WO2022008837A1

WO2022008837A1 - Method for signalling the approach of a vehicle to a smartphone user and corresponding application

Info

Publication number: WO2022008837A1
Application number: PCT/FR2021/051245
Authority: WO
Inventors: Carlos PILOTO FONSECA; Rubén ALFONSO REYES; Emmanuel Ruiz
Original assignee: CopSonic; Regie Autonome Des Transports Parisiens
Priority date: 2020-07-09
Filing date: 2021-07-06
Publication date: 2022-01-13
Also published as: FR3112383B1; FR3112383A1

Abstract

The present invention relates to a method for signalling the approach of a vehicle to a smartphone user, in which the vehicle emits (310) an ultrasonic signal over a plurality of discrete frequencies, the distribution of the deviations between the discrete frequencies corresponding to a signature characteristic of the vehicle type. An application of the smartphone carries out (320) a spectral analysis of the ultrasonic signal received by the smartphone and measures (330) the strength of the received signal in a plurality of frequency channels. It researches (340) whether there is a plurality of frequency channels having substantially the same distribution of deviations and for which the strength of the received signal in each of them is higher than a predetermined threshold. If so, the application generates (350) an audio and/or visual and/or tactile alert on the smartphone.

Description

METHOD OF SIGNALING THE APPROACH OF A VEHICLE TO THE USER OF A SMARTPHONE AND CORRESPONDING APPLICATION

DESCRIPTION

TECHNICAL AREA

The present invention generally relates to the field of vehicle signaling systems as well as that of alert systems installed on a smartphone.

PRIOR ART

The ubiquity of consulting websites, messaging services or various applications is certainly one of the great advantages brought by smartphones. It is however not without risks for the user, when he moves in the public space because his attention, monopolized by the consultation, makes him ignore the dangers which surround him. This phenomenon, known as "smombie", resulting from the combination of smartphone and zombie, has caused many victims among pedestrians in urban centers with high traffic intensity, so much so that some cities have started to equip crossings pedestrian warning systems.

Various alert systems have been proposed in the state of the art.

Some systems, such as those described in applications CN103927904 and CN107274722 use the GPS coordinates provided by the pedestrian's smartphone and the navigation data provided by the vehicle's GPS system to estimate a probability of collision between the vehicle and the pedestrian, and in inform the latter by means of his smartphone. The calculation unit can be on board the vehicle or else be installed in a terminal close to the pedestrian crossing. These systems are complex and assume the establishment of a communication between the smartphone and the vehicle or the terminal.

Other systems, such as that described in the application KR101862986 are based on the recognition of an obstacle by the camera of the smartphone. However, these systems are not very selective in the sense that the alert level is independent of the real danger, in particular of the relative speed of the obstacle/pedestrian. As a result, the probability of a false alarm is high, either that the user is warned too late with the serious consequences that this entails, or that the user is warned in vain, which then reduces his subsequent vigilance.

Finally, other systems, such as that described in application KR20180032926, simply signal the presence of a danger by sending a Wi-Fi signal. road works to deploy them in traffic lights or bury them in the roadway. They also assume that Wi-Fi is permanently activated on pedestrian smartphones, which is rarely the case in outdoor environments and leads to increased energy consumption.

The object of the present invention is therefore to propose a method of signaling the approach of a vehicle to the user of a smartphone which does not have the defects of the prior art, in other words which is simple and robust, has a high level of discrimination and does not suppose to activate the Wi-Fi function permanently.

DISCLOSURE OF THE INVENTION

The present invention is defined by a method of signaling the approach of a vehicle to the user of a smartphone, in which the vehicle emits an ultrasonic signal on a plurality of discrete frequencies ( f ₀ ,...,f _Nx ), the distribution of the deviations between said discrete frequencies corresponding to a characteristic signature of a type of vehicle; an application of the smartphone performs a spectral analysis of the ultrasonic signal received by the smartphone and measures the intensity of the signal received in a plurality of frequency channels, said application searches whether there are a plurality of frequency channels having substantially the same said distribution deviations and for which the intensity of the signal received in each of them is greater than a predetermined threshold; and, if so, generates an audible and/or visual and/or tactile alert on the user's smartphone. The vehicle can emit the ultrasonic signal when it is in a predetermined geographical area or when its speed is greater than a predetermined threshold.

Alternatively, the vehicle may emit the ultrasonic signal when a vehicle horn is actuated by the driver. In the latter case, a sound watermarking signal can be superimposed on the sound signal, the sound watermarking signal being modulated to transmit information indicating a kinematic parameter of the vehicle.

Typically, the ultrasonic signal is emitted on a frequency doublet having a predetermined frequency deviation. The ratio of the frequency difference to the average frequency of said doublet can then be chosen to be less than 10%.

Advantageously, said application searches for a plurality of channels exhibiting substantially the same said distribution of deviations, starting with the plurality of channels respectively comprising said discrete frequencies, then simultaneously incrementing each of these channels.

In any case, the alert can only be generated if the smartphone detects a module speed or acceleration greater than a predetermined value.

Further, the visual alert may only be generated when the smartphone detects a gaze from the user.

The application can also generate different types of alerts depending on the geographical position of the smartphone.

According to a variant embodiment, at least one of said discrete frequencies is modulated to transmit information indicating a kinematic parameter of the vehicle. In this case, the smartphone can generate a first level alert when the kinematic parameter of the vehicle is lower, in absolute value, than a first threshold value and a second level alert when this parameter is higher in absolute value than a second value threshold.

The invention also relates to an application installed on a user's smartphone, which, when executed, performs spectral analysis of an ultrasound signal received by the smartphone and measures the intensity of the received signal in each channel of a plurality of frequency channels; searches whether there is a plurality of frequency channels substantially exhibiting a distribution of deviations stored in memory and for which the intensity of the signal received in each of them is greater than a predetermined threshold; and, if so, generates an audible and/or visual and/or tactile alert on the smartphone.

Finally, the invention relates to a signaling device on board a vehicle, intended to signal the approach of this vehicle to the user of a smartphone, characterized in that it comprises: at least one piezoelectric transducer; a generator adapted to generate a signal at a plurality of discrete frequencies (f ₀ ,...,f _Nx ), the distribution of the deviations between said discrete frequencies corresponding to a characteristic signature of a type of vehicle; the piezoelectric transducer adapted to emit an ultrasound signal from the signal at said plurality of discrete frequencies.

BRIEF DESCRIPTION OF DRAWINGS

Other characteristics and advantages of the invention will appear on reading a preferred embodiment of the invention, described with reference to the attached figures, including:

Fig. 1 schematically represents an example of context of use of the present invention;

Figs. 2A and 2B represent an example of spectrum of an ultrasonic signal emitted by the vehicle and of a signal received by the smartphone of FIG. 1;

Fig. 3 schematically represents a flowchart of the signaling method of the approach of a vehicle to the user of a smartphone according to one embodiment of the invention. DETAILED DISCUSSION OF PARTICULAR EMBODIMENTS

The context of the present invention is that of a traffic lane on which vehicles, in particular electric vehicles, can travel. The traffic lane can be a road, a railway track, a tram track, etc. The traffic lane is crossed by a pedestrian crossing, protected or not.

Fig. 1 schematically represents an example of context of use of the present invention.

A vehicle 110 travels on a traffic lane 120 and approaches a pedestrian crossing 130. The vehicle is equipped with transducers, 140, typically piezoelectric transducers, adapted to emit ultrasonic beams towards the front of the vehicle. These transducers may be integrated into a shock absorber, for example those employed in a pre-existing anti-collision system. The radiation pattern of the piezoelectric transducers advantageously has a wide main lobe, preferably having an angular aperture of the order of 120° in the horizontal plane and secondary lobes of low intensity relative to the main lobe. The distribution of the transducers and their radiation patterns are advantageously chosen so as to ensure complete coverage in front of the vehicle.

When the vehicle is an electric vehicle fitted with an AVAS system (Acoustic Vehicle Alert System), the transducers may be the loudspeakers mounted at the front of the vehicle and intended to emit an artificial sound when the vehicle is moving at a speed greater than a predetermined minimum value.

The pedestrian 150 is equipped with a smartphone, on which a specific application described later has been installed. The smartphone can use an Android or iOS operating system for example. It is equipped with a microphone, the bandwidth of which extends from 20Hz to a hundred KHz, makes it possible to receive sound and ultrasonic waves.

The specific application can be permanently activated in the background or can be executed only on user request. It may be provided for natively in the smartphone and selected by the user in the configuration menu or else be downloaded by the user from a server.

The application will be able to access certain resources of the smartphone, in particular the signal detected by the microphone, the GPS position of the smartphone, the measurements of sensors mounted in the smartphone, such as linear or angular speed sensor, linear acceleration sensor or angular, gravimeter, magnetometer, gyroscope, step counter etc.

Advantageously, the ultrasonic signal emitted by the transducers of the vehicle has a spectrum of lines located at a plurality of discrete frequencies, f ₀ ,...,f _{N x} . The distribution of gaps between the discrete frequencies, namely of -f_ = _x, z = l, .., / Vl is a feature of the vehicle type. For example, this distribution may be different for trams, motor vehicles, trains, buses, autonomous vehicles, etc.

The values of the frequencies i = 0, .., Nl, and consequently deviations in frequency of f = -f _l, z = l, .., / Vl, are known to the specific application and stored in memory. When a new type of vehicle is equipped with the approach signaling device according to the invention, the application can simply be updated to take it into account.

Preferably, the frequency deviations df,i=1,..,/V-1, will be chosen to be small with respect to the mean value of the discrete frequencies. For example, the

ratio of each of the frequency deviations to the average value will be less than 10% and, preferably, less than 1%, ie df<0.01, Vz.

According to a particular embodiment, the ultrasonic signal emitted by the vehicle will be a frequency doublet, in other words the spectrum of the signal will include two lines at the frequencies / ₀ and , the frequency difference d / = / ₁ -/ ₀ being a characteristic of the type of vehicle.

The sound signal received by the smartphone will also have a spectrum of lines whose frequencies are respectively affected by the Doppler shifts:

where n _q is the projection on the axis connecting the transducer to the smartphone, of the relative speed, v _r of the vehicle in relation to the pedestrian and u is the speed of sound in the air. The respective deviations between the lines of the received signal are therefore:

The relative difference between the doppler shift and the line shift of the received signal is therefore:

n _L given that — “cl for land vehicles. u

It is thus understood that the difference between the frequencies of successive lines is less affected by the Doppler shift than the frequencies of the lines themselves. The distribution of the deviations between the frequencies of the successive lines is advantageously used as the signature of the signal emitted by the vehicle.

To do this, the specific application embedded on the user's smartphone performs a spectral analysis of the ultrasound signal received, in practice an FFT over a time window of the sampled signal. The size of the FFT is chosen so as to be able to discriminate between discrete frequencies. For example if the ultrasonic signal is in the 20-40 kHz band, the sampling frequency, f _s , will be equal to the Nyquist frequency, i.e. 80 kHz. If we assume that the gap between the discrete frequencies is of 600 Hz, one can choose an FFT of size 2 M = 1024 , which corresponds to a number of frequency bins (frequency bins), M , equal to 512, the intervals being of width equal to 156 Hz, sufficient to discriminate between the different discrete frequencies.

Figs. 2A and 2B represent by way of example respectively the spectrum of an ultrasonic signal emitted by the vehicle and that of the ultrasonic signal received by the user's smartphone.

In the example illustrated, the transmitted signal has a triplet of lines at the frequencies /o,/ _i ,/ ₂ . For reasons of scale, only the part of the spectrum around these frequencies has been represented.

On reception, it can be seen that the lines have been respectively shifted by the values A ^d ° ^p , A ^d ₂ ^op , A ^d ° ^p . However, the distribution of the frequency deviations Af _x ,Af ₂ is substantially identical to the distribution of the frequency deviations d _l ,df ₂ in the spectrum of the transmitted signal.

The application estimates the intensity of the signal received in each of the M channels, c ₀ ,...,c _Mx and finds whether there are N channels having substantially the same distribution of deviations as in the spectrum of the transmitted signal, and for which the intensity of the signal received in each of them is greater than a predetermined threshold value. The threshold value is chosen from the noise level generally measured in the band of interest in an urban context, and from the probability of false alarm. If there are N channels,

verifying:

2 where S(c _m ) is the intensity of the signal received in channel c _m , s is the threshold value and f(c _m ) is the frequency of channel c _m . By substantially identical distribution of deviations, - d , i = \,..,N-\ is meant that it is identical to that relating to the emitted signal, to within the degree of spectral resolution, in other words to that of the resolution of the TFF:

Advantageously, to accelerate the search, the frequency channels whose frequency is less than / ₀ are not taken into consideration because they correspond to a negative relative speed of the vehicle with respect to the pedestrian (vehicle moving away from the pedestrian). For a given type of vehicle searched, the search preferably begins only from c _init such that / ₀ −f{c _init ) < fjM .

The search can then progress by increasing channel index. As soon as a first channel c _fflo satisfying > s , we search if a subsequent channel c such that

2 verifies S(c > s . In the affirmative, we continue with the

subsequent channel up to the channel corresponding to the last discrete frequency.

Of course, the application may be able to search for vehicles of several types, in which case the search will begin with the channel corresponding to the lowest frequency / ₀ among those of the different possible types of vehicles. The search can be carried out in series or in parallel for the different types of vehicles. In the case of a sequential search, it will be possible to begin with the type of vehicle presenting the greatest danger for the user.

Different strategies for searching for the distribution of the frequency deviations may be envisaged without thereby departing from the scope of the present invention.

If the vehicle detects an approaching vehicle of known type, it can determine its relative speed from the doppler shifts A ^d ° ^p . Indeed, once the N channels have been identified, the deviations D/j can be determined by:

The relative speed, v _r , of the vehicle in relation to the pedestrian (more precisely its projection, v _g ) can then be estimated by performing a linear regression on the coordinate points (<5/j,Â/ : = {r ^~ ) ^Ό (7) where p is the slope obtained by the linear regression. Alternatively, the linear regression could relate to the coordinate points and the speed to be estimated by the

same formula.

According to a variant embodiment of the invention, a kinematic parameter of the vehicle, for example its speed or its acceleration, could be transmitted by modulating at low frequency at least one of the lines of the ultrasonic signal, for example that at the frequency of base / ₀ . The modulation will advantageously be a BPSK or DBPSK modulation, or even a low-excursion frequency modulation with respect to fjM so as not to disturb the search for frequency deviations.

According to another variant embodiment, given that the speed of sound varies according to the temperature (in square root of the temperature expressed in degrees Kelvin) and the degree of hygrometry, the discrete frequencies of the transmitted signal can be compensated according to of the temperature or of the degree of hygrometry, so that the f ratio — remains constant in spite of these variations. u

The ultrasonic signaling signal may be emitted continuously by the vehicle. However, different variants could be envisaged to selectively restrict the emission thereof. According to a first variant, the ultrasonic signal will only be emitted when the vehicle is only in a predetermined geographical zone. This zone can correspond to an urban zone for example or be that defining a pedestrian crossing.

According to a second variant, the ultrasonic signal will only be emitted if its speed is greater than a predetermined threshold.

According to a third variant, the ultrasonic signal will only be emitted if its audible warning device (car horn, tram gong, train siren) is activated. In this case, the sound signal can be watermarked, that is to say superimposed with an audio watermarking signal transmitting a kinematic parameter of the vehicle. This watermarking signal is then detected by the smartphone in the usual audio band.

These different variants can be combined without departing from the scope of the present invention.

On the smartphone side, the generation of the alert may be conditioned by a certain number of criteria so as to prevent untimely alerts or even to graduate them according to the level or type of danger.

For example, alerts of different levels could be generated according to the geographical position of the user. When it is close to a pedestrian crossing or a traffic lane, the danger will be at the maximum level. On the other hand, when the latter is not in a space dedicated to traffic, the alert will be of a lower level or even purely and simply inhibited.

In another configuration, the alert will only be generated if the smartphone user is moving, i.e. when the modulus of the speed and/or acceleration measured by the smartphone is greater than a predetermined value or if the step counter varies.

In yet another configuration, the application will only generate an alert when the relative vehicle speed, estimated from the Doppler shifts as described above, is greater than a predetermined value or even when the vehicle speed as transmitted by modulation of at least one line of the ultrasonic signal or in the watermarking signal, is greater than a predetermined value.

The alert generated by the application may be visual and/or sound and/or tactile. The type of alert can be selected or conditioned by the status of the smartphone. For example, if headphones are plugged in, the audible alarm will be preferred and if the user's gaze is detected (Smart Screen function on Android for example) the visual alert will be preferred. The tactile alert may be a vibration or a particular sequence of vibrations of the smartphone.

Fig. 3 schematically represents a flowchart of a method for signaling the approach of a vehicle to the user of a smartphone according to one embodiment of the invention.

At step 310, the vehicle emits an ultrasonic signal exhibiting a spectrum of lines situated at a plurality of discrete frequencies f _t , i=0,..,N1. The distribution of frequency deviations 5f _i =f _{i -}

is given by the type of vehicle.

This issue can be permanent, triggered or subject to the fulfillment of certain conditions as seen above.

The values of the discrete frequencies can be compensated according to the variations of temperature and hygrometry.

A kinematic parameter such as the speed/acceleration of the vehicle can be transmitted by modulating at low frequency at least one of the lines of the ultrasonic signal.

The smartphone application performs a loop comprising steps 320-350.

At step 320, the application performs a spectral analysis of the ultrasonic signal, typically an FFT on the samples of this signal in a time window then calculates, at 330, the intensity of the signal received in the different frequency channels.

At step 340, the application searches if there is a plurality N of frequency channels, , satisfying the conditions (4), in other words if they present

substantially the same distribution of frequency deviations as that of the transmitted spectrum and if the intensity of the signal received in each of them is greater than a predetermined threshold.

If not, the application returns to step 320 for a new spectral analysis. If so, the application generates an audible and/or visual and/or tactile alert at 350, where applicable subject to an additional condition being met, as described in the aforementioned variants.

This alert can remain for a predetermined duration at the end of which the application returns to step 320. Alternatively, it can wait for an acknowledgment from the user before ceasing, the application returning to step 320 once the alert acknowledged.

Claims

1. Method of signaling the approach of a vehicle to the user of a smartphone, characterized in that: the vehicle emits (310) an ultrasonic signal on a plurality of discrete frequencies (f ₀ ,..., f _Nx ), the distribution of the deviations between said discrete frequencies corresponding to a characteristic signature of a type of vehicle; an application of the smartphone performs (320) a spectral analysis of the ultrasound signal received by the smartphone and measures (330) the intensity of the signal received in a plurality of frequency channels, said application searches (340) if there is a plurality of frequency channels having substantially the same said distribution of deviations and for which the intensity of the signal received in each of them is greater than a predetermined threshold; and, if so, generates (350) an audible and/or visual and/or tactile alert on the user's smartphone.

2. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 1, characterized in that the vehicle emits the ultrasonic signal when it is in a predetermined geographical area.

3. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 1, characterized in that the vehicle emits the ultrasonic signal when its speed is greater than a predetermined threshold.

4. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 1, characterized in that the vehicle emits the ultrasonic signal when a vehicle horn is activated.

5. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 4, characterized in that when the vehicle horn is activated, a sound tattoo signal is superimposed on the signal sound, the sound watermarking signal being modulated to transmit information indicating a kinematic parameter of the vehicle.

6. Method of signaling the approach of a vehicle to the user of a smartphone according to one of the preceding claims, characterized in that the ultrasonic signal is emitted on a frequency doublet having a predetermined frequency deviation.

7. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 6, characterized in that the ratio of the frequency deviation to the average frequency of said doublet is less than 10%.

8. Method of signaling the approach of a vehicle to the user of a smartphone according to one of the preceding claims, characterized in that the application searches for a plurality of channels having substantially the same said distribution of deviations starting with the plurality of channels respectively comprising said discrete frequencies, then simultaneously incrementing each of these channels.

9. Method of signaling the approach of a vehicle to the user of a smartphone according to one of the preceding claims, characterized in that the alert is generated only if the smartphone detects a speed or an acceleration modulus greater than a predetermined value.

10. Method of signaling the approach of a vehicle to the user of a smartphone according to one of the preceding claims, characterized in that the visual alert is only generated when the smartphone detects a gaze from the 'user.

11. Method of signaling the approach of a vehicle to the user of a smartphone according to one of the preceding claims, characterized in that the application generates alerts of different types depending on the geographical position of the smartphone .

12. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 1, characterized in that at least one of said discrete frequencies is modulated to transmit information indicating a kinematic parameter of the vehicle.

13. Method of signaling the approach of a vehicle to the user of a smartphone according to claim 12, characterized in that the smartphone generates a first level alert when the kinematic parameter of the vehicle is lower, in absolute value , to a first threshold value and a second level alert when this parameter is greater in absolute value than a second threshold value.

14. Application installed on a user's smartphone, characterized in that, when executed, performs (320) a spectral analysis of an ultrasound signal received by the smartphone and measures (330) the intensity of the received signal in each channel of a plurality of frequency channels; searches (340) if there are a plurality of frequency channels having substantially a distribution of deviations stored in memory and for which the intensity of the signal received in each of them is greater than a predetermined threshold; and, if so, generates (350) an audible and/or visual and/or tactile alert on the smartphone.

15. Signaling device on board a vehicle, intended to signal the approach of this vehicle to the user of a smartphone, characterized in that it comprises at least one piezoelectric transducer; a generator adapted to generate a signal at a plurality of discrete frequencies (f ₀ ,...,f _Nx ), the distribution of the deviations between said discrete frequencies corresponding to a characteristic signature of a type of vehicle; the piezoelectric transducer adapted to emit an ultrasound signal from the signal at said plurality of discrete frequencies.