FR3091778A1 - IDENTIFICATION OF A STOPPED VEHICLE AND ITS LOCATION - Google Patents

Abstract

The invention relates to a method for identifying a vehicle (1) as being at a standstill with a view to determining a location of this stationary vehicle, comprising: a prior learning step in the signals picked up by the sensor unit, - and a current step of: * detection, in the signals (4) picked up by the sensor unit, of the signature of the event corresponding to a stopping of the vehicle and an exit of the user from the vehicle, * determination of the last geolocation coordinates of the vehicle thus detected as being stopped, and * sending to the remote server (5) of said geolocation coordinates. * detection of a need for maintenance of the vehicle vehicle (1), * sending, by the remote server (5), of a maintenance service proposal notification on the mobile terminal (8), * emission of a signal to start the maintenance service, the signal comprising at least the geolocation coordinates. [Fig. 1]

Description

IDENTIFICATION OF A STOPPED VEHICLE AND ITS LOCATION

The object of the invention is to identify a vehicle as being stationary with a view to determining the location of this stationary vehicle.

Electronic devices on board motorized vehicles (such as, for example, electronic dashboards) can include a satellite geolocation means, typically of the GPS chip type. Such embodiments make it possible to identify a current location of the vehicle when it is in operation and with the engine running (in particular by the user for guidance).

Technical problem

On the other hand, on the one hand, all vehicles are not necessarily equipped with such a GPS chip. On the other hand, when the engine is off and the user has left his vehicle, this GPS chip is no longer electrically powered and it becomes impossible to retrieve the last geolocation coordinate of the vehicle when stationary. However, this information is useful in particular for the implementation of an upkeep/maintenance service for a part of the vehicle which does not require the presence of the user of the vehicle. Typically, the operators of such a service would not have, with the current means of a GPS chip implanted in the vehicle, any information on the last geolocation of the vehicle, ensuring that the vehicle is indeed in the following situations:
- the vehicle is stationary, engine off,
- the user has actually left the vehicle,
these situations then being conducive to the above-mentioned upkeep/maintenance service.

The present invention improves the situation.

To this end, the invention proposes a method, implemented by computer means, for identifying a motorized vehicle as being stationary with a view to determining a location of this stationary vehicle. , in which a portable device is provided, available to a user of the vehicle, the portable device comprising a signal sensor unit, a geolocation module and a communication interface with a remote server.

In particular, the method comprises:- a prior step of learning in the signals picked up by the sensor unit, at least one signature of a predefined event, corresponding to a stopping of the vehicle and an exit, of the vehicle, of the user carrying the portable device,
- and a common step of:
* detection, in the signals picked up by the sensor unit, of said signature of the event of stopping the vehicle and leaving the user,
* determination of the last geolocation coordinates of the vehicle thus detected as being stationary, and
* sending the said geolocation coordinates to the remote server,.
The remote server being capable of communicating with a mobile terminal available to the user, the method may further comprise:
- detection of vehicle maintenance needs,
- sending, by the remote server, of a maintenance service proposal notification to the mobile terminal,
- the reception, by the remote server, of a maintenance service request transmitted by the mobile terminal.

Thus, the invention can have recourse to prior learning of a sequence for stopping the engine and exiting the user from the vehicle, for example by implementing artificial intelligence, by previously identifying each sequence of signals received from one or more sensors, by marking (or “tag”) of each sequence, as developed below. Thanks to such an arrangement, in the current step where it is sought to determine the last geolocation coordinates of the presumed stopped vehicle, if the succession of these learned sequences (stopping of the vehicle and exit from the 'user) is effectively identified, then this succession of events has taken place with certainty and therefore the last geolocation coordinates are indeed those of a stationary vehicle without a driver.

In the foregoing, an application of the above method is presented, in particular for the implementation of a maintenance service for stationary motor vehicles, such as, for example, the replacement of windscreen wipers after a given period, or others. Users of such vehicles may have previously subscribed to this service and installed a corresponding mobile application on their mobile terminal.

For example, the mobile terminal is configured to send the maintenance service request without any particular action by the user or in response to an action by the user on the mobile terminal (via an installed mobile application, typically).

In this embodiment, it may be advantageous for the method to further include, in response to receiving the maintenance service request, the transmission of a maintenance service initiation signal (intended for human operators who can then intervene on the location of the vehicle whose last geolocation coordinates have been retrieved by the server). Thus, in this embodiment again, it may also be advantageous for the signal to include at least these geolocation coordinates.

According to a particular embodiment, the preliminary learning step comprises:
- acquiring a plurality of signals picked up by the sensor unit,
- identifying, in each signal of the plurality of signals, if at least a portion of the signal corresponds to the predefined event,
- associate with the at least one portion of the signal a marker representative of the predefined event,
- training, by learning, an artificial intelligence in the detection, in signals picked up by the sensor unit, of said signature of the event, the training being carried out on the basis of each signal of the plurality of signals picked up and of the markers associated with said signal portion.

By way of non-limiting examples, artificial intelligence can be implemented by “deep” learning, by a trained neural network, or by “machine” learning. (from the English “machine learning”). In a practical example, a human operator identifies in the signals from the sensors, each of the relevant situations (engine off, movement of the user to get out of the vehicle) and thus marks each of these situations to train a neural network to recognize them. .

We can define a "neural network" generally by a set of transfer functions whose coefficients are gradually modified (as the learning progresses) according to a new signal acquisition captured and marked as an event to recognise.

In some embodiments, it may be advantageous to ensure the identity of the vehicle, prior to sending the geolocation coordinates to the remote server. Thanks to this provision, it is possible to identify that the user has left his personal vehicle, and not another vehicle, such as a taxi for example. There are many ways to ensure the identity of the vehicle. For example, the vehicle can itself transmit to the portable device an identifier of the vehicle by a short-range communication means, such as Bluetooth for example. Indeed, provision can be made for a vehicle memory connected to this short-range communication means to be able to store at least one vehicle identifier, and send the identifier of this vehicle to the remote server via the portable device.

However, once the engine is off, in most practical situations, no Bluetooth module is operational to transmit this identifier unless a battery is provided specifically for this means.

Another way of identifying the vehicle then comprises a step of identifying the vehicle by detecting a fingerprint of the vehicle, as described below.

In this embodiment, the sensor unit comprises an accelerometer capable of sensing vibrations of the engine of the vehicle in operation, and thus the method further comprises:
prior learning to identify a characteristic vibration of the engine in at least one predefined phase of engine operation, this characteristic vibration then being representative of an identity of the vehicle,
and a current detection, in the signals picked up by the sensor unit, of this characteristic vibration,
the sending to the remote server of the geolocation coordinates then being conditioned by a positive detection of the aforementioned characteristic vibration.

Thanks to such an arrangement, it is possible to remotely identify a vehicle even if it is not equipped with an electronic transmission device of the Bluetooth type (for example a vehicle of an old model) or even if this transmission medium is not electrically powered.

To ensure the event of stopping the vehicle and leaving the user, it is possible to identify on the one hand that the vehicle is stationary, and on the other hand that its user has left it. . In order to ensure that the vehicle is stationary, according to one embodiment, the signature of said event includes a characteristic vibration of engine operation, over a time interval, in at least one of the following situations:
- neutralization of the engine,
- engine shutdown.

In order to ensure that its user has gotten out of it, according to one embodiment, the sensor unit comprises at least one element from among an accelerometer, an altimeter, a gyroscope, a pressure sensor, a light sensor, a microphone. The signature of said event includes a specific measurement of a user's movement, over a time interval, in at least one of the following situations:
- the user removes his seat belt,
- the user opens the door,
- the user makes a movement to leave the seat,
- the user performs a movement to pass through the door,
- the user closes the door,
- the user actuates a closing mechanism of the vehicle.

For example, the specific measurement is a characteristic acceleration. For example the door is the driver's side door. For example the seat is the driver's seat.

According to one embodiment, the signature of said event comprises a succession of at least two characteristic accelerations corresponding to respective situations among the situations listed above, taken successively in the order presented above.

In the case presented here where the sensor unit comprises an accelerometer capable of detecting characteristic accelerations of the user, such an accelerometer can also detect vibrations, in particular vibrations characteristic of the engine of the vehicle, which makes it possible to achieve savings means in the portable device.

There are several ways to make a portable device. In one embodiment, the portable device is a mobile terminal (smartphone type or other connected object equipped with a GPS module). Such equipment may be sufficient to detect characteristic user movements and accelerations, as well as characteristic engine vibrations. However, it is possible that the user leaves his smartphone placed in the glove compartment of the passenger compartment of the vehicle and that the smartphone does not detect all the movements of the user. In another embodiment which does not have such a drawback, the portable device may consist of a connected portable accessory (connected bracelet or connected spectacle frame), carrying the sensor unit, and a mobile terminal capable of communicating with wearable accessory.

This mobile terminal can then carry out the processing for identifying characteristic signals in the signals picked up by the sensor unit. Nevertheless, the calculations to be implemented for these processing operations are tedious and it may be preferable to carry out these calculations with the server.

In this embodiment, provision can then be made for the portable device to comprise the sensor unit and a first radiofrequency communication module, capable of communicating with a second radiofrequency module of a mobile terminal, available to the user, and connected to the remote server via a wide area network. These can typically be Bluetooth or Wi-Fi communication modules.

In such an embodiment, the method may include:
- A transmission by the first module to the second module of the signals received,
- A transmission from the mobile terminal to the server of the signals received, and
- Processing with the server to identify a characteristic signal among the signals received.

In this embodiment, the mobile terminal can advantageously include the aforementioned geolocation module. In the event of identification of the signature of the event in the signals received, the server requests from the mobile terminal the last geolocation coordinates of the vehicle which the mobile terminal can then transmit.

Furthermore, the aforementioned “characteristic signal” may be the aforementioned signature of the event of stopping the vehicle and leaving the user. It can also be a question of identifying the vehicle according to its characteristic vibration. It may also involve recognizing the user himself, as provided for in the embodiment described below.

According to this particular embodiment, the portable device comprises a portable accessory, connected, available to the user and carrying the sensor unit, this sensor unit comprising at least one accelerometer. In particular, the accelerometer measures at least movements of the user to detect characteristic gestures of the user, these characteristic gestures then being representative of an identity of the user.

In one embodiment, the connected accessory is actually a spectacle frame available to the user equipped with an accelerometer and it has actually been observed that a physical movement out of the vehicle (or other preliminary gestures ) characterize an individual, like a biometric fingerprint.

As indicated above, users may have previously subscribed to the maintenance service and installed the mobile application on their terminal. As moreover, each user can be identified by a physical movement (exiting the vehicle or other gestures) such as a biometric fingerprint, the sending of the maintenance service proposal notification to the user's mobile terminal may be conditioned by the identification of the user by the aforementioned characteristic gestures of the user.

Thus, recognizing a user by his characteristic gestures makes it possible to send the notification to the terminal of this user precisely. In addition, a user who has just exited the vehicle will not be disturbed while driving by the message received on his terminal (since he has exited the vehicle). The message is also received at a relevant time, since the user has just left his vehicle and is aware of the state of the vehicle.

In an embodiment where the aforementioned maintenance concerns a part of the vehicle likely to wear out (typically such as windshield wipers), the detection of a need for maintenance of the vehicle may include:
- prior learning in the signals picked up by the sensor unit, of a wear state signature of a predefined part of the vehicle,
- And a current detection, in the signals picked up by the sensor unit, of said wear state signature of the predefined part.

For example, the predefined part is a pair of windscreen wipers from the vehicle and the signature of the state of wear is a characteristic vibration of the friction of the worn windscreen wipers on the windshield.

In an alternative embodiment, the detection of a need for maintenance of the vehicle can comprise:
- prior learning in the signals picked up by the sensor unit, of a characteristic signature of a stress on a predefined part of the vehicle,
- And a current detection, in the signals picked up by the sensor unit, of a drift of said characteristic signature of the stress on the predefined part.

A stress of the predefined part is an action exerted by a force or a load on the predefined part or an action of the predefined part caused by a force or a load. For example, a predefined part is a wiper blade and a request for the wiper blade is an actuation of a wiper blade cycle by a control lever. For example, a sensor (microphone or other) of signals produced by the friction of the windscreen wiper (with a characteristic sound in the event of wear) can be provided. This sensor can be integrated in the aforementioned sensor unit for example.

It is possible to provide a form of service implementation, strictly speaking, in which a memory is also provided storing a pre-established list of operators each associated with an operator's mobile terminal telecommunications address, and each dynamically associated with a availability indicator and to presence coordinates, the availability indicator taking at least one of the values from the list consisting of: "available", "not available", and the vehicle maintenance service may include:
- select, from the pre-established list of operators, an operator whose presence coordinates are within a determined radius of action around said geolocation coordinates, and whose availability indicator has an "available" value,
- issue, by the remote server, a maintenance service request on the operator's mobile terminal associated with the selected operator,
- in response to the reception, by the remote server, of a maintenance service performance acceptance message transmitted by the operator's mobile terminal, storing in the memory, in association with the selected operator, a new value operator availability indicator "not available".

The present invention further relates to a system for implementing the above method, comprising:
- a portable device, available to a user of the vehicle, and including a signal sensor unit, a geolocation module, and a communication interface with a remote server, and
- said remote server.

The invention further relates to a server of a system as described previously, this server being configured to implement:
- a prior step of learning in signals picked up by the sensor unit, at least one signature of a predefined event, corresponding to a stopping of the vehicle and an exit, from the vehicle, of the user wearing the portable device,
- a common step of:
* detection, in the signals picked up by the sensor unit, of said signature of the event of stopping the vehicle and leaving the user,
* and request for the last geolocation coordinates of the vehicle thus detected as being stationary.

The invention also relates to a computer program comprising instructions for the implementation of the method above, when these instructions are executed by a processor.

This may typically be the processor (reference 22 in Figure 1) of a computer processing circuit further comprising a memory (6) for storing computer program instructions, as well as a communication interface (7) with an extended communication network (21), as described later with reference to Figure 1.

Moreover, other characteristics and advantages of the invention will appear on reading the description which follows. This is purely illustrative and must be read in conjunction with the appended drawings on which:

Fig. 1

illustrates a diagram of a system for implementing a method of detecting a vehicle location according to the present invention;

Fig. 2

illustrates a step diagram of the method of detecting a vehicle location; and

Fig. 3

illustrates two acquisition graphs of acceleration versus time during engine start of two different vehicle models.

Fig. 4

illustrates a graph of acquisition of acceleration versus time for a person performing a series of steps involving walking and the use of a vehicle;

Fig. 5

illustrates a vibration graph representative of the state of an engine at a given time.

The identification of a stationary vehicle and its location is described below.

FIG. 1 illustrates an example of a system 100 for the implementation of a method for detecting the location of a stationary vehicle, in particular with the aim of carrying out maintenance on this vehicle.

The system 100 comprises in particular the vehicle 1, a connected spectacle frame 2 worn by the user 9, a smart telephone, also called in English “smartphone” 8, an extended communication network 21 comprising in particular the internet and a remote server 5.

As visible in Figure 1, a user 9 of the connected glasses frame 2 leaves the vehicle 1. In the example of Figure 1, the vehicle 1 is a passenger car. Also, user 9 has smartphone 8.

User 9 has subscribed to a maintenance service for his vehicle 1.

For this, the user 9 has installed a maintenance service application on his smartphone 8 and has an associated user account managed by the remote server 5.

As shown in Figure 1, the remote server 5 comprises a computer processing circuit comprising a processor 22 and a memory 6, as well as a communication interface 7 with an extended communication network 21.

The memory 6 stores, in the user account, an identifier of the user, an identifier of the vehicle of the user, and a telecommunications address of the smartphone 8 of the user, for example a telephone number. It can also store temporary data (calculation or other) and in particular the instruction data of the computer program of the invention. It can thus be a memory unit 6, overall.

The remote server 5 is programmed to send to the application, via the communication interface 7, a maintenance service notification for the vehicle 1 so that the vehicle 1 benefits from maintenance at its last location, when these two conditions are met. fulfilled:
Condition A) Vehicle 1 requires maintenance;
Condition B) User 9 is out of vehicle 1 and does not need it immediately.

In order to ensure condition B) and to determine the last location of the vehicle in this condition, a process for identifying a vehicle 1 as being stationary with its driver out of the vehicle, is implemented.

A first embodiment of the identification method is described below.

As described below in more detail, the identification process may include the following steps:
Identification (S0) of the user 9;
Identification (S1) of vehicle 1;
Identification (S2) of the last GPS coordinates of the vehicle 1.

An identification (S0) of the user 9 is described below.

The connected frame of the pair of glasses 2 (hereinafter called “connected glasses frame 2”) comprises a sensor unit 3 comprising an accelerometer capable of measuring accelerations linked to movements of the user 9. The connected glasses frame 2 is further configured to filter the vibratory signals received in order to isolate the useful part thereof. In particular, the connected spectacle frame 2 comprises for this purpose a computer processing circuit comprising a processor and a memory.

Each person has a gait composed of movements that are specific to him. Also, the gait of a person can be a gesture characterizing the user in the same way as his fingerprints.

In particular, the user 9 performs when driving a series of head movements, for example to look in the mirrors. Thus, the frequency and the amplitude of the movement of the head that the user 9 makes to look in his central rear-view mirror can form a gestural signature characteristic of the identity of the user 9. The way in which the user 9 replaces his glasses on his nose can also constitute a characteristic gesture signature of his identity. In addition, the natural gait of the user 9 when he walks, opens the door of his car and enters it can constitute as many gestural signatures of the user 9. In the same way, when the user leaves his car, slamming the driver's side door, etc. it is a sequence of gestures that can characterize both the user and his actual exit from the vehicle.

The connected spectacle frame 2 transmits to the remote server 5 the acceleration measured during the characteristic movement of the user 9, via the smartphone 8. For this, the connected spectacle frame 2 and the smartphone are equipped with interfaces short-range communication devices, here Bluetooth interfaces. In addition, the smartphone 8 comprises for this purpose a computer processing circuit comprising a processor and a memory. The smartphone 8 transmits the acceleration to the remote server 5 via the extended communication network 21.

The remote server 5 comprises a computer processing circuit able to determine that the acceleration corresponds to a characteristic movement of the identity of the user 9. To allow this determination, the computer processing circuit comprises an artificial intelligence which has been previously trained.

This training may have been carried out by an operator piloting equipment implementing artificial intelligence and who can thus mark (or “tagger”) actions (or sequences of actions) by entering them via his equipment. During training, the operator also uses a vehicle preferably of the same model.

The operator performs each action of the type several times, for example:
- Stop the vehicle engine,
- Remove your seat belt,
- Open the door on the driver's side,
- Make a movement to leave the driver's seat,
- Make a movement to pass through the door on the driver's side,
- Close the door on the driver's side,
- Operate a closing mechanism of the vehicle.

These actions (or at least part of these actions), taken successively, can constitute a sequence, as will be seen later, in particular of stopping and exiting the user from his vehicle. To inform via its equipment of what action it is on the acceleration measurements, the marking can consist of a tag (for example "removal of the seat belt") associated with a time interval (for example "2s - 8s”) of the acceleration measurement, for each of the elements of the sequence above. The acquired acceleration measurements can for example be sent from the equipment to the remote server 5 in association with the action markers. Artificial intelligence is thus trained to recognize the aforementioned actions.

Then, during the first use of the service by a user, for example, the user can optionally enter on the application of his smartphone 8 a user identifier and/or an identifier of his vehicle (the license plate number vehicle registration, for example).

In one embodiment, the connected spectacle frame 2 worn by the user is configured to capture and transmit to the remote server 5 acceleration measurements (in particular movements of the user's head, vibrations, etc.), in association with the identifier of the user 9. During his first use of the service and subsequent uses, the user 9 equipped with the connected spectacle frame 2 thus reproduces naturally, without even thinking about it, one or more of the elements of the aforementioned sequence of actions. The actions can thus be identified by the remote server 5, thanks to the previous learning.

In addition, as described above, the acceleration measurements of each of these actions can also serve, when they are acquired for the user 9 in particular, as a signature of the identity of the user 9.

Thus artificial intelligence is trained in two main stages: a first stage, by training a network of neurons using action markers, corresponding to training to recognize signatures of actions carried out by a operator, then a second subsequent step, by autonomous learning, corresponding to training in recognizing in particular that the actions are carried out by a given user 9. Advantageously, the user 9 does not have to perform any particular action to train the autonomous intelligence to recognize it. Also, user 9 does not perceive any intrusion in his daily life.

Thus, the computer processing circuit of the remote server 5 is able to determine the identity of the user 9.

Advantageously, this determination makes it possible in particular to verify that the connected spectacle frame 2 is used by the user 9 registered in the user account, and not by another user.

Once the identity of the user 9 has been verified, the vehicle 1 can be identified by the identification step (S1) of the vehicle 1.

The steps of identification (S1) of vehicle 1 and of identification (S2) of the last GPS coordinates of vehicle 1 are described below with reference to FIGS. 1 and 2.

The engine of vehicle 1 emits (S100) a signature signal 4. The signature signal 4 is a vibration signal characteristic of the engine and is detected by the sensor unit 3 of the connected spectacle frame 2.

For ease of illustration, the capture of the signature signal 4 is shown in Figure 1 outside the vehicle 1. In fact, in this embodiment, the signature signal 4 is picked up by the sensor unit 3 when user 9 is inside vehicle 1.

The sensor unit 3 is an accelerometer capable of detecting vibrations. The characteristic vibration signal of the motor is a vibration of the motor in at least one predefined phase of motor operation. In general, this predefined phase can be selected from the list including acceleration, predefined gear change, deceleration, neutral, engine ignition.

For example, it may be poles of vibration of the engine at standstill or at constant speed, from which poles the different frequencies and the associated amplitudes characterize the engine, as illustrated in Figure 5. Alternatively or in addition, it can s It is also a particular phase of engine operation, the vibrations of which are characteristic over time, such as the acceleration phases visible in FIGS. 3 and 4, commented on below.

The signature signal 4 is the engine vibration signal (in time and/or in frequency). Typically, the motor vibrates and this vibration is partly transmitted to the passenger compartment in which the user and his connected accessory 2 carrying an accelerometer are located. The vibration that the accelerometer thus undergoes can define a temporal vibration signal of the type illustrated in FIG. 3. Indeed, FIG. vehicle model and on the right, 18 a second model).

In addition, it is possible to estimate a Fourier transform of this temporal signal to deduce the vibration poles of figure 5.

The snapshot of the frequency of the modes and their associated amplitude is characteristic of the engine of vehicle 1, and can be used to identify the engine in the same way as an identity card of vehicle 1.

As represented in FIG. 1 by the wavelets 19 and 20, the connected spectacle frame 2 transmits (S101) to the smartphone 8, via the short-range communication interfaces, the vibration signal. The smartphone 8 transmits (S102) the vibration signal to the remote server 5, via the extended communication network 21.

The communication interface 7 of the remote server 5 is configured to receive the vibration signal.

The computer processing circuit of the remote server 5 has been previously configured as an artificial intelligence trained to identify signature signals.

The training can be carried out by making markings, also called in English “tags”, on the graphs of a series of numerous graphs of training signals previously acquired and comprising the signature to be recognized later. Typically this "number of graphs" can be several hundred, or even several thousand for robust training.

In the present case, markers were associated with the acquisitions of training signature signals to identify a precise phase of operation of the vehicle engine (for example its extinction to identify it as being at a standstill).

It is possible, more generally, to keep a database of signatures of different engines up to date. A neural network can be used and the transfer function of each neuron is adjusted by feedback when training the neural network. For this, the acquisitions of training signature signals are sent to the neural network for identification. For each acquisition, the neural network is instructed a posteriori on the type of motor identified which must be associated with each acquisition until the neural network is able to recognize itself, for a new signature signal acquisition drive, the type of motor from which a new signal acquisition originated.

Thus, by using its neural network thus trained, the remote server 5 identifies (S103) the signal comprising a signature 4 as representative of the identity of the engine and more particularly of the vehicle 1.

More generally, artificial intelligence can also be trained in two main stages, as indicated previously: a first stage, by training a neural network using markers of operating phases, corresponding to training in recognizing signatures of operating phases of an engine in general, and a second subsequent step, by autonomous learning, corresponding to training in recognizing that the actions are carried out by a particular type of engine.

The second subsequent step can be extended by continuous autonomous learning on the engine of vehicle 1, or more generally on the vehicle itself. Indeed, an example of a mass-produced vehicle can cause vibrations in service which can change during the life of this vehicle, depending on its particular use, which use can be different from the use made of other vehicles of the same model.

Thus, the remote server 5 has identified the vehicle 1 and the user 9 by one of its characteristic gestures.

In addition, the remote server 5 is configured to verify that the identity of the user 9 and the identification of the vehicle 1 found by the steps described above correspond to the same user account registered in the remote server 5.

Advantageously, this verification makes it possible in particular to identify that the user 9 is in his personal vehicle, and not in a bus or a taxi. The remote server 5 can then send (S104) to the smartphone 8 a maintenance proposal notification in order to ensure the agreement of the user 9 for the maintenance to take place. Upon receipt of the notification, the user 9 can indicate (S105) his agreement for the maintenance operation to take place. Once the remote server 5 has determined that the user 9 is in his personal vehicle 1, condition B) can be tested as described below.

For practical reasons, the maintenance operation must take place when the user 9 does not need to use his vehicle 1. For this, the method includes a step of identifying (S2) the last GPS coordinates of the vehicle 1 and to this end, it is necessary to determine that the vehicle 1 is parked and to ensure that the user 9 has left his vehicle, the engine of the vehicle 1 being switched off.

A method for determining that the user 9 leaves his vehicle 1 is described below. User 9 has made a journey using his vehicle 1. He parks vehicle 1. User 9 turns off the engine of vehicle 1 and, as shown in FIG. 1, leaves vehicle 1. When he leaves the vehicle 1, the user 9 transmits (S106) a characteristic signal 10, comprising a signature of an event (stopping the engine and leaving the vehicle). As shown in Figure 1, the sensor unit 3 of the connected spectacle frame 2 captures the signature of this signal 10 both the characteristic vibration of the engine when switching off and the characteristic gesture of the user to exit the vehicle . The sensor unit 3 comprises an accelerometer configured to acquire these various measurements of vibrations and movements. The accelerometer can be implanted in the connected glasses frame 2.

This signal 10 is sent (S107) by the connected spectacle frame 2 to the smartphone 8 which transmits it (S108) to the remote server 5.

The remote server 5 comprises a computer processing circuit able to determine, in the second signature signal 10, a signature of the event of engine extinction and exit, by the user 9, of the vehicle 1.

The remote server 5 is equipped with an artificial intelligence trained to recognize event signatures in the signals that it receives from the mount 2 via the smartphone 8. For example, FIG. 4 represents an acceleration measurement acquisition in function of time, acquired by the accelerometer of the connected spectacle frame 2 for the following sequence of events:
- User 9 walks (S11) to reach vehicle 1.
- User 9 enters (S12) vehicle 1.
- User 9 starts (S13) the engine.
- The engine performs (S14) a sequence at different engine revolutions.
- User 9 exits (S15) from vehicle 1.
- User 9 walks (S16) to move away from vehicle 1.

To identify this succession of events, an operator may have previously marked each event on the acceleration measurements. For example, the marking may consist of a tag (for example "S11") associated with a time interval (for example "40s -50s") of the acceleration measurement.

Thus, the artificial intelligence of the remote server 5 is able to recognize (S109), in the event signature signal, the events S15 and S16, in this order, and to further identify a typical time difference between these two events. , which thus characterizes a typical succession of these events. When the remote server 5 has identified the two events S15 and S16, it sends a request (S110) for current geolocation coordinates to the smartphone 8. The application of the smartphone 8 is configured so that the smartphone 8 interrogates the geolocation module of the smartphone and transmits the smartphone's current geolocation coordinates, then it sends them (S111) to the remote server 5, in response to the geolocation data request.

The interrogation of the smartphone 8 by the server 5 takes place very shortly after the user leaves his vehicle (for example less than five seconds). Thus, the remote server 5 obtains the information of the last location of the vehicle 1, definitely parked for a long enough period allowing the maintenance operation. The remote server 5 can then send the smartphone 8 a maintenance service proposal notification.

Thus, the remote server 5 has the place where the maintenance of the vehicle can be undertaken, that is to say the location of the vehicle when stopped.

According to a variant, in order to be completely sure of condition B), the remote server 5 can send (S104) to the smartphone 8 the maintenance proposal notification after having identified that the user 9 has left the vehicle 1, in order to to ensure the agreement of the user 9 for the maintenance to take place. Upon receipt of the notification, the user 9 can indicate his agreement for the maintenance operation to take place. Advantageously, this variant makes it possible to prevent the remote server 5 from planning maintenance in the event that the user 9 had only temporarily left his vehicle 1.

The aforementioned maintenance consists of changing the windscreen wipers 23 of the vehicle 1 when the windscreen wipers 23 are worn, in an embodiment presented below.

The wiper replacement service 23 consists of sending a courier with a new pair of wipers to the location of the last geolocation coordinates of the stationary vehicle. The courier then replaces the windshield wipers 23 of vehicle 1 while user 9 is not using his vehicle 1. In this example, condition A), that is to say the fact that vehicle 1 requires a maintenance, can be provided as explained below.

The windscreen wipers 23 are considered worn after a defined period of use, for example one year.

Alternatively, the wipers 23 are considered worn according to a defined time of use weighted by criteria of frequency and intensity of use of the vehicle and/or frequency and intensity of local rainfall, and/ or exposure to ultraviolet rays.

Alternatively, the wipers 23 are considered worn by detecting and analyzing a signature signal which can be detected, transmitted and analyzed as follows. This signal, picked up by the sensor unit 3 of the connected spectacle frame 2, may actually comprise a characteristic signature of the state of wear of the windscreen wipers 23. For example, this signature may be a characteristic vibration of a friction of the brushes 23 on the windshield with a particular noise.

Then, before sending a courier to the place of the last GPS coordinates, the remote server 5 can select for example, from a list of couriers stored in its memory 6, a courier available within a predetermined radius of action around the last coordinates GPS of the vehicle 1. In particular, in the memory 6, each courier of the list of couriers is associated with a smartphone, is also associated with an option, also called in English: "flag", of availability taking a value indicating its availability ( "available" or "not available"), and each courier is also associated with presence coordinates. For each courier, the value of the option and the presence coordinates are updated in the remote server 5, for example by the sending of periodic update requests by the smartphone of the courier. The selected courier receives a maintenance request on his smartphone, sent by the remote server 5. The request includes an identifier of the vehicle 1, for example its license plate, as well as its last GPS coordinates. The courier responds to this request to validate with the remote server 5 that he is actually going to replace the windshield wipers of vehicle 1.

Thanks to such an implementation, it is possible to offer a vehicle maintenance service 1 without user intervention. Indeed, the user 9 does not need to remember to change his wipers 23 regularly. In addition, the maintenance service is able to be carried out without inconvenience for the user 9. Indeed, the service detects that the user 9 is absent from his vehicle 1 and offers maintenance when the user 9 has not need for his vehicle 1. Moreover, according to certain aspects involving the verification of the identity of the vehicle 1, other advantages can be obtained, in particular security, but also to identify that the user 9 is in his personal vehicle in order to avoid carrying out a maintenance proposal on a vehicle other than the personal vehicle of the user 9.

Of course, the present invention is not limited to the embodiments described above by way of examples; it extends to other variants.

Thus, for example, in order to ensure condition B) and to determine the last location of vehicle 1, a second embodiment of the identification method can be implemented as presented below.

According to this embodiment, it is not necessary to carry out an identification of the user 9 but to detect a stoppage of the vehicle and an exit of the user from the vehicle only from the user's smartphone, which constitutes thus, only the portable device comprising the sensor unit 3. The remote server 5 is always configured to determine that the user 9 turns off the engine of the vehicle 1 and leaves the vehicle 1. This determination is made on the basis of the acquisition, by the smartphone 8, of a signal possibly comprising a signature of this succession of events. This signal is transmitted to the server and as soon as the signature of events is identified by the server, the smartphone sends its geolocation coordinates, with a terminal identifier (telephone number in the GSM network, or number of its SIM card, or others). This smartphone identifier makes it possible to check in a database that the characteristic vibration of the engine corresponds to the owner of this vehicle according to his smartphone identifier and that a notification can be sent to this smartphone for the start of the maintenance operation .

In this embodiment, to identify the characteristic vibration of the engine, the accelerometer of the smartphone can be used, or alternatively, a microphone that smartphones usually include can pick up an acoustic signal in which the server can look for a characteristic signature. This microphone can also be used to detect that the user has left his vehicle by detecting a door closing sound signature followed by footsteps for example. Thus, the aforementioned accelerometer may not be used.

Typically, the sound of a door closing of a particular vehicle can be recognized by learning, but also it is possible to determine the door that has been closed (front driver or passenger, or rear, etc.).

In another variant, the smartphone can also be equipped with a GPS chip ("Global Positioning System") and it is detected that the user has left his vehicle by detecting a change in the speed of movement of the smartphone, first on board in a moving vehicle then carried by the user, who is walking. The application makes it possible to store, by the memory, a sequence of geolocation coordinates associated with timestamps. This sequence is acquired by preprocessing data from the GPS chip embedded in the smartphone. The sequence of coordinates can thus be used to detect that the user has left his vehicle, and this by detecting. As a variant, provision can be made for the connected spectacle frame 2 to be equipped with a GPS chip for the aforementioned applications.

Although a remote server comprising an artificial intelligence has been described to allow this determination, it is possible instead to provide a computer processing circuit comprising such an artificial intelligence implanted rather in the portable device (smartphone or connected spectacle frame).

Furthermore, although the portable device represented in FIG. 1 is in the form of a connected eyeglass frame 2, other types of portable device can make it possible to carry out the invention. For example, the portable device can be a connected bracelet, a smart phone, also called a “smartphone”, a connected belt, a connected ring or necklace, a connected pin, a connected handbag or briefcase, etc.

There are a wide variety of motorized vehicles 1 whose detection process can identify the location. For example, the vehicle is an automobile such as a passenger car. In another example, the vehicle is a truck, motorcycle or moped.

In one embodiment, the vehicle is an electric car. In this case, the signature signal 4 is for example a vibration signal characteristic of the braking system or of the suspensions, in at least one predefined phase of operation of the braking system or of the suspensions.

Although the example of maintenance service mentioned above is a wiper replacement service, there are a wide variety of maintenance services, adaptable to any other vehicle needs, that can be carried out during a identification of a vehicle as being stationary with a view to determining a location of this stationary vehicle as described above. For example, the maintenance may consist of replacing other worn parts of vehicle 1, such as headlights from the outside, tires, hubcaps, or the like. The example of maintenance described above is also adaptable to a washing of the vehicle 1, for example to a cleaning of the exterior part of the vehicle. In such an example of maintenance, the vehicle body is equipped with at least one dirt deposit sensor. The sensor is configured to send an alert upon detection of the exceeding of a threshold of dirt deposited on the bodywork. The server is configured to trigger the service if the user has left his vehicle, upon receipt of the alert.

In general, the use of the service as described does not require subscribing to a regular maintenance service and can also be a one-time use.

The remote server 5 can be implemented in a unitary or distributed manner. The assembly formed by the portable device (for example the connected spectacle frame 2) and the mobile terminal (for example the smartphone 8) can be implemented in a unitary or distributed manner.

Furthermore, the aforementioned training of the neural network can involve an operator who can be the user as described above. However, as a variant, it can be an autonomous learning of an artificial intelligence. In a typical example of such an embodiment, a computer means recognizes for example an upward movement of the spectacle frame 2 and interprets (according to a pre-recorded table of interpretation rules) that a user gets up from his seat to get out of the vehicle. The artificial intelligence marks (or “tags”) this sequence as a vehicle exit by a user. Then, in learning, if the user goes out in such a way that the movement of the mount makes an upward trip of a distance always included in the same restricted interval (for example between 1.00 m and 1.20 m), then this user precisely is clearly identified as leaving his vehicle.

Furthermore, various embodiments have been described allowing a remote server 5 to identify a vehicle and/or a user using an artificial intelligence trained to identify signature signals. In an advantageous option, the smartphone can also be configured to recognize a unique signature, based on a first prior identification of the unique signature by the server. In such an option, the signature is transmitted by the server 5 to the smartphone after the first prior identification, and stored in the memory in relation to an identifier of the vehicle and/or of the user. Thus the smartphone can be configured to carry out certain verifications that are less costly in terms of computing power than the identification carried out by the smartphone via artificial intelligence.

Claims (16)

Method, implemented by computer means, for identifying a motorized vehicle (1) as being stationary with a view to determining a location of this stationary vehicle, in which a portable device (2), available to a user of the vehicle, the portable device comprising a signal sensor unit (3), a geolocation module and a communication interface (7) with a remote server (5),
The process comprising:
- a preliminary step of learning, in the signals picked up by the sensor unit, of at least one signature of a predefined event, corresponding to a stopping of the vehicle and an exit, of the vehicle, of the user carrying the portable device (2),
- and a current stage of:
* detection (S106), in the signals picked up by the sensor unit, of said signature of the event of stopping the vehicle and leaving the user,
* determination of the last geolocation coordinates of the vehicle (1) thus detected as being stationary, and
* sending (S111) to the remote server (5) said geolocation coordinates,
The remote server being able to communicate with a mobile terminal available to the user, the method further comprises:
- detection of a need for vehicle maintenance (1),
- a sending, by the remote server (5), of a maintenance service proposal notification on the mobile terminal (8),
- the reception, by the remote server (5), of a maintenance service request sent by the mobile terminal (8),
- In response to the reception of the maintenance service request, the transmission of a start signal for the maintenance service, the signal comprising at least the geolocation coordinates. A method according to claim 1, wherein the prior learning step comprises:
- acquire a plurality of signals picked up by the sensor unit,
- identify, in each signal of the plurality of signals, whether at least a portion of the signal corresponds to the predefined event,
- associate with at least a portion of the signal a marker representative of the predefined event,
- training, by learning, an artificial intelligence in the detection, in signals picked up by the sensor unit, of said signature of the event, the training being carried out on the basis of each signal of the plurality of signals picked up and of the markers associated with said signal portion. A method according to any preceding claim, wherein, the sensor unit comprising an accelerometer capable of sensing vibrations of the engine of the running vehicle, the method further comprises:
- prior learning to identify a characteristic vibration of the engine in at least one predefined operating phase of the engine, the characteristic vibration being representative of an identity of the vehicle,
- and a current detection, in the signals picked up by the sensor unit, of said characteristic vibration,
The sending (S111) to the remote server (5) of said geolocation coordinates being conditioned by a positive detection (S100) of said characteristic vibration. A method according to claim 3, wherein the signature of said event comprises a characteristic vibration of engine operation, over a time interval, in at least one of the following situations:
- engine neutralization,
- engine shutdown. A method according to any one of the preceding claims, wherein the sensor unit (3) comprising at least one of an accelerometer, an altimeter, a gyroscope, a pressure sensor, a light sensor, a microphone,
the signature of said event includes a characteristic acceleration, a specific measurement of a movement of the user, over a time interval, in at least one of the following situations:
- the user (9) removes his seat belt,
- the user (9) opens the door on the driver's side,
- the user (9) makes a movement to leave the driver's seat,
- the user (9) makes a movement to pass through the door on the driver's side,
- the user (9) closes the door on the driver's side,
- the user (9) activates a closing mechanism of the vehicle. Method according to Claim 5, characterized in that the signature of said event comprises a succession of at least two characteristic accelerations corresponding to respective situations among said situations taken successively in order. Method according to any one of the preceding claims, in which the portable device (2) comprises said sensor unit (3) and a first radiofrequency communication module, able to communicate with a second radiofrequency module of a mobile terminal, at the disposal of the user, and connected via a wide area network to the remote server (5), and the method comprises:
- Transmission by the first module to the second module of the signals picked up,
- A transmission from the mobile terminal to the server of the signals received, and
- Processing with the server to identify a characteristic signal among the signals picked up. Method according to claim 7, in which, the mobile terminal comprising the geolocation module, in the event of identification of the signature of said event in the signals picked up, the server requests from the mobile terminal the said last geolocation coordinates of the vehicle (1). Method according to one of the preceding claims, in which the portable device (2) comprises a portable accessory, connected, available to the user and carrying the sensor unit (3), said sensor unit comprising at least one accelerometer,
And in which the accelerometer measures at least the movements of the user in order to detect characteristic gestures of the user, said characteristic gestures being representative of an identity of the user. The method of claim 9, wherein the accelerometer further measures engine vibration. Method according to one of Claims 9 or 10, in which the sending of the maintenance service proposal notification on the user's mobile terminal (8) is conditioned by the identification of the user by said characteristic gestures of the user. Method according to one of Claims 10 and 11, in which the detection of a need for maintenance of the vehicle comprises:
- prior learning in the signals picked up by the sensor unit, of a wear state signature of a predefined part of the vehicle (1),
- And a current detection, in the signals picked up by the sensor unit, of said state of wear signature of the predefined part. A method according to claim 12, wherein the detection of a need for maintenance of the vehicle comprises:
- prior learning in the signals picked up by the sensor unit, of a signature characteristic of a solicitation of a predefined part of the vehicle,
- And current detection, in the signals picked up by the sensor unit, of a drift in said signature characteristic of the stress on the predefined part. System for implementing the method according to one of the preceding claims, comprising:
- a portable device (2), available to a user of the vehicle, and including a signal sensor unit (3), a geolocation module, and a communication interface (7) with a remote server (5), and
- said remote server (5). Server of a system according to claim 14, configured to implement:
- a preliminary step of learning, in the signals picked up by the sensor unit, of at least one signature of a predefined event, corresponding to a stopping of the vehicle and an exit, of the vehicle, of the user carrying the portable device (2),
- a current stage of:
* detection (S106), in the signals picked up by the sensor unit, of said signature of the event of stopping the vehicle and leaving the user,
* and request for the last geolocation coordinates of the vehicle (1) thus detected as being stationary. Computer program comprising instructions for implementing the method according to one of claims 1 to 13, when said instructions are executed by a processor.
.