GB2613002A - Method and system of vehicle occupants stress detection by using a pressure sensor network - Google Patents

Abstract

Method of detecting stress level of at least two vehicle occupants. Multiple pressure sensors, each weighted, are disposed on various components of the cabin, including seat, armrests, steering wheel, gear knob, floor, door handle. Only pressure and frequency of touch of each sensor associated with occupied seats are monitored during a predetermined monitoring time period. Pressure and frequency margin are selected to define their respective alert interval. Alert interval weighted sum being when pressure values are inside pressure alert interval and when frequency values are inside alert interval weighted by each sensor weight; and a margin interval weighted sum being the total number of occurrences when pressure values exceed pressure margin and frequency values exceed frequency margin weighted by each sensor weight are defined. Vehicle stress level is high if sum between alert interval weighted sum and margin interval weighted sum is above alert interval threshold. An alert is then generated to trigger action from the driver and/or driver assist systems if stress is high. Max and min level may be a percentage coefficient of the nominal value. Stress levels may be discretised or classified into different degrees of danger. Different actions may be predefined for different degrees of danger.

Description

DESCRIPTION

Method and system of vehicle occupants stress detection by using a pressure sensor network

Field of the invention

The present invention relates to the detection of vehicle occupants stress based on the occupants' body interactions with some specific areas inside the vehicle as sensed by a pressure sensor network.

Terms to be used in the invention Throughout this invention the following terms have the following corresponding meanings: - pressure sensors, alternatively called in short sensors unless specified otherwise, are pressure sensors configured by their respective manufacturers to sense pressure exercised by human touch, - vehicle is any land vehicle that is or can be provided with the pressure sensors as described above: a passenger car, a truck, a trailer roll, a bus, a crane, an earth moving equipment, a tractor, a motorcycle, -vehicle occupant, alternatively called in short occupant, is a person riding inside the vehicle as described above. The vehicle occupant can be the driver or can be any passenger, the terms driver and passenger having their common meaning.

Background of the invention

Driving is a complex task that requires full driver concentration and a calm attitude. Unfortunately, regardless of being good or bad drivers, all drivers are exposed to stressful driving situation. Bad weather, long commutes, busy traffic, distraction due to multitasking, work/personal related issues can all make more likely that the drivers will become stressed.

Road accidents reduction is one of the most important topic automotive industry is trying to solve. And researches have showed that stress while driving is one major factor contributing to the total number of accidents.

Stress can have both mental and physical effects on a driver.

The perception of surrounding traffic is crucial for drivers and vehicle occupants. The drivers rely on their senses to control their vehicles in a goal-oriented and reliable manner. Unfortunately, the stress is one of the most important factors that may affect the driver's senses.

The increasing use of on-board electronics and in-vehicle information systems has made the evaluation of driver task demand an area of increasing importance to industry.

Most commonly researched methods to detect driver's stress level are: - Eyes and facial expression monitoring systems via integrated on-board 10 cameras.

- Driver's action monitoring such as accelerating, braking, or engaging the clutch.

- Health monitoring sensors (electromyogram, electrocardiogram, skin conductivity, respiration effort, temperature).

-On-road metrics (GPS traces, traffic, novelty of the area...).

Disadvantages of prior art

Firstly, the monitoring systems for driver's health, behaviour or eye/ facial expression prove to be very costly and difficult to obtain which limit drastically their 20 use.

Secondly, prior art deals with detecting only the stress level of the driver. However, in some cases, the driver may not be aware of the danger and can take a wrong decision leading to accidents or the vehicle.

Problem to be solved by the invention The technical problem to be solved is to find a robust and easy to implement solution to detect the stress level for more than one vehicle occupant.

Summary of the invention

In order to overcome the disadvantages of prior art, in a first aspect of the invention it is presented a method of stress detection of at least two vehicle occupants including the driver, the method comprising the following steps: Si Arranging a plurality of pressure sensors in a Pressure Sensor Network, including: positioning each pressure sensor in the proximity of one vehicle occupant's seat, configuring a predetermined monitoring time period defining for each vehicle occupant's seat an occupancy pressure sensor configuring parameters of each pressure sensor: a Nominal Pressure Value, a Nominal Frequency Value, a sensor weight, configuring an additional parameter for the vehicle occupancy pressure sensor: a Yes/No Presence Value, S2 Disconnecting the pressure sensors corresponding to the vehicle occupant's seats for which the occupancy pressure sensor has the No value, and continuing the steps of the method only for the pressure sensors corresponding to the vehicle occupant's seats for which the occupancy pressure sensor has a Yes value, S3 Measuring by each pressure sensor of the pressure value and of the frequency value of the touch exercised by each vehicle occupant on said respective pressure sensor during the predetermined monitoring time period and sending measurements to a Data Processing Unit, S4 Defining a vehicle stress rule by a Rules Unit: setting for each sensor: at least one Pressure Margin value being either one or both between a Max.

Pressure Level value and a Min. Pressure Level value, where the Min. Pressure Level value < the Nominal Pressure Value < the Max Pressure Level value, and at least one Frequency Margin value being either one or both between a Max.

Frequency Level value and a Min. Frequency Level value, where the Min. Frequency Level value < Nominal Frequency Value < Max. Frequency Level value defining alert intervals for each sensor: at least one pressure Alert Interval between the Nominal Pressure Value and one of the Pressure Margin values, at least one frequency Alert Interval between the Nominal Frequency Value and one of the Frequency Margin values, defining two parameters referring to occurrences for the ensemble of the sensors: an alert interval weighted sum = the total number of occurrences during the predetermined monitoring time period when the pressure values are inside one of the at least one pressure Alert Interval and when the frequency values are inside one of the at least one frequency Alert Interval weighted by each sensor weight, a margin interval weighted sum = the total number of occurrences during the predetermined monitoring time period when the pressure values exceed one of the Pressure Margin values, and the frequency values exceed one of the Frequency Margin values, weighted by each sensor weight, defining an alert interval threshold = the total number of occurrences during the predetermined monitoring time period above each of the alert interval weighted sum and margin interval weighted sum, defining a vehicle stress rule: the vehicle level of stress for all occupants is high if the sum between the alert interval weighted sum and the margin interval weighted sum is above the alert interval threshold, sending the vehicle stress rule to the Data Processing Unit, S5 Checking by the Data Processing Unit if the measurements of the occurrences received within the predetermined monitoring time period fulfil the vehicle stress rule.

56 Action by a Response Control Unit, If the vehicle stress rule is fulfilled, sending by the Data Processing Unit of an alert signal to the Response Control Unit in order to trigger an action from the driver and/or driver automatic assist systems.

If the vehicle stress rule is not fulfilled taking no action.

In a second aspect of the invention it is presented a system of vehicle occupants stress detection comprising a Pressure Sensor Network, a Rules Unit, a Data Processing Unit, a Response Control Unit, a communication network between the Pressure Sensor Network, the Rules Unit, the Data Processing Unit and the Response Control Unit, configured to apply the method of the invention of any embodiment.

In a third aspect of the invention it is presented computer program comprising instructions which, when the program is executed by the system of the invention, 5 causes the system to carry out the steps of the method of the invention of any embodiment.

Further advantageous embodiments are the subject matter of the dependent claims.

Advantages The main advantages of using the invention are as follows: - The method of the invention is robust allowing to easily fit into the majority of the embedded systems without requiring high memory or high processing power, which makes it highly affordable.

- The invention addresses those situations when any of the passengers may become aware of a danger, not only the driver, thus increasing generally the level of safety when driving.

- The system of the invention is easy to implement due to the fact that pressure sensors are widely used in automotive industry and due to the fact that the system can be interconnected with other stress detection and drive assist solutions.

List of drawings Further special features and advantages of the present invention can be taken from the following description of an advantageous embodiment by way of the accompanying drawings: Fig. 1 illustrates the working of the method and the components of the system.

Fig. 2 depicts an example of positioning the pressure sensors for the driver.

Fig. 3 depicts an example of positioning the pressure sensors for the front seat passenger.

Fig. 4 depicts an example of positioning the pressure sensors for each of the back seats.

Fig. 5 depicts an example of positioning the pressure sensors on the vehicle seats. Fig. 6 depicts the principle behind defining the alert intervals.

Detailed description

Referring to Fig. 1, the system of the invention comprises the following components: - A Pressure Sensor Network PSN, -A Rules Unit RU, - A Data Processing Unit DPU, - A Response Control Unit RCU A communication network between the Pressure Sensor Network PSN, the Rules Unit RU, the Data Processing Unit DPU and the Response Control Unit RCU The communication network uses protocols according to prior art such that but not limited to CAN bus, LIN, Ethernet, WIFI, etc. In step 1 it is arranged the Pressure Sensor Network PSN comprising a plurality of sensors.

The Pressure Sensor Network PSN is a network comprising a plurality of pressure sensors Si placed in specific areas inside the vehicle in the proximity of each vehicle occupant's seat, including the driver's seat.

The minimum number of vehicle occupants for working the invention is two: the driver and one passenger.

The maximum number of vehicle occupants for working the invention is pre-determined according to studies and experiments especially for the vehicles having a large number of seats. In other words, even if the vehicle has a large number of seats-for example 40 in case of a bus, not all the seats will be equipped in their proximity with sensors, but only some selected seats most probably the ones placed in the first row because they have better road visibility.

Throughout the invention, the term "seat" shall refer only to those seats of the vehicle that have been selected to be provided in their proximity with sensors of the Pressure Sensor Network PSN for the purpose of working of the invention and the term "occupant" shall refer to the person-driver or passenger, occupying one of the seats provided in their proximity with sensors.

The arrangement of the Pressure Sensor Network PSN includes: - Positioning each pressure sensor Si in the proximity of one vehicle occupant's seat, - Defining for each vehicle occupant's seat an occupancy pressure sensor, - Configuring parameters of each pressure sensor Si, - Configuring an additional parameter for the vehicle occupancy pressure sensor: The placement of the sensors Si is chosen in correspondence with the number of seats of the vehicle based on the studies disclosing the interaction between the occupants and some selected surfaces inside the vehicle in his near proximity based on the finding that the occupants have the tendency, when in stressful state, to interact with said selected surfaces, such as but not limited to middle console armrest door armrest, door pull handle. The method is based on the finding that in general, the more stressed the vehicle occupant is, the more he interacts with the selected surfaces that are in his near proximity.

An example of positioning the pressure sensors Si inside the vehicle is described below with reference to Figs. 2 to 4 for the purpose of exemplifying the invention and not for limiting it. The positioning of the pressure sensors Si of Figs. 2 to 4 can suffer variations based on the interior vehicle architecture. In Figs. 2 to 4 the vehicle is a passenger car. The person skilled in the art will understand that the same concept applies for positioning the pressure sensor Si inside any other land vehicle that can be provided with pressure sensors and with communication system between the components of the system: a passenger car, a truck, a trailer roll, a bus, a crane, an earth moving equipment, a tractor, a motorcycle, etc. With reference to Fig. 2, the pressure sensors Si for the driver are placed in the driver's proximity: steering wheel, shift gear knob, door armrest, door pull handle, middle console armrest, left footrest area, roof handle, knee contact area on middle console. In each specific place one or more pressure sensors Si can be positioned.

For example, in Fig. 2 there are two pressure sensors Si on the steering wheel. With reference to Fig. 3, the pressure sensors Si for the front seat passenger are placed in the front seat passenger's proximity: door armrest, middle console armrest, footrest area, door pull handle, roof handle, outside surface of the glove compartment lid, dashboard surface, knee contact area on middle console. In each specific place one or more pressure sensors Si can be positioned. For example, in Fig. 3 there are two pressure sensors Si in the foot rest area.

With reference to Fig. 4, the pressure sensors Si for each of the back seat passengers are placed in the rear seats passenger's proximity: top area door, armrest door, armrest middle console, roof handle, door pull handle, seat side pad, seat back cushion-knee area footrest area.

An example of positioning of the pressure sensors Si corresponding the seats of the vehicles is given in Fig. 5 for the purpose of exemplifying the invention and not for limiting it: headrest, seat side pads, lumbar support, seat wings, seat cushion pad, seat side bolster.

Other areas for sensors placement can be chosen based on interior vehicle 10 architectures.

Each sensor Si from the plurality is configured to carry out the following: - to measure the pressure exercised by each vehicle occupant on their respective sensor surface during a predetermined monitoring time period Tm, the pressure expressed as pressure value Vpi, -to measure the frequency of touching of said respective sensor by said each vehicle occupant during the predetermined monitoring time period Tm, the frequency expressed as frequency value Vfi.

- to send the measurements to the Data Processing Unit DPU. The measurements of the pressure value Vpi of the frequency value Vfi and of the predetermined monitoring time period Tm are carried out according to prior art using a timestamp t for each measurement.

The following parameters are configured for each pressure sensor Si: - a Nominal Pressure Value NPV -is a reference for the pressure value Vpi corresponding to a normal level of stress of the occupant based on studies and 25 tests, - a Nominal Frequency Value NFV -is a reference for the frequency value Vfi corresponding to a normal level of stress of the occupant based on studies and tests - a sensor weight kJ referring to the importance of the respective sensor Si, the value of each weight ki based on studies and tests.

For example, the sensor weight(s) kJ for the driving wheel are the highest among all weights, because the driving wheel is very important to the driving.

For each vehicle occupant's seat, it is defined an occupancy pressure sensor Si, whose role is to indicate the presence or absence of the occupant in the vehicle.

It is usually placed somewhere on the surface of the seat. The occupancy pressure sensor is one of the sensors Si that is additionally configured for this purpose, by means of a Yes/No Presence Value YNPV logical parameter that will assess if the occupancy pressure sensor is generating pressure values, in other words, if the seat is occupied by an occupant or not. This logical parameter has only two values: YES and NO.

In step 2 the occupancy pressure sensor detects the presence or absence of the respective vehicle occupant in each of the possible vehicle occupants' seats and disconnects the pressure sensors for all the occupants seats where the parameter Yes/No Presence Value YNPV returns a NO value. The method will be carried out only for the pressure sensors for all the occupant's seats where the parameter Yes/No Presence Value YNPV returns a YES value.

In step 3 the plurality of pressure sensors Si for all the occupant's seats where the parameter Yes/No Presence Value YNPV returns a YES value measures the pressure value Vpi and the frequency value Vfi during the predetermined monitoring time period Tm applying the time stamp to each measurement and sends the measurements to the Data Processing Unit DPU.

The Data Processing Unit DPU comprises at least one computer processing unit CPU core at least one volatile memory RAM and at least one non-volatile memory ROM, the respective configuration of which is according to prior art.

Non limiting examples of Data Processing Units DPU are laptops, computers or controllers, electronic control units.

The Data Processing Unit DPU is configured: - to receive the measurements from the Pressure Sensor Network PSN, - to interpret the results of measurements according to a vehicle stress rule that is received from the Rules Unit RU, - to send alert signals to the Response Control Unit in case the vehicle level of stress for all occupants is high according to the vehicle stress rule.

The vehicle stress rule is the one that establishes if the level of stress for all occupants is high or not. Said vehicle stress rule is defined by the Rules Unit RU.

The Rules Unit RU comprises at least one computer processing unit CPU core at least one volatile memory RAM and at least one non-volatile memory ROM, the respective configuration of which is according to prior art.

Non limiting examples of Rules Unit RU are laptops, computers or controllers, electronic control units.

In a preferred embodiment, the Rules Unit RU is a different hardware entity from the Data Processing Unit DPU, for example the Rules Unit RU is provided by the manufacturer of the pressure sensors.

In another preferred embodiment, the Rules Unit RU is the same hardware entity with the Data Processing Unit DPU, functioning as a Rules and Data Processing Unit RDPU.

The Rules Unit RU is configured to define the vehicle stress rule in step 4.

In step 4 the Rules Unit RU carries out the following: Firstly the Rules Unit RU sets for each sensor Si: at least one Pressure Margin PM value being either one or both between a Max Pressure Level MPL value and a Min. Pressure Level MPL value, where the Min. Pressure Level MPL value < the Nominal Pressure Value NPV < the Max. Pressure Level MPL value, and at least one Frequency Margin FM value being either one or both between a Max Frequency Level MFL value and a Min. Frequency Level MFL value, where the Min. Frequency Level MFL value< Nominal Frequency Value NFV < Max. Frequency Level MFL value.

The selection of the Pressure Margin PM values and respectively Frequency Margin FM values depend on the type of pressure sensors Si. For some sensors Si, only one of the margins is relevant: either the maximum or the minimum, whereas for other sensors, both margins are relevant.

For example for the sensors Si placed on the driving wheel both margins are relevant because it is considered that in a normal situation the driver is exerting a continuous pressure on the wheel. If the pressure value Vpi is below the Min. Pressure Level MPL value, the driver may be too relaxed or may fall asleep, whereas if the pressure value Vpi is above the Max. Pressure Level MPL value, this indicates a rising level of the stress.

In another example, for the sensors Si placed on the dashboard-front seat passenger, the normal values are when the dashboard is not touched, thus the Min.

Pressure Level MPL value is relevant.

Then the Rules Unit RU defines alert intervals for each sensor Si, as depicted suggestively in Fig. 6: at least one pressure Alert Interval Al between the Nominal Pressure Value NPV and one of the Pressure Margin PM values, at least one frequency Alert Interval Al between the Nominal Frequency Value NFV and one of the Frequency Margin FM values, Then the Rules Unit RU defines two parameters referring to occurrences MS; for the ensemble of the sensors Si: an alert interval weighted sum SUMnAi = the total number of occurrences MS; during the predetermined monitoring time period Tm when the pressure values Vpi are inside one of the at least one pressure Alert Interval Al and when the frequency values Vf; are inside one of the at least one frequency Alert Interval Al weighted by each sensor weight k, calculated as follows: SUMnAl= >MS1 otto X ki and a margin interval weighted sum SUMnv = the total number of occurrences MS; during the predetermined monitoring time period Tm when the pressure values Vp; exceed one of the Pressure Margin PM values, and the frequency values Vf; exceed one of the Frequency Margin FM values, weighted by each sensor weight ki, calculated as follows: SUMnv= Er," M.St (v) x ki Then the Rules Unit RU defines an alert interval threshold Ntotal = the total number of occurrences MS; during the predetermined monitoring time period Tm above each of the alert interval weighted sum SUMnAl and the margin interval weighted sum SUMnv.

At the end of step 4 the Rules Unit RU defines the vehicle stress rule: the vehicle level of stress for all occupants is high if the sum between the alert interval weighted sum SUMnA; and the margin interval weighted sum SUMn" is above the alert interval threshold N total.

SUMa; + SUMnv > Ntotal The vehicle stress rule is sent at the end of step 4 by the Rules Unit RU to the Data Processing Unit DPU for implementation.

In step 5 the Data Processing Unit DPU receives the measurements of the occurrences MS; from the Pressure Sensor Network PSN during the predetermined monitoring time period Tm fulfil the vehicle stress rule and checks if vehicle stress rule is fulfilled or not. This step is carried out continuously throughout the functioning of the engine for successive predetermined monitoring time periods Tm as long as the condition of at least two vehicle occupants is fulfilled.

The checking of the vehicle stress rule is carried out as follows: The Data Processing Unit DPU counts for all sensors S; the number of occurrences MS; when the pressure value Vp; belongs to at least one of the pressure alert intervals Al and the frequency value Vf; belongs to at least one of the frequency alert intervals Al, and the number of occurrences MS; when the pressure value Vp; exceeds one of the Pressure Margin PM values and the frequency value Vf; exceeds one of the Frequency Margin FM values Then, the Data Processing Unit DPU computes the alert interval weighted sum SUMnA; and the margin interval weighted sum SUMnv and compares their sum SUMa; + SUMnv with the alert interval threshold Ntotal.

There are various types of interactions to be assessed by the Data Processing Unit DPU based on the position of the pressure sensors Si: continuous interaction -surfaces like "drive wheel", "footrest areas", "seats cushion pads", "seats lumbar support" are in continuous contact with the driver/ passengers. That means the sensors Si placed in that areas should generate continuous readings of the pressure values. The values will be assessed by the DPU based on comparison with the at least one of the Pressure Margin PM values, regular interaction -surfaces like "shift gear knob", "door armrest", "middle console armrest" are often in contact with the driver/passengers, but not continuously. In this case, the Data Processing Unit DPU will assess not only the at least one of the Pressure Margin PM values but also the frequency of touching, that is the comparison with the at least one of the Frequency Margin FM values.

rare interaction-surfaces like "door pull handle", "roof pull handle", "dashboard", "seat side pads" are rarely in interaction with the car's occupants. In this case, the Data Processing Unit DPU will assess not only the at least one of the Pressure Margin PM values but also the frequency of touching, that is the comparison with the at least one of the Frequency Margin FM values.

The analysis performed by the Data Processing Unit DPU is as follows: the measurements received from the Pressure Sensor Network PSN is compared with the parameters of each pressure sensor Si.

the measurements received from the Pressure Sensor Network PSN are 15 compared with measurements of the previous predetermined monitoring time periods Tm and compared with the alert interval threshold Ntotal.

A deep learning algorithm can be implemented in the Data Processing Unit DPU and/or in the Rules Unit RU. The purpose of the deep learning algorithm is to learn over time the behaviour during driving for frequent occupants, such as for example the driver and his family. For this type of analysis, the Data Processing Unit DPU will adjust the parameters of each pressure sensor, whereas the Rules Unit RU will adjust the Pressure Margin PM value, the Frequency Margin FM value, the alert intervals, the parameters referring to occurrences, etc. In case the vehicle stress rule is fulfilled, the Data Processing Unit DPU sends in step 6 an alert signal to the Response Control Unit RCU in order to trigger an action from the driver and/or driver automatic assist systems.

In case the vehicle stress rule is not fulfilled, there is no alert signal sent in step 6 to the Response Control Unit RCU and, consequently no action is necessary from the driver and/or driver automatic assist systems.

The Response Control Unit RCU is configured to be connected with the audio, video, infotainment systems of the vehicle and, when the alert signal is received from the Data Processing Unit DPU will trigger a type of stimulus that will make the driver or the driver and the passengers aware of a potential danger.

The Response Control Unit RCU is configured to be connected to the drive assist systems of the vehicle. In case the audio and video stimulus will not improve the values from the PSN, the RCU may trigger automatic assist system actions that will avoid the potential risky situation such as braking, speed reduction, lane change, warning lights, etc. In a preferred embodiment, the at least one Alert Interval Al is set as follows: Max. Pressure Level MPL value-is the Nominal Pressure Value NPV increased with a predetermined max pressure coefficient pmax%, Min. Pressure Level MPL value-is the Nominal Pressure Value NPV decreased with a predetermined min. pressure coefficient prnin %, Max. Frequency Level MFL value-is the Nominal Frequency Value NFV value increased with a predetermined max frequency coefficient fmax%, Min. Frequency Level MFL value-is the Nominal Frequency Value NFV value decreased with a predetermined min frequency coefficient fmin%, Positive Pressure Level Alert PPLA value is the Nominal Pressure Value NPV increased with a predetermined alert coefficient ac% Negative Pressure Level Alert NPLA value is the Nominal Pressure Value NPV decreased with the predetermined alert coefficient ac% Positive Frequency Level Alert MFLA value is the Nominal Frequency Value NFV value increased with the predetermined alert coefficient ac% Negative Frequency Level Alert MFLA value is the Nominal Frequency Value NFV value decreased with the predetermined alert coefficient ac% There are two types of Alert Interval Al, as follows: a positive alert interval above the Positive Pressure Level Alert PPLA, respectively Positive Frequency Level Alert MFLA but below the Max.

Pressure Level MPL value and, respectively, the Max. Frequency Level MFL value, or a negative alert interval above the Min. Pressure Level MPL value, respectively the Min. Frequency Level MFL value and the Negative Pressure Level Alert NPLA value, respectively Negative Frequency Level Alert MFLA value The coefficients applied to the minimum values are equal or unequal to the coefficients applied to the maximum values.

Non-limiting example of coefficients: the predetermined alert coefficient aa% = 20%, the predetermined max pressure coefficient pmax% = 50%, the predetermined min. pressure coefficient!Drain% = 50%, the predetermined max frequency coefficient fmax% = 50%, the predetermined min frequency coefficient fmia% =50%.

In the example above, the Alert Interval Al is calculated as follows for the pressure value Vpi: The negative alert interval is Nominal Pressure Value NPV -Nominal Pressure 10 Value NPV X 50% < negative alert interval < Nominal Pressure Value NPV -Nominal Pressure Value NPV X 20%, The positive alert interval is Nominal Pressure Value NPV + Nominal Pressure Value NPV X 20% < positive alert interval Nominal < Pressure Value NPV + Nominal Pressure Value NPV X 50%. The same calculation method applies for the frequency value Vfi.

The Nominal Pressure Value NPV, the Nominal Frequency Value NFV, the predetermined max pressure coefficient pmax%, predetermined min. pressure coefficient prain % predetermined max frequency coefficient fmax predetermined max frequency coefficient fmax and the Alert Interval Al are determined based on studies and testing and can be adjusted over time based on new studies and testing and/or on deep learning algorithm. The studies consider, among others, parameters like gender, weight and height of the drivers/passengers. As the pressure sensors are grouped based on the number of seats of the vehicle, depending on the studies and testing, it is possible to define different parameters for the sensors corresponding to different occupants of the vehicle, typically different for the driver and the passengers.

The predetermined alert coefficient ac% is the one that adapts to the particulars of each vehicle and sensor and purpose of the method.

Thus, as depicted in Fig.6 in an embodiment, the full range between the maximum and the minimum value for both the pressure value Vpi and the frequency value Vfi is divided in three portions: two alert intervals-one negative and one positive and the interval between the alert intervals has the meaning of an accepted level of stress. In other embodiment, not represented graphically, there may be only one alert interval if only one of the Pressure Margin PM values and/or Frequency Margin FM values is relevant.

In order to fine-tune the response in step 6, in another preferred embodiment, more than one alert interval threshold and more than one margin interval threshold can be predefined. This corresponds to assigning different degrees of danger to the situation based on how much the measurements exceed the lowest value alert interval threshold and one margin interval threshold. In this case at least two alert interval thresholds are set, and, correspondingly at least two margin thresholds and the vehicle stress rule is discretized in sub-level vehicle stress rules, the number of sub-levels corresponding to the number of successive thresholds for the alert intervals and for the margin intervals.

When defining different degrees of dangers, it is possible to predefine specific different types of action in step 6, for the situations when action is required.

For example, in case two levels the alert interval thresholds the margin interval thresholds are defined: say the lower-value threshold-corresponding to a lower degree of danger and the higher-value threshold, corresponding to a higher degree of danger leading to two sub-levels.

The lower degree of danger triggers in step 6 only an audio or visual signal to the driver, whereas the higher degree of danger triggers in step 6 automatic response from the vehicle such as braking, speed reduction, lane change, etc. Examples of realization Example No. 1: the vehicle is a passenger car having two occupants: the driver 25 and a passenger in the front seat. The pressure sensors Si are located as shown in Fig. 2 for the driver and in Fig. 3 for the front seat passenger.

- The driver's seat has 10 sensors Si in total: sensor weight ki = 1 for the two sensors Si on the steering wheel driver and for the shift gear knob and sensor weight ki= 0.5 for other 7 sensors Si -the value chosen for simplicity.

- The front passenger's seat has also 10 sensors Si in total: sensor weight K = 1 for the two sensors Si on the footrest area and for the door pull handle sensor weight K= 0.5 for other 7 sensors Si -the value chosen for simplicity.

-the predetermined max pressure coefficient pmax% = 50%, - the predetermined max frequency coefficient fmax% = 50%, predetermined monitoring time period T, = 1 minute - Ntotal = 27 Drivers' state is agitated before the trip start.

Due to the driver's state, the preventive driving behaviour is no longer fully followed. The speed of the car starts increasing, the safety distance to the other traffic participants is decreasing, overtaking situations are made in short distances, fast cycles of acceleration-deceleration are occurring.

The driver's interaction with the selected surfaces inside the vehicle in his near proximity is changing rapidly. The gripping force on the drive wheel is increasing, only one hand is on the drive wheel and the other hand is constantly gripping the shift gear knob.

At the same time, the passenger in the front seat starts noticing the potential risk and subconsciously, the passenger's body takes a defensive posture. The passenger is grabbing firmly the door pull handle and is pushing the legs towards the footrest areas. The elbow is no longer laid relaxed on the middle console armrest but rather pressed down with a higher force.

The Pressure Sensor Network PSN starts generating readings outside the nominal values during successive predetermined monitoring time periods Tm as 25 follows: - The gripping force on the steering wheel -where the sensor weight kJ, has maximum value, is increased with more than 50% and only one range of measurements of pressure value Nip; on the steering wheel is generated due to using one hand on the steering wheel. Remember, in Fig. 2 there are two pressure sensors Si on the steering wheel.

- The gripping force on the shift gear knob -where the sensor weight K has maximum value, is increased and constant measurements of pressure value Vpi are generated due to constant hand touching. The increase of the frequency value Vfi of the touch on the knob is indicating that the driver is changing gears faster.

- The pressure sensor Si on the front passenger's door pull handle-where the sensor weight kJ has maximum value starts generating input values. It means that the passenger is touching that area. At the same time the pressure value Vpi of gripping force on the pull handle increases with more than 50% higher than the nominal value.

- The pressure pressure value Vpi of the sensors on the passenger footrest are suddenly increasing with more than 50% above nominal values. Here also the 10 sensor weight kJ has maximum value.

- At a certain moment, within the predetermined monitoring time period Tm = 1minute, the Data Processing Unit DPU receives the following measurements: 3 occurrences for the driver on one of the steering wheel's sensors Si, 2 occurrences for the driver on the shift gear knob, 1 occurrence for the remaining 7 sensors Si, all occurrences places within the respective positive pressure Alert Interval Al and the positive frequency Alert Interval Al, 4 occurrences for the passenger for each of the footrest area's sensors Si, 3 occurrences for the door pull handle, 1 occurrence for the remaining 7 sensors Si all occurrences placed within the respective positive pressure Alert Interval Al and the positive frequency Alert Interval Al, - 2 occurrences for the driver on one of the steering wheel's sensors Si and 1 occurrence of values of pressure value Vpi for the driver on the shift gear knob, 1 occurrence for 3 of the remaining 7 sensors S, when the pressure values Vpi exceeds the Max. Pressure Level MPL value and the frequency values Vfi exceeds the Max. Frequency Level MFL value Vpi.

- 2 occurrences for the passenger for each of the footrest area's sensors Si, 3 occurrences for the door pull handle, 1 occurrence for 2 of the remaining 7 sensors Si when the pressure values Vpi exceeds the Max. Pressure Level MP L value and the frequency values Vfi exceeds the Max.

Frequency Level MFL value Vpi - The Data Processing Unit DPU calculates the two weighted sums: SUMnAl= E?nis, (Aux k = (3+2) X1 + 7X 0.5 for the driver + (4X2 +3)X1 +7X0.5 for the passenger = 24, and SUMnv= Ey Ms, (v) x K = (2+1)X1 + 3X0,5 for the driver + (2X2+3)X1+2X0.5 for the passenger = 12,5 - At this moment the vehicle stress rule is fulfilled because: SUMai + SUMnv > Ntotal 24+ 12,5 > 27 Consequently, the Data Processing Unit DPU sends the alert signal to the Response Control Unit RCU in order to trigger the action.

In this case the alert signal is an audio signal together with the message "POTENTIAL RISK" is displayed on the instrument cluster and on the infotainment unit.

Example No. 2: the vehicle is a passenger car having three occupants: the driver who is a taxi driver and one two passengers in the back seats.

The pressure sensors Si are located as shown in Fig. 2 for the driver and in Fig. 4 for the back seat passengers.

- The driver's seat has 10 sensors Si in total: sensor weight ki = 1 for the two sensors Si on the steering wheel driver and for the shift gear knob and sensor weight kJ= 0.5 for other 7 sensors Si -the value chosen for simplicity.

- The front passenger's seat has also 10 sensors Si in total sensor weight ki = 1 for the two sensors Si on the footrest area and for the door pull handle sensor weight kJ= 0.5 for other 7 sensors Si -the value chosen for simplicity.

- The back passenger's seats have each one of them 8 sensors Si in total, the armrest middle console sensor is common to both seats sensor weight ki = 1 for the two sensors on the footrest area and for the door pull handle sensor weight Kr 0.5 for other 5 sensors -the value chosen for simplicity.

- the predetermined max pressure coefficient pmax% = 50%, - the predetermined max frequency coefficient fmax% = 50%, - predetermined monitoring time period Tm = 1 minute - Ntotal = 27 - Traffic conditions: normal -Driver state: distracted due to phone conversation - Discretized vehicle stress rule: if action is not taken in the immediately subsequent monitoring time period Tm after receipt of the first alert signal, further alert signals are generated with a higher pitch and intensity.

The taxi driver is engaging into a phone conversation shortly after ride start.

The driver is not using the hands free or the connectivity to the car systems.

Consequently, he has only one hand of the steering wheel.

The preventive driving behaviour is no longer fully followed. The distance to brake is shortened. The driver tends to lean towards the side where the phone is handheld. The elbow is firmly pressed down on the armrest. The knee is continuously touching the contact area in the middle console.

At the same time, the passengers in the back seats take a defensive stand.

They are grabbing firmly the door pull handles and they are pushing the knees towards the front seats back cushions.

The Pressure Sensor Network PSN starts generating readings outside the nominal values during successive predetermined monitoring time periods Tm as follows: - There is only only one range of measurements of pressure value Vpi on the steering wheel is generated due to using one hand on the steering wheel - The seat side bolster and side wings sensors read higher pressure value 25 Vpi values and increasing frequency values Vfi.

- Continuous measurements of the pressure value VI); and increasing frequency values Vfi are generated by the knee rest area on the middle console.

- The pressure sensors Si on the passengers' door pull handles are generating measurements of increasing pressure values Vpi and the of the frequency value Vfi of the touch of the door pull handles increases rapidly. It means that the passengers are touching with more force and more frequently that area.

- The sensors Si placed on the seat back cushion are generating measurements of increasing pressure values Vpi and the of the frequency value Vfi of the touch of the door pull handles increases rapidly. It means that the passengers are touching with more force and more frequently that area.

- The computing of the weighted sums is exactly as exemplified in example No. 1. When the vehicle stress rule is fulfilled, there are two levels of action: the alert signal is an audio signal together with the message "POTENTIAL RISK" is displayed on the instrument cluster and on the infotainment unit.

- In case the driver is not taking action in the immediately subsequent monitoring time period Tm = 1 minute, the second level of action is to generate video and audio stimuli and with a higher pitch and intensity during said immediately subsequent monitoring time period Tm = 1 minute.

Example No. 3: the vehicle is a passenger car having four occupants: the driver, the front passenger and two children in the back seats.

The pressure sensors Si are located as shown in Fig. 2 for the driver, Fig. 3 for 15 the front seat passenger and in Fig. 4 for the back seat passengers.

- Traffic conditions: highway driving with low traffic.

- All parameters are set like in Example No.2 - Discretized vehicle stress rule: if action is not taken in the immediately subsequent monitoring time period Tm after receipt of the first alert signal, the 20 vehicle engages automatic braking.

Due to long ride the car participants tend to enter in a more relaxed state. The children in the back seat are close to falling asleep. The passenger in the front seat takes a more relaxed posture lifting the feet from the floor. The driver is also showing signs of relaxation. He has only one hand on the steering wheel, the left-hand elbow is resting on the top side of the door trim.

The Pressure Sensor Network PSN starts generating readings outside the nominal values during successive predetermined monitoring time periods Tm as follows: - There is only one range of measurements of pressure value Vpi on the steering wheel is generated due to using one hand on the steering wheel, the pressure value Vpi is decreasing as the driver is relaxing, until the gripping force falls first within the negative alert interval and then below the Min. Pressure Level MPL value, - The seat side bolster and side wings sensors read higher pressure value \fp; and higher frequency values Vfi and send continuous measurements to the Data Processing Unit DPU.

- Discrete measurements of the pressure value Vpi and increasing frequency 5 values Vfi are generated by the knee rest area on the middle console.

- No remeasurements generated by the sensors placed on the passengers' footrest areas.

- Within the successive predetermined monitoring time period Tm = 1 minute, the Data Processing Unit DPU receives the same type of information as in Example No. 1 but with a major difference: the occurrences refer to the pressure values Vpi and frequency values Vfi placed within the negative alert interval Al for both the pressure Vpi and frequency value Vfi and, respectively below the Min. Pressure Level MPL value and the Min. Frequency Level MFL value.

- When the vehicle stress rule is fulfilled, in this case there are two levels of action: the alert signal is an audio signal together with the message "POTENTIAL RISK" is displayed on the instrument cluster and on the infotainment unit. If action is not taken in the immediately subsequent monitoring time period Tm after receipt of the first alert signal, the vehicle engages automatic braking according to the discretized stress rule.

Example No.4

In another preferred embodiment, the invention is connected to other vehicle's intelligent stress detection system.

For example, the vehicle already has implemented a system for monitoring the eyes of the driver during driving that includes specific rules for detecting an event that requires immediate action.

The Rules Unit will additionally be configured with an interconnecting trigger parameter having two values: YES and NO. The value YES means that the other vehicle's intelligent stress detection system has detected an event, and the value 30 NO means that there is no event.

The event detected by the other vehicle's intelligent stress detection system receives a specific weight W. If the vehicle has more than one other vehicle's intelligent stress detection system, each such system has its specific weight Wi In this case the vehicle stress rule is adapted to the presence of the other vehicle's intelligent stress detection system by using the following formula: (SUMai + SUMnv) X W> Ntotal where W= 1 if the value of the interconnecting trigger is NO and W> 1 if the value of the interconnecting trigger is YES In case the vehicle has more than one other vehicle's intelligent stress detection system, the vehicle stress rule is adapted to the presence of the other vehicle's intelligent stress detection system by using the following formula: (SUMai + SUMnv) X Wi X14/2 K.. X We> Ntotal In this way the invention has the advantage that the method can be interconnected with other stress detection and drive assist solutions.

While certain embodiments of the present invention have been described in detail, those familiar with the art to which this invention relates will recognize various alternative designs and embodiments for practicing the invention as defined by the following claims.

REFERENCE SIGNS

Sensor Si sensor weight ki pressure value Vpi the frequency value Vfi Nominal Pressure Value NPV, Nominal Frequency Value NFV, Yes/No Presence Value YNPV, occurrences MS; when Vpi E pressure Al and Vf; E frequency Al, and occurrences MS; when Vpi exceeds one PM and Vfi exceeds one FM at least one pressure Alert Interval Al between the Nominal Pressure Value NPV and one of the Pressure Margin PM values, at least one frequency Alert Interval Al between the Nominal Frequency Value NFV and one of the Frequency Margin FM values alert interval weighted sum SUMnAi= E7 ms, (Al) x kJ margin interval weighted sum SUMnv= ms, (v) x k;

Claims (7)

1. CLAIMS1. Method of stress detection of at least two vehicle occupants including the driver, characterized in that it comprises the following steps: Si Arranging a plurality of pressure sensors (Si) in a Pressure Sensor Network (PSN), including: positioning each pressure sensor (Si) in the proximity of one vehicle occupant's seat, configuring a predetermined monitoring time period (Tm) defining for each vehicle occupant's seat an occupancy pressure sensor configuring parameters of each pressure sensor (Si): a Nominal Pressure Value (NPV), a Nominal Frequency Value (NFV), a sensor weight (K), configuring an additional parameter for the vehicle occupancy pressure sensor: a Yes/No Presence Value (YNPV), S2 Disconnecting the pressure sensors (Si) corresponding to the vehicle occupant's seats for which the occupancy pressure sensor has the No value, and continuing the steps of the method only for the pressure sensors (Si) corresponding to the vehicle occupant's seats for which the occupancy pressure sensor has a Yes value, S3 Measuring by each pressure sensor (Si) of the pressure value (Vpi) and of the frequency value (Vfi) of the touch exercised by each vehicle occupant on said respective pressure sensor (Si) during the predetermined monitoring time period 25 (Tm) and sending measurements to a Data Processing Unit (DPU), S4 Defining a vehicle stress rule by a Rules Unit (RU): setting for each sensor (Si): at least one Pressure Margin (PM) value being either one or both between a Max. Pressure Level (MPL) value and a Min. Pressure Level (MPL) value, where the Min. Pressure Level (MPL) value < the Nominal Pressure Value (NPV) < the Max. Pressure Level (MPL) value, and at least one Frequency Margin (FM) value being either one or both between a Max Frequency Level (MFL) value and a Min. Frequency Level (MFL) value, where the Min. Frequency Level (MFL) value < Nominal Frequency Value (NFV) < Max. Frequency Level (MFL) value defining alert intervals for each sensor (Si): at least one pressure Alert Interval (Al) between the Nominal Pressure Value (NPV) and one of the Pressure Margin (PM) values, at least one frequency Alert Interval (Al) between the Nominal Frequency 10 Value (NFV) and one of the Frequency Margin (FM) values, defining two parameters referring to occurrences (MSi) for the ensemble of the sensors (Si): an alert interval weighted sum (SUMnAi) = the total number of occurrences (MS;) during the predetermined monitoring time period (Tm) when the pressure values (\fp) are inside one of the at least one pressure Alert Interval (Al) and when the frequency values (W) are inside one of the at least one frequency Alert Interval (Al) weighted by each sensor weight (K), a margin interval weighted sum (SUMnv) = the total number of occurrences (MS;) during the predetermined monitoring time period (Tm) when the pressure values (Vpi) exceed one of the Pressure Margin (PM) values, and the frequency values (n) exceed one of the Frequency Margin (FM) values, weighted by each sensor weight (10, defining an alert interval threshold (Ntotal) = the total number of occurrences (MS;) during the predetermined monitoring time period (Tm) above each of the alert interval weighted sum (SUMnAd) and margin interval weighted sum (SUMnv), defining a vehicle stress rule: the vehicle level of stress for all occupants is high if the sum between the alert interval weighted sum (SUMnAi) and the margin interval weighted sum (SUMnv) is above the alert interval threshold (N total). (SUMai) + (SUMnv) > (N total) sending the vehicle stress rule to the Data Processing Unit (DPU), Checking by the Data Processing Unit (DPU) if the measurements of the occurrences (MS;) received within the predetermined monitoring time period (Trn) fulfil the vehicle stress rule.56 Action by a Response Control Unit (RCU) If the vehicle stress rule is fulfilled, sending by the Data Processing Unit (DPU) of an alert signal to the Response Control Unit (RCU) in order to trigger an action from the driver and/or driver automatic assist systems.If the vehicle stress rule is not fulfilled taking no action.
2. 2. The method of Claim 1 wherein the at least one Alert Interval (Al) is set as follows: Setting a Max. Pressure Level (MPL) value as the Nominal Pressure Value (NPV) increased with a predetermined max pressure coefficient (pmax%), Setting a Min. Pressure Level (MPL) value as the Nominal Pressure Value (NPV) decreased with a predetermined min. pressure coefficient (pm(n%), Setting a Max. Frequency Level (MFL) value as the Nominal Frequency Value (NFV) value increased with a predetermined max frequency coefficient (fmax%), Setting a Min. Frequency Level (MFL) value as the Nominal Frequency Value (NFV) value decreased with a predetermined min frequency coefficient (fmin%), Setting a Positive Pressure Level Alert (PPLA) value as the Nominal Pressure Value (NPV) increased with a predetermined alert coefficient (a0%) Setting a Negative Pressure Level Alert (NPLA) value as the Nominal Pressure Value (NPV) decreased with the predetermined alert coefficient (ac%) Setting a Positive Frequency Level Alert (MFLA) value as the Nominal Frequency Value (NFV) value increased with the predetermined alert coefficient (ac%) Setting a Negative Frequency Level Alert (MFLA) value as the Nominal Frequency Value (NFV) value decreased with the predetermined alert coefficient (ac%) Setting the Alert Interval (Al) as the interval when both the pressure value (Vpn) and the frequency value (Vfn) as follows: a positive alert interval (Al) above the Positive Pressure Level Alert (PPLA), respectively Positive Frequency Level Alert (MFLA) but below the Max. Pressure Level (MPL) value and, respectively, the Max. Frequency Level (MFL) value, or a negative alert interval (Al) above the Min. Pressure Level (MPL) value, respectively the Min. Frequency Level (MFL) value and the Negative Pressure Level Alert (NPLA) value, respectively Negative Frequency Level Alert (MFLA) value.
3. 3. The method of Claim 1 or 2 wherein different degrees of danger are predefined by discretizing the vehicle stress rule in a plurality of sub-level vehicle stress rules.
4. 4. The method of Claim 3 wherein different types of action in step 6 are predefined corresponding to each degree of danger.
5. 5. System of vehicle occupants stress detection comprising: A Pressure Sensor Network (PSN), A Rules Unit (RU), - A Data Processing Unit (DPU), A Response Control Unit (RCU), - A communication network between the Pressure Sensor Network (PSN), the Rules Unit (RU), the Data Processing Unit (DPU) and the Response Control Unit (RCU).characterized in that it is configured to apply the method of stress detection of at least two vehicle occupants including the driver of any of the claims 1 to 4.
6. 6. The system of claim 5 wherein the Rules Unit (RU) and the Data Processing Unit (DPU) are arranged in a single hardware entity, functioning as a Rules and Data Processing Unit (RDPU).
7. 7. The system of claim 5 wherein the Rules Unit (RU) and the Data Processing Unit (DPU) are arranged in different hardware entities.B. A computer program characterized in that it comprises instructions which, when the program is executed by the system of vehicle occupants stress detection of any of the claims 5 to 7, causes the system to carry out the steps of the method of stress detection according to any of the claims 1 to 4.