module of this processing unit 7.

The physiological parameters are modeled so as to obtain data representative of the state of G user. It can be represented in a characterization space. It corresponds for example to a point in the characterization space. Once the modeling of the representative datum has been carried out, a processing step can be carried out, so as to determine the state of the user U, for example according to the placement of the representative datum in the characterization space. This step makes it possible to map the state of the user U.

[0130] By way of example, the data representative of a physiological state of the user can correspond to a physiological state variable such as metabolism, fever, or else the thermal comfort of the user, or an aggregate such state variables.

In the case of an interpretation of the emotional state of the user, the characterization space of the emotional state can be a two-dimensional space formed of a first axis corresponding to the valence and of a second axis corresponding to the arousal. According to a variant schematized in FIG. 6, the space for characterizing the emotional state of the user can be a three-dimensional space. This three-dimensional space for characterizing the emotional state is formed by a first axis corresponding to the valence V, a second axis corresponding to the arousal A and a third axis corresponding to the dominance D. The valence V characterizes the level of pleasure (V+) or inconvenience (V-) associated with an emotion. In other words, the valence makes it possible to characterize how the user experiences a situation, positively (V+), being happy, or negatively (V-), being sad, angry, unhappy. Arousal A can be defined as the intensity of the response generated by a situation, an emotion. In other words, the arousal makes it possible to define whether the user experiences the situation with a lot of excitement (A+), or on the contrary with little excitement, that he is relaxed, asleep, passive (A-). Dominance D, on the other hand, characterizes the level of mastery and control of the user's response to this emotion. In other words, the dominance D makes it possible to characterize whether G user is in control, responsible for his reaction and his actions (D+) or on the contrary if he is overwhelmed, impulsive, under the control of his emotions (D-). In the example of mapping the emotional state in the three-dimensional space A, V, D, W corresponds for example to a state of anxiety, X corresponds for example to a state of ecstasy, Y corresponds for example to a state of serenity, Z corresponds for example to a state of depression. The point of intersection of the three axes of the three-dimensional space A, V, D corresponds to the neutral position of the emotional state of G user U.

FIG. 7 illustrates an example of placement of data representative of the emotional state obtained by modeling physiological parameters, in the three-dimensional characterization space A, V, D. This data varies over time. Thus, a first representative datum is provided at a time t0 and the representative datum provided at a time t corresponds to the present. Continuous measurements are preferably carried out, so as to obtain the curve represented in FIG. 7. However, discrete measurements can also be carried out.

As an alternative or in addition, to define data representative of the cognitive state of T user, a cognitive state evaluation model is used. In a non-exhaustive manner, the data representative of the cognitive state may correspond to one or more combined or aggregated cognitive state variables, such as a level of attention or distraction, a level of falling asleep, and possibly a state under the influence of narcotic substances. [0135] In addition, the data representative of the state of G user are defined for any time, for example before, during, after an exercise such as a remediation exercise, before, during or following a corrective action applied to at least one sequence, or even before, during, after activation of a stimulus.

The processing of the evaluation model of the state of G user can be improved by learning. In particular, the mapping can be modified by taking into account the impact of the user's culture and/or education, making it possible to obtain or enrich a user profile for each occupant of the motor vehicle 100. This can be done by deep learning. The mapping can also be adapted by taking into account the impact of the personal life and/or the experience of the user U. This can be done by statistical learning (or “machine leaming”) thus making it possible to adapt the cartography in real time. Deep or statistical learning can be done during the analysis and learning phase A.

The execution phase of the exercise, advantageously the remediation phase R, comprises the execution of the application selected by the user U or automatically by the interactive system 1 and validated by the user U The latter can then carry out at least one session of the remediation exercise.

Referring to Figures 3 and 4, the interactive exercise can take place according to one or more sequences from a sequence S 1 of choice, for example in an application menu, a relaxation or preparation sequence S2, a sequence S3 for carrying out the interactive exercise, in particular remediation, by the user, a sequence S4 for explaining the interactive exercise, in particular remediation, a sequence S5 of feedback from the user on the exercise, in particular remediation, carried out. The different sequences SI to S5 make it possible to accompany the user during the execution of the application. The relaxation or preparation sequence S2 is prior to the execution sequence S3 of the interactive exercise, in particular of remediation, by the user. The sequence S4 for explaining the remediation exercise can be executed after the sequence S3 for carrying out the exercise. This sequence S4 makes it possible to explain to the user the usefulness, the objective of the interactive exercise, in particular of remediation, its importance or even its therapeutic effect. During the explanation sequence S4, the user can receive at least one piece of information concerning his performance, and/or his progress. The user can also receive at least one item of information concerning its state, for example during at least one of the sequences of the interactive exercise, possibly before and/or following a corrective action as described below. During the last feedback sequence S5, the user can answer an interactive questionnaire, fill in information. This experience feedback information can be stored on the memory medium which can be internal to the vehicle and/or on a storage medium external to the vehicle. They can be analyzed by at least one processing means of the interactive system, in particular the analysis module, to improve the interactive exercise by learning.

During one or the other of these sequences S1-S5, so as to allow G user to tend towards a neutral state, an actuator can activate stimulus located in the characterization space, symmetrically to the representative datum the state of the user, relative to the neutral position. As a variant, in order to amplify a state, an actuator can activate a stimulus positioned in a direction opposite to the neutral position in the characterization space, with respect to the initial position of the data representative of the state of the user .

At least some of these sequences are essential, in particular the sequence S3 for carrying out the interactive exercise, in particular remediation by the user, and the sequence SI of choice in a menu of the application. Certain sequences may be optional, in particular, one or more of the sequences from relaxation or preparation sequences S2, explanation S4, feedback S5.

In addition, the sequences can be adapted specifically for the user U by learning. In particular, the duration, intensity, rhythm, speed, or difficulty of one or more sequences can be increased or, on the contrary, decreased. Examples of different sequence sequences, chosen arbitrarily, are represented in FIG. 4, each example corresponding to a respective user U1, U2, U3, U4. This adaptation of the sequences is advantageously done according to the state of the user concerned during one or more of the sequences S1 to S5.

Referring again to Figures 1 to 5, the analysis and learning phase A can be implemented at least in part by the processing unit 7 of the interactive system 1, in particular the module of analysis of the processing unit 7. It makes it possible to adapt, to personalize, the interactive exercise, in particular of remediation, according to a correlation law, from the state of the user U when he do this exercise. The correlation law can be defined from a matrix or correction table. As a variant, the correlation law can be defined by a transfer function, a distribution function. According to yet another variant, the correlation law can be defined by a decision graph, a decision tree. The correlation law can also be defined by any other means or element or function, making it possible to link the reaction of G user during a sequence, and in particular to a modification of a characteristic or a parameter of the sequence and/ or to a stimulus or group of stimuli applied during this sequence, to an adjustment. The correlation matrix or table, the graph or the decision tree, the transfer function can be recorded on the memory medium and/or be accessible from the server.

[0144] Furthermore, the correlation law can be contextualized. In other words, it can also comprise at least one contextual information in correlation with an adjustment of at least one parameter of the sequence of the interactive exercise. This contextual information can be temporal, timestamp information, and/or location information and/or an event such as a vacation or work period or even the identification of a regular journey, or a phase of user life u.

Such a correlation law can be pre-established, pre-defined.

Alternatively, the correlation law can be established and/or corrected iteratively by learning.

For this, the analysis and learning phase A includes one or more steps to determine how to potentially adjust at least one parameter or one characteristic of one or more sequences S1-S5 of the interactive exercise, in particular to remediation. In particular, a correlation can be identified between a data representative of the state of the user defined and an adjustment of at least one parameter of the sequence of the interactive exercise to be applied. The parameter can be adjusted during the interactive exercise, especially during remediation. Alternatively or additionally, the parameter can be adjusted for a later session of this exercise.

In general, the analysis and learning phase A comprises a step for comparing at least one piece of data representative of the state of the user defined from parameters obtained by the measuring device 5 with at least reference data. In alternatively or in addition, at least one parameter linked to the user such as a state variable can be compared with a reference datum.

The reference datum can be chosen from among datum representative of a target state, a target value of a parameter linked to the user such as a state variable, or a predefined threshold value or pivot value or defined by learning, or even data representative of a previous state of the user during a previous sequence or the current sequence, or even a previous value of a parameter linked to the user such as a variable acquired during a previous sequence or the current sequence.

[0150] Depending on the comparison results, the correlation law making it possible to link a datum representative of the state of G user to an adjustment, can be adapted.

For this, the method makes it possible to apply at least one corrective action and to analyze the writing of the user U following the corrective action to determine the impact of this corrective action on the user U and in based on the analysis results deduce how to adjust it. This analysis can be implemented by the analysis module, according to at least one physiological parameter of the user acquired during the execution of the corrected sequence and at least one reference datum.

More specifically, the method comprises the following steps: a. apply a corrective action during at least one sequence of the interactive exercise, when the application is executed, b. acquiring or recording at least one parameter linked to user G, such as one or more physiological parameters, during the execution of the corrected sequence according to the corrective action, c. receiving said at least one physiological parameter recorded during the execution of the corrected sequence according to the corrective action and defining accordingly a datum representative of the state of G user U during the execution of the corrected sequence according to the corrective action, d. compare the datum representative of the state of G user U during the execution of the corrected sequence according to the corrective action with at least one reference datum, and e. adapt the corrective action based on the comparison results.

The corrective action can be applied by the processing unit 7, in particular by the correction module previously described. And this corrective action can be applied during the remediation phase R, when the interactive exercise is a remediation exercise.

The physiological parameters of the user U can be read by at least one sensor 50-54 of the measuring device 5, during the execution of the corrected sequence according to the corrective action. For this, the measurement phase M is implemented during and following the corrective action.

The processing unit 7, in particular the interpretation module, can receive these physiological parameters and define data representative of the state of G user U during the execution of the sequence corrected according to the corrective action . For this, the reception and interpretation phase I is implemented during and following the corrective action.

The processing unit 7, in particular the analysis module, can receive and compare the data representative of the state of the user U during the execution of the sequence corrected according to the corrective action with the data reference. Examples of implementation according to the corrective action are described below.

The adaptation of the corrective action can be implemented by the processing unit 7 of the interactive system 1, in particular the correction module of the processing unit 7.

More specifically, during the adjustment phase, at least one sequence from among the relaxation or preparation sequence S2, the sequence S3 for carrying out the interactive exercise, in particular remediation, by the user, the sequence S4 explanation, and the S5 feedback sequence can be adjusted. When adjusting one or more sequences of the interactive exercise, at least one parameter or characteristic can be modified among the type, duration, rhythm, intensity, speed, or even execution or not of a sequence.

A modification of the type can be to change the type of tasks or actions carried out during at least one sequence, and this can also be the type of remediation when it comes to the sequence S3 of execution of remedial exercise.

The duration of at least one sequence can be lengthened or shortened. The degree of intensity of at least one sequence can be increased or decreased. Also, one or more sequences can be speeded up or slowed down. Concerning the rhythm, the sequence of the different sequences can be adapted, with a possible modification of their durations and/or their intensity. As a variant or in addition, the adjustment may relate to the activation or not of at least one stimulus.

According to another variant, during an adjustment, one or more characteristics or variables of a stimulus activated during at least one of the sequences can be modified, such as, in a non-exhaustive manner, the type, the duration, rhythm, intensity, modulation. For example, this may concern: a. for the light environment, a modification of the intensity, the type of direct or indirect light, a matt or shiny appearance, the type of fixed or variable modulation, or even the duration; b. for the surrounding odor, a modification of the intensity of the fragrance, a modification of the mixture of perfumes or pheromones, of the type of diffusion or of the duration; vs. for the sound environment, a change in the intensity or type of sound, noise or music, a change in modulation, type of broadcast or even duration; d. for the thermal environment, a change in temperature, air flow, degree of humidity, type of diffusion or duration; e. for a contact stimulus via massaging or vibrating equipment such as a seat, a change in the type of massage, intensity or duration.

[0163] First method

According to a first method, the corrective action is selected randomly.

[0164] By way of illustrative example, a random corrective action can be to add or remove a duration that is randomly generated from a sequence, to randomly modify the intensity, the speed, of a sequence, to activate or deactivate a stimulus, or to modify a characteristic of a stimulus applied during this sequence.

The corrective action can for example be linearized.

In order to avoid instability of the system, the corrective actions can be limited, in particular in number, in frequency, in variation in amplitude and/or in time. Moreover, at least some of the parameters to be corrected, such as the duration, the intensity, the rhythm, the speed, of the sequences and/or of the stimuli, can be limited. These terminals can be predefined or possibly adapted by learning, for example according to the preferences of G user collected during the experience feedback sequence S5.

Next, the reactions of G user U according to the corrective actions applied randomly are analyzed in order to determine those having a positive impact on the state of G user U.

The impact of the corrective actions on the state of G user U can be determined according to a target state. In this case, this target state is the reference data against which the corrective action will be adjusted.

This target state is defined with respect to a given interactive exercise, for example remediation. It can be defined so that G user U is the most receptive to the remediation or to the current or future sequence.

[0170] The target state can be predefined. The method may optionally include a step for defining a target state. This target state can advantageously be modified by learning, in particular thanks to interactive questions during the feedback sequence S 5 .

The data representative of target states can be constructed from data from at least one group of reference individuals or reference panel. These are, for example, data measured and recorded in the laboratory, or via interactive systems according to the invention when the individuals carry out the interactive exercise, for example on board motor vehicles and subsequently transmitted and stored on a remote server, thus making it possible to enrich the model of evolution of the target state. For example, this corresponds to an optimum state which defines an envelope by taking the mean value of the behaviors of the reference group of individuals and at least one or two standard deviations around this mean.

During a comparison step, the data representative of the state of the user U during the execution of a corrected sequence, that is to say following a corrective action, is compared with the data representative of the target state. This comparison step can be implemented by the analysis module described previously.

The difference between the data representative of the state of the user U and the data representative of the target state can then be quantified, in particular in time and in amplitude. These comparison and quantification steps can be implemented by the analysis module described above. In order to facilitate the understanding of such an analysis, a very simplified example is represented in FIG. 8. This FIG. 8 shows a reference interval I_R of data or target values “x” as a function of time “t” for a given sequence. In a non-limiting manner, this reference interval I_R can be centered around an average curve and its limits are located at least two standard deviations on either side of the average curve. The target data or values can be time-varying in the sequence, as shown in this example. A first curve Cl corresponds to the evolution of a datum representative of the state of one of the individuals of the reference panel as a function of time “t”. A second curve C2 schematizes the evolution of the data representative of the user's state G as a function of time “t” during the sequence. The sequence can be divided into several tasks. In this example, a random corrective action can be applied between a task T_N and a successive task T_N+1 of the sequence, as shown by the arrow F.

The user's curve C2 is compared with the curve Cl corresponding to the target state. The difference in evolution between the user's curve C2 compared to the curve Cl corresponding to the target state can be qualified in time and in amplitude.

Depending on the comparison results, the corrective action, for example to be applied during a subsequent execution of the sequence or during the current sequence, can be adjusted. Indeed, if the state of the user (curve C2) tends or reaches the target state (curve Cl), the impact of the corrective action is determined as positive. This corrective action is then validated. It can be reproduced or amplified in a future iteration. Conversely, if the state of the user (curve C2) moves away from the target state (curve Cl), the impact is determined as negative. This corrective action is not validated in this case. It can be stopped or reduced during a next iteration.

Thus, the random corrective action can be adjusted, corrected, in an iterative manner by learning. This makes it possible to define or adapt the correlation law accordingly by matching a state of the user to the adjustment defined by learning.

The learning can be done in addition according to a satisfaction questionnaire during the feedback sequence S5 (figure 3).

The above example describes the comparison of data representative of the state of the user with data representative of a target state. This example may apply also for the comparison of a parameter linked to the user such as a state variable, with a target value (or with an average of target values) in order to quantify at least one deviation. Mention may be made, as a non-exhaustive example, of a variable of physiological state, or of emotional state such as valence, or arousal, or of cognitive state, or even an aggregate or a combination of several of these variables. 'state.

[0180] Second method

A second method makes it possible to modify at least one parameter of at least one sequence SI to S5 of the interactive exercise in a controlled manner taking into account a parameter linked to the user or the state of the user.

In other words, the mapping of the user's state makes it possible to decide on a corrective action on one or more sequences.

According to this second method, the correlation law or distribution function can for example be defined by a decision graph, a decision tree. Such a decision tree can be used for discrete variables (color for example) or even continuous variables. Such a decision tree can be predefined, standard at least initially and be intended to be adapted by learning.

An example of a standard decision tree or graph, which can be predefined, is schematized in a simplified way in FIG. 9. This standard graph can subsequently evolve by learning.

Thus, depending on the data representative of the state of G user U determined following the interpretation phase I, the interaction method includes at least one step to identify the corresponding corrective action to be applied according to the 'decision tree. This identification step can be implemented by the analysis module previously described.

As for the first method, the corrective action takes place by acting on at least one characteristic of a sequence, but this time chosen according to the decision tree. The corrective delta, the difference to be applied, can be random.

By way of example, if the physiological state of G user is in a zone corresponding to a sleepy state Epi and/or if his emotional state reflects a state of rage or anger Eml, a modification Ml can be reduce the duration of one or more sequences and adapt the rhythm of these sequences. For a stressed physiological state Ep2 and/or a neutral emotional state Em7, a modification M2 can be to keep the duration stable and to adapt the rhythm.

If the physiological state of the user corresponds to nervousness Ep3 and/or if his emotional state corresponds to fear Em2, a modification M3 can be to reduce the rhythm of the sequences and the sequence of explanations can be longer.

According to another example, if the physiological state of the user is in a zone corresponding to a state of fatigue Ep4 and/or if his emotional state is in a zone corresponding to a state of disgust Em3, a modification M4 may be to reduce the duration of one or more sequences.

On the contrary, if his physiological state corresponds to a calm physiological state Ep5 and/or if his emotional state reflects joy Em4 or sadness Em5, a modification M5 can be to extend the duration of one or more sequences .

According to yet another example, if his physiological state corresponds to a calm physiological state Ep5 and/or if his emotional state corresponds to a state of surprise Em6, a modification M6 can be to extend the duration of one or more sequences and to adapt the rhythm of these sequences.

The example described above can also apply for at least one parameter linked to the user such as a state variable, for example an emotional state variable, or a physiological state variable or cognitive state or an aggregate or a combination. In particular, if the valence is too negative, the duration of one or more sequences can be reduced. On the contrary, if the valence is more positive the duration of one or more sequences can be increased. Similarly, if the arousal is very positive, the duration of one or more sequences can be increased or even the intensity or the difficulty of one or more sequences can be increased. Otherwise, if the arousal is very negative, the duration of one or more sequences and/or the intensity may be significantly reduced.

In these different examples, the duration removed or added, the adaptation of the rhythm, can be defined randomly. These corrective elements can be adapted by learning. Thus, the decision tree may contain a bias maintained by learning.

Also, depending on the state of the user, at least one of the sequences, when it is not essential, may not be executed, deployed. For example, if the system interactive 1, in particular the processing unit 7, determines that the user U is tired, one or more of the optional sequences, such as the sequence S2 for relaxation, S4 for explanation, or S5 for feedback, may not not be executed. The choice of sequences to be executed is therefore adjusted.

[0295] As an alternative or in addition, the corrective action can also make it possible to play on the multi-sensory environment of the user U according to his state. At least one stimulus to be applied can be defined by the decision tree. By way of example, if the physiological state of G user is in an area corresponding to a sleepy state Epi, at least one stimulus to wake him up can be applied, such as a type of music, or a set of lights, or even a variation of the thermal environment, etc. Another example can be to change the music if the interactive exercise is accompanied by music and the previous music has the effect, for example, of making the user tense.

[0196] The corrective action therefore varies according to the users and their condition. According to the examples of FIG. 4, chosen arbitrarily, the sequences are adjusted differently for each respective user U1, U2, U3, U4.

In addition, the interaction process makes it possible to identify whether or not the corrective actions are beneficial for a given user U. When a corrective action is applied, the reaction may differ from user to user and be received positively by one user but negatively by another user.

According to the second method, the reference datum against which the corrective action will be adjusted can be at least one threshold value (pivot value) in particular of a state variable or of a datum representative of a state.

For example, at least one parameter linked to the user U such as an emotional state variable such as for example valence, arousal, dominance or even a physiological state variable such as metabolism or even a cognitive state variable such as the level of attention or distraction noted following the corrective action, in particular a function or a derivative of the measured value, can be compared with the threshold value.

If the measured value reaches or is greater than the threshold value, the corrective action applied is validated. During a next iteration it can be reproduced, or even increased. Conversely, if the measured value does not reach this threshold value, the parameters or features of previous interactive exercise sequences can be restored. The corrective action can also be reduced.

[0201] One or more threshold values can be predefined. The threshold values can also be defined and adapted by learning, for example according to the satisfaction questionnaires, in particular during the S5 feedback sequence (figure 3).

[0202] The decision tree can be corrected iteratively by learning in order to take account of user preferences.

Referring again to Figures 1 to 5, these comparison and determination steps can be implemented by the processing unit 7, in particular the analysis module previously described.

[0204] Alternatively or in addition to one or the other of these methods, for determining the impact of the corrective action on the state of the user U, the data representative of the state of the user U following the corrective action can be compared with the data representative of the state of the user U before this corrective action. This data can reflect the state of the user U during the execution of a previous sequence or during the current sequence before the corrective action.

To this end, the measurement phase M is implemented before and after the corrective action. During the reception and interpretation phase I, data representative of the user's state U is defined for the periods before and after this corrective action.

[0206] Alternatively, it is at least one parameter linked to the user U, such as an emotional or physiological or even cognitive state variable, which can be compared before and after the corrective action.

According to one or the other of the methods previously described, the increments to increase or reduce the corrective action can be implemented by the correction module. Increments can be made using a dichotomy method.

[0208] Either of these methods thus makes it possible to improve the adjustment of the interactive exercise, in particular of remediation, by learning. [0209] In addition, the elements learned by learning according to one or the other of these methods, can be confirmed, validated, or corrected by G user during the experience feedback sequence S5, for example by the through an interactive questionnaire.

[0210] In addition to one or other of these methods, the analysis and learning phase A can also take into account contextual data, or even data related to the use of the motor vehicle, such as example of navigation data.

To this end, the method can include a step in which the user U can define an event (for example a departure on vacation). Alternatively or in addition, the method may comprise a step of analyzing data related to the use of the motor vehicle 100 so as to identify phases of life such as a daily journey, or even time zones (morning, noon, evening ). These steps can be implemented by the processing unit 7, in particular the analysis module previously described.

The contextual elements identified following this analysis can be confirmed and/or enriched subsequently by the user U, for example during the feedback sequence S5.

[0213] During the analysis and learning phase A, the corrective actions are also adapted according to the identified contexts. The learned correlation law can thus also be contextual. This makes it possible to obtain a correlation law allowing personalization that is both linked to the individual and contextual.

[0214] The interaction process may also include a phase for notifying the user U in order to inform him of his state at any time or of any adjustment to the interactive exercise.

[0215] Furthermore, according to another aspect, the interaction process can make it possible to choose and propose an application of an interactive exercise, in particular remediation, to be carried out by the user U according to his state. This selection can be made by the processing unit 7 of the interactive system 1.

[0216] To this end, the map of the state of the user U can be correlated with one or more maps corresponding to different applications stored on the server 11. Among the different applications of remedial exercises available in the library, those allowing the state of the user U to be brought back to a neutral position can be selected. For example, the application or applications for which a representative datum is located symmetrically to the datum representative of the state of G user with respect to the neutral position in the characterization space, can be selected. The map of the applications can be either on the server 11 or be downloaded via the interactive system 1, for example on a storage space provided for this purpose of the processing unit 7.

The user U can then validate the application which is notified to him or one of the applications if a list, possibly classified, is notified to him. This validation can be done by interaction with the interface 3, for example by pressing on a touch screen whether or not it is embedded in the motor vehicle 100. The interaction method can include a step of detecting a validation action of the user. There follows a phase of downloading from the server 11 of the application validated by the user U, for example by a download module of the processing unit 7. The validated application is downloaded onto a storage space provided for this purpose of the processing unit 7.

[0218] Thus, the interaction method does not aim for an interaction such as remediation which is only punctual, but seeks to personalize an interactive exercise at any time, in particular its sequences S1-S5, according to the state physiological, emotional, cognitive of the user U.

The personalization linked to the state of the user U is obtained by varying in a completely random manner or at least partly in a controlled manner, the characteristics or parameters of one or more of the sequences S1-S5. This is the corrective action according to one or other of the methods previously described.

[0220] Then, the reaction of the user U is analyzed. Depending on this analysis, a simple correlation can be defined or adapted: if the reaction of the user U is positive, the corrective action is lengthened or amplified, but shortened or stopped in the case of a negative reaction. These comparison results make it possible to define and correct the correlation law by learning.

Finally, this correlation law can also take into account predefined contextual information, entered by the user U or even determined by learning.

Claims

[Claim 1] Interactive system (1) with at least one user (U; Ul, U2,

U3, U4), including:

— a measuring device (5) configured to acquire at least one physiological parameter of the user (U; Ul, U2, U3, U4) during at least one sequence (SI, S2, S3, S4, S5) of a interactive exercise performed by the user (U; U1, U2, U3, U4) through the interactive system (1), and

— a processing unit (7) configured to be connected to the measuring device (5), comprising an interpretation module configured to receive said physiological parameter and to define, from said physiological parameter, data representative of the physiological state of the user (U; U1, U2, U3, U4) by means of a corresponding evaluation model, characterized in that the processing unit (7) comprises:

— an analysis module configured to identify, as a function of the data representative of the physiological state of the user (U; Ul, U2, U3, U4) defined, an adjustment of at least one parameter of at least a sequence (S2,

53, S4, S5) of the exercise to be applied, according to at least one correlation law making it possible to link said representative datum to an adjustment, and

— a correction module configured to modify the parameter of the sequence (S2, S3, S4, S5) of the interactive exercise according to the identified adjustment.

[Claim 2] Interactive system (1) according to the preceding claim, in which:

— the correction module is configured to apply a corrective action during at least one sequence (SI, S2, S3, S4, S5) of the interactive exercise, when the application is executed,

- said at least one sensor (50, 51, 52, 53, 54) is configured to acquire at least one physiological parameter of the user (U; Ul, U2, U3, U4) during the execution of the sequence (S2 , S3, S4, S5) corrected according to the corrective action,

— the interpretation module is configured to receive said at least one physiological parameter acquired during the execution of the sequence (S2, S3,

54, S5) corrected according to the corrective action and to define, a datum representative of the physiological state of the user (U; Ul, U2, U3, U4) during the execution of the corrected sequence according to the corrective action, and

- the analysis module is configured to compare the data representative of the physiological state of the user (U; U1, U2, U3, U4) during the execution of the sequence corrected according to the corrective action to said at least a reference datum, and

— the correction module is configured to adapt the corrective action according to the comparison results.

[Claim 3] Interactive system (1) according to the preceding claim, in which the said at least one reference datum is chosen from a datum representative of a target state, a datum representative of the previous state of the user during the execution of a previous or current sequence (S2, S3, S4, S5), a threshold value of a datum representative of a state or of at least one state variable depending on at least one physiological parameter.

[Claim 4] Interactive system (1) according to one of Claims 2 or 3, in which:

— the analysis module is configured to determine according to the comparison results if the corrective action improves the physiological state of G user (U; Ul, U2, U3, U4) during the execution of the sequence (S2, S3, S4, S5) corrected, and

— the correction module is configured to increase the corrective action if the physiological state is improved, otherwise to reduce or stop the corrective action, during the current sequence or a subsequent execution of the sequence.

[Claim 5] Interactive system (1) according to one of the preceding claims, in which the correction module is configured to adjust at least one parameter of at least one sequence (S2, S3, S4, S5) of the exercise interactive among type, duration, rhythm, intensity.

[Claim 6] Interactive system (1) according to one of the preceding claims, in which:

— the interactive exercise is a remedial exercise to be carried out by the user (U; Ul, U2, U3, U4) to modify his physiological state, comprising a predefined number of sequences among:

• a sequence (SI) of choices in an application menu, • a sequence (S2) of relaxation or preparation,

• a sequence (S3) for carrying out the remediation exercise by the user,

• a sequence (S4) explaining the remediation exercise, and

• a sequence (S5) of feedback from the user on the remediation exercise carried out and on the sequence adjustment applied, and in which

— the correction module is configured to adjust the parameter of at least one sequence from among the sequence (S2) of relaxation or preparation, the sequence (S3) of carrying out the remediation exercise per user, the sequence (S4) explanation of the remediation exercise, and the sequence (S 5) of feedback from the user on the remediation exercise carried out.

[Claim 7] Interactive system (1) according to one of the preceding claims, configured to interact with a user (U; U1, U2, U3, U4) occupying a motor vehicle (100), the interactive system (1) being at the least partly on board the motor vehicle (100).

[Claim 8] Method of interaction with at least one user (U; Ul, U2,

U3, U4), configured to be implemented at least in part by an interactive system (1) according to one of the preceding claims, said method comprising a phase of execution of an application of an interactive exercise to be performed by the user (U; Ul, U2, U3, U4) according to at least one sequence (SI, S2, S3, S4, S5), characterized in that said method comprises the following phases:

— a measurement phase (M) of at least one physiological parameter of G user by the measurement device (5) of the interactive system (1),

— a reception and interpretation phase (I) of said at least one physiological parameter by an interpretation module of a processing unit (7) of the interactive system (1), to define, from said at least one parameter physiological, data representative of the physiological state of the user (U; Ul, U2, U3, U4) by means of a corresponding evaluation model,

— an analysis and learning phase (A) by an analysis module of the processing unit (7) to identify, according to the data representative of the physiological state of the user (U; Ul, U2, U3, U4) defined, an adjustment of at least one parameter of at least one sequence (S2, S3, S4, S5) of the interactive exercise to be applied, according to at least one correlation law making it possible to link said representative datum to an adjustment, and

— an adjustment phase by a correction module of the processing unit (7), to modify the parameter of the sequence (S2, S3, S4, S5) of the interactive exercise according to the identified adjustment.

[Claim 9] Interaction method according to the preceding claim, comprising the following steps:

— apply a corrective action during at least one sequence (S2, S3, S4, S5) of the interactive exercise, when the application is executed, by the correction module,

— acquire at least one physiological parameter of the user (U; Ul, U2, U3, U4) during the execution of the sequence (S2, S3, S4, S5) corrected according to the corrective action, by the measuring device (5),

— receiving said at least one physiological parameter during the execution of the sequence (S2, S3, S4, S5) corrected according to the corrective action and defining a datum representative of the physiological and/or emotional and/or cognitive state of the user (U; Ul, U2, U3, U4) during the execution of the sequence (S2, S3, S4, S5) corrected according to the corrective action, by an interpretation module of the processing unit ( 7),

— compare the data representative of the physiological and/or emotional and/or cognitive state of the user (U; Ul, U2, U3, U4) during the execution of the corrected sequence (S2, S3, S4, S5) according to the corrective action to at least one reference datum by the analysis module of the processing unit (7), and

— adapt the corrective action to be applied during a subsequent execution of the sequence (S2, S3, S4, S5) of the interactive exercise, according to the results of comparison, by the correction module of the processing unit (7).

[Claim 10] Interaction method according to the preceding claim, in which the corrective action is random.