GB2592425A

GB2592425A - Vehicular control assistance system and method

Info

Publication number: GB2592425A
Application number: GB2002834.6A
Authority: GB
Inventors: Tsubaki Koji
Original assignee: Continental Automotive GmbH
Current assignee: Continental Automotive GmbH
Priority date: 2020-02-28
Filing date: 2020-02-28
Publication date: 2021-09-01
Also published as: GB202002834D0

Abstract

A vehicular control assistance system (100) comprising: a data retriever (128) configured to retrieve data from a data collection system (102) of a vehicle; a historical driving pattern developer (152) configured to develop a driver historical driving pattern of a driver based on historical driving data; an actual driving pattern developer (154) configured to develop an actual driving pattern of the driver based on driver input data comprised in the data retrieved by the data retriever (128) and current trip historical data comprised in the historical driving data; and a driving pattern assessor (142) configured to determine whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion. The system may further comprise a driver evasion travel profile predictor configured to develop an evasion travel profile that the driver is predicted to take to avoid a critical event (i.e. a crash or collision) based on the current and historic driving data.

Description

VEHICULAR CONTROL ASSISTANCE SYSTEM AND METHOD FIELD OF THE INVENTION

The invention relates to a vehicular control assistance system for a motor vehicle, such as a car, and a corresponding vehicular control assistance method.

BACKGROUND

Since many traffic accidents are due to human error during driving, vehicular control assistance systems or driver assistance systems have been designed to prevent accidents or to mitigate the severity of accidents.

Vehicular control assistance systems, which operate under several levels of autonomy, help a driver to control a vehicle. Hence, it is desirable to continually improve on the vehicular control assistance systems in order to continuously reduce the number of and the severity of accidents.

SUMMARY

An objective is to provide an improved vehicular control as-25 sistance system or method in order to reduce the number of and the severity of accidents.

According to a first aspect of the invention, there is provided a vehicular control assistance system comprising: a data re-triever configured to retrieve data from a data collection system of a vehicle; a historical driving pattern developer configured to develop a driver historical driving pattern of a driver based on historical driving data; an actual driving pattern developer configured to develop an actual driving pattern of the driver based on driver input data comprised in the data retrieved by the data retriever and current trip historical data comprised in the historical driving data; and a driving pattern assessor configured to determine whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion.

The vehicular control assistance system is thus able to determine 5 whether the driver is maliciously attempting to create a critical event or is merely incompetent, for example, because the driver is a novice or is ill.

The historical driving pattern developer is configured to develop the driver historical driving pattern of the driver based on the historical driving data. One advantage of the historical driving pattern developer is that the vehicular control assistance system is able to use the information from the historical driving pattern developer, such as the driver historical driving pattern de-veloped, to understand the driver better and to assess whether the driver is, for example, maliciously attempting to create a critical event.

The actual driving pattern developer is configured to develop the actual driving pattern of the driver based on the driver input data comprised in the data retrieved by the data retriever. One advantage of considering the driver input data, is that the actual driving pattern developer may accurately develop the actual driving pattern of a driver by taking into account the driver's current driving inputs or non-driving inputs.

The actual driving pattern developer is configured to develop the actual driving pattern of the driver based on the current trip historical data comprised in the historical driving data. One advantage of considering the current trip historical data comprised in the historical driving data, is that the actual driving pattern developer may accurately develop the actual driving pattern of the driver by taking into account a driver's current driving behaviour for the current trip.

One advantage of the actual driving pattern developer is that the vehicular control assistance system is able to use the information from the actual driving pattern developer, such as the actual driving pattern developed, to understand the driver better and to assess whether the driver is, for example, maliciously attempting to create a critical event.

The driving pattern assessor is configured to determine whether the deviation of the actual driving pattern from the historical driving pattern meets the driving pattern deviation criterion. Thus, the driving pattern assessor is advantageously able to assess whether a driver is maliciously attempting to create a critical event or is merely incompetent, for example, because the driver is a novice or is ill.

Optionally, the driver historical driving pattern and the actual driving pattern comprise a plurality of categories. Optionally, the driving pattern deviation criterion comprises the historical driving pattern falling into a different category from the actual driving pattern.

Optionally, the driver historical driving pattern and the actual driving pattern comprise a scale. Optionally, the driving pattern deviation criterion comprises the deviation of the actual driving pattern from the historical driving pattern by more than a driving pattern percentage threshold value Optionally, the driving pattern deviation criterion comprises: a deviation of an actual acceleration pattern from a historical acceleration pattern of the driver; a deviation of an actual braking pattern from a historical braking pattern of the driver; a deviation of an actual driving speed pattern from a historical driving speed pattern of the driver; or a deviation of an actual lateral acceleration pattern from a historical lateral acceleration pattern of the driver.

Optionally, the vehicular control assistance system further comprises a driver evasion travel profile predictor configured to develop a driver predicted evasion travel profile that the driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever and the historical driving data.

The driver evasion travel profile predictor is configured to develop the driver predicted evasion travel profile based on the driver input data comprised in the data retrieved by the data retriever. One advantage of considering the data from the driver input detector, is that the driver evasion travel profile predictor may accurately predict the driver predicted evasion travel profile by taking into account the driver's current driving inputs or non-driving inputs. Hence, the driver evasion travel profile predictor recognises that a driver in a particular situation may adopt different evasion travel profiles depending on the driver's condition or intention. For instance, a driver intently concentrating on driving may react in a different manner from the same driver attempting to adjust the air-conditioning system while driving.

The driver evasion travel profile predictor is configured to develop the driver predicted evasion travel profile based on the historical driving data. One advantage of considering the historical data, is that the driver evasion travel profile predictor may accurately predict the driver predicted evasion travel profile by taking into account the driver's driving history. Hence, the driver evasion travel profile predictor recognises that different drivers may adopt different evasion travel profiles in a particular situation.

Optionally, the vehicular control assistance system further 30 comprises a predicted travel profile critical event assessor for determining whether the driver predicted evasion travel profile would create the critical event.

Optionally, the vehicular control assistance system further 35 comprises an actual evasion travel profile developer configured to develop an actual evasion travel profile that the driver takes in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever.

The actual evasion travel profile developer is configured to 5 develop the actual evasion travel profile based on the driver input data comprised in the data retrieved by the data retriever. One advantage of considering the driver input data is that the actual evasion travel profile developer may accurately determine an actual evasion travel profile by taking into account the 10 driver's current driving inputs or non-driving inputs.

Optionally, the vehicular control assistance system further comprises an actual travel profile critical event assessor configured to determine whether the actual evasion travel profile 15 would create the critical event.

One advantage of considering the driver input data, is that the actual travel profile critical event assessor may accurately assess whether the actual evasion travel profile would create a critical event by taking into account the driver's current driving inputs or non-driving inputs. Hence, the actual travel profile critical event assessor recognises that a driver in a particular situation may adopt different evasion travel profiles depending on the driver's condition or intention.

Optionally, the vehicular control assistance system further comprises a driver evasion travel profile predictor configured to develop a driver predicted evasion travel profile that the driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever and the historical driving data; an actual evasion travel profile developer configured to develop an actual evasion travel profile that the driver that the driver takes in order to avoid the critical event or to mitigate the effects of the critical event, based on the driver input data; and a travel profile assessor configured to determine whether a deviation of the driver predicted evasion travel profile from the actual evasion travel profile meets a travel profiles deviation criterion.

The actual evasion travel profile developer is configured to develop the actual evasion travel profile based on the driver input data comprised in the data retrieved by the data retriever. One advantage of considering the driver input data is that the actual evasion travel profile developer may accurately determine an actual evasion travel profile by taking into account the driver's current driving inputs or non-driving inputs.

Optionally, the travel profile deviation criterion comprises: a deviation of an actual speed of the vehicle from a predicted speed of the vehicle; a deviation of an actual acceleration of the vehicle from a predicted acceleration of the vehicle; a deviation of an actual position of the vehicle from a predicted position of the vehicle; or a deviation of an actual steering angle of a steering wheel of the vehicle from a predicted steering angle of the steering wheel of the vehicle.

Optionally, the travel profile deviation criterion comprises: a total period of time that an actual speed of the vehicle deviates from a predicted speed by above a first speed deviation threshold value over a first predetermined period of time is above a total speed deviation time threshold value; a number of times that the actual speed of the vehicle deviates from the predicted speed of the vehicle by above a second speed deviation threshold value over a second predetermined period of time is above a first threshold value; an actual vehicle position deviates from a predicted vehicle position by a position threshold value; a number of times that an actual steering angle of a steering wheel of the vehicle deviates from a predicted steering angle of the steering wheel of the vehicle by a steering angle threshold value over a third predetermined period of time is above a second threshold value; or an integration of a steering angle deviation-time graph over a fourth predetermined period of time is above an integrated steering angle deviation threshold value.

Optionally, the vehicular control assistance system further comprises a driving strategy developer for developing an evasion strategy comprising manoeuvres for avoiding the critical event or to mitigate the effects of the critical event, based on the data retrieved by the data retriever.

One advantage of the driving strategy developer is that it is able to develop at least one evasion strategy to prevent the critical event from occurring or to mitigate the effects of the critical 35 event.

Optionally, the driving strategy developer is configured to develop more than one evasion strategies.

One advantage of the driving strategy developer is that it is able to develop more than one evasion strategies, and the vehicular control assistance system may then decide which evasion strategy is suitable to be implemented in each different situation or which evasion strategy is the best evasion strategy to be implemented in a particular situation.

Optionally, the vehicular control assistance system further comprises a critical event assessor configured to, based on the data retrieved by the data retriever: determine whether a critical event would occur; calculate an estimated time period before the critical event would occur; calculate a comfort evasion time period threshold value that indicates a minimum time period necessary for the vehicle to avoid the critical event without requiring aggressive manoeuvres; and determine whether the estimated time period before the critical event would occur is higher or lower than the comfort evasion time period threshold value.

Optionally, the vehicular control assistance system further comprises a sleeping driver assessor configured to determine whether the driver has fallen asleep based on the driver input data comprised in the data retrieved by the data retriever.

One advantage of considering the driver input data, is that the sleeping driver assessor may accurately and robustly determine whether the driver is asleep, unconscious or sleepy. For instance, a driver who is asleep would not change gear, apply the brakes or adjust the driver seat. Thus, a driver condition detector may detect that the driver's eyes are open and the driver's head is straight, but the driver may in fact be daydreaming and not paying any attention to the road.

Optionally, the vehicular control assistance system further 35 comprises a driving assistant configured to control the vehicle at a plurality of levels of autonomy.

Optionally, a motor vehicle comprising the vehicular control assistance system.

According to a second aspect of the invention, there is provided a vehicular control assistance system comprising: a data retriever configured to retrieve data from a data collection system of a vehicle; a historical driving pattern developer configured to develop a driver historical driving pattern of a driver based on historical driving data; an actual driving pattern developer configured to develop an actual driving pattern of the driver based on driver input data and current trip historical data comprised in the historical driving data; a driving pattern assessor configured to determine whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion; a driver evasion travel profile predictor configured to develop a driver predicted evasion travel profile that the driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever and the historical driving data; a predicted travel profile critical event assessor for determining whether the driver predicted evasion travel profile would create the critical event; an actual evasion travel profile developer configured to develop an actual evasion travel profile that the driver takes in order to avoid the critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever; an actual travel profile critical event assessor configured to determine whether the actual evasion travel profile would create the critical event; a travel profile assessor configured to determine whether a deviation of the driver predicted evasion travel profile from the actual evasion travel profile meets a travel profiles deviation criterion; wherein the travel profile deviation criterion comprises: a deviation of an actual speed of the vehicle from a predicted speed of the vehicle; a deviation of an actual acceleration of the vehicle from a predicted acceleration of the vehicle; a deviation of an actual position of the vehicle from a predicted position of the vehicle; a deviation of an actual steering angle of a steering wheel of the vehicle from a predicted steering angle of the steering wheel of the vehicle; a total period of time that an actual speed of the vehicle deviates from a predicted speed by above a first speed deviation threshold value over a first predetermined period of time is above a total speed deviation time threshold value; a number of times that the actual speed of the vehicle deviates from the predicted speed of the vehicle by above a second speed deviation threshold value over a second predetermined period of time is above a first threshold value; an actual vehicle position deviates from a predicted vehicle position by a position threshold value; a number of times that an actual steering angle of the steering wheel of the vehicle deviates from a predicted steering angle of the steering wheel of the vehicle by a steering angle threshold value over a third predetermined period of time is above a second threshold value; or an integration of a steering angle deviation-time graph over a fourth predetermined period of time is above an integrated steering angle deviation threshold value; a driving strategy developer for developing an evasion strategy comprising manoeuvres for avoiding the critical event or to mitigate the effects of the critical event, based on the data retrieved by the data retriever; a critical event assessor configured to, based on the data retrieved by the data retriever: determine whether the critical event would occur; calculate an estimated time period before the critical event would occur; calculate a comfort evasion time period threshold value that indicates a minimum time period necessary for the vehicle to avoid the critical event without requiring aggressive manoeuvres; and determine whether the estimated time period before the critical event would occur is higher or lower than the comfort evasion time period threshold value; a sleeping driver assessor configured to determine whether the driver has fallen asleep based on the driver input data comprised in the data retrieved by the data retriever; and a driving assistant configured to control the vehicle at a plurality of levels of autonomy.

The vehicular control assistance system is thus able to determine whether the driver is maliciously attempting to create a critical event or is merely incompetent, for example, because the driver is a novice or is ill.

The historical driving pattern developer is configured to develop 5 the driver historical driving pattern of the driver based on the historical driving data. One advantage of the historical driving pattern developer is that the vehicular control assistance system is able to use the information from the historical driving pattern developer, such as the driver historical driving pattern de-10 veloped, to understand the driver better and to assess whether the driver is, for example, maliciously attempting to create a critical event.

One advantage of considering the driver input data, is that the 5 actual travel profile critical event assessor may accurately assess whether the actual evasion travel profile would create a critical event by taking into account the driver's current driving inputs or non-driving inputs. Hence, the actual travel profile critical event assessor recognises that a driver in a 10 particular situation may adopt different evasion travel profiles depending on the driver's condition or intention.

One advantage of the driving strategy developer is that it is able 10 to develop at least one evasion strategy to prevent the critical event from occurring or to mitigate the effects of the critical event.

According to a third aspect of the Invention, there is provided a vehicular control assistance method for a vehicle comprising the steps of: retrieving data from a data collection system of the vehicle; developing a driver historical driving pattern of a driver based on historical driving data; developing an actual driving pattern of the driver based on driver input data comprised in the data retrieved from the data collection system and current trip historical data comprised in the historical driving data; and determining whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion.

The vehicular control assistance method is thus able to determine 35 whether the driver is maliciously attempting to create a critical event or is merely incompetent, for example, because the driver is a novice or is ill.

BRIEF DESCRIPTION OF THE DRAWINGS

These and other features, aspects, and advantages will become better understood with regard to the following description, 5 appended claims, and accompanying drawings where: Figure 1 shows a vehicular control assistance system; Figure 2 shows a diagram for a vehicular control assistance method using the vehicular control assistance system of Figure 1; Figure 3 Illustrates the process steps involved when a critical event assessor determines whether a critical event may occur and whether an estimated time period before the critical event would occur is higher or lower than a comfort evasion time period threshold value; and Figure 4 illustrates the process steps Involved when a sleeping driver assessor determines whether the driver has fallen asleep.

In the drawings, like parts are denoted by like reference numerals.

DESCRIPTION

In the summary above, in this description, in the claims below, and in the accompanying drawings, reference is made to particular features (including method steps) of the invention. It is to be understood that the disclosure of the invention in this specification includes all possible combinations of such particular features. For example, where a particular feature is disclosed in the context of a particular aspect or embodiment of the invention, or a particular claim, that feature can also be used, to the extent possible, in com-bination with and/or in the context of other particular aspects and embodiments of the invention, and in the inventions generally.

Theterm"comprises"andgrammaticaleguivalentsthereofareused herein to mean that other components, ingredients, steps, et cetera are optionally present. For example, an article "comprising" (or "which comprises") components A, B, and C can consist of (that is, contain only) components A, B, and C, or can contain not only components A B, and C but also one or more other components.

Where reference is made herein to a method comprising two or more defined steps, the defined steps can be carried out in any order or simultaneously (except where the context excludes that possibility), and the method can include one or more other steps which are carried out before any of the defined steps, between two of the defined steps, or after all the defined steps (except where the context excludes that possibility).

The term "at least" followed by a number is used in to denote the start of a range beginning with that number (which may be a range having an upper limit or no upper limit, depending on the variable being defined). For example, "at least 1" means 1 or more than 1. The term "at most" followed by a number is used herein to denote the end of a range ending with that number (which maybe a range having 1 or 0 as its lower limit, or a range having no lower limit, depending on the variable being defined). For example, "at most 4" means 4 or less than 4, and "at most 40%" means 40% or less than 40%. When, in this specification, a range is given as "(a first number) to (a second number)" or "(a first number) -(a second number)", this means a range whose lower limit is the first number and whose upper limit is the second number. For example, 25 to 100 mm means a range whose lower limit is 25 mm, and whose upper limit is 100 mm.

Figure 1 shows a vehicular control assistance system 100 comprising a data collection system 102, a decision-making system 104 and a response system 106.

The data collection system 102 is configured to capture data required by the vehicular control assistance system 100 and to transmit the data to the decision-making system 104. The data collection system 102 may comprise at least one of the following modules: an environment detector 110, a vehicle detector 112, a driver condition detector 114, a driver input detector 116, a map module 118 and a localisation module 120.

The environment detector 110 comprises sensors to detect the surroundings of a vehicle. The environment detector 110 may comprise sensors such as lidar (light detection and ranging) sensors, radar (radio detection and ranging) sensors, ultrasound sensors, cameras (including infrared cameras), or combinations thereof. The environment detector 110 is configured to capture data from the vehicle's environment and to send the data to the decision-making system 104.

The vehicle detector 112 comprises sensors to collect data about the vehicle. The vehicle detector 112 may comprise sensors configured to collect vehicular dynamics information on at least one of a steer angle, a wheel rotation speed, the vehicle's speed, the vehicle's acceleration (linear or lateral), the vehicle's rotation rate on its longitudinal axis, transverse axis or yaw axis, a chassis height, a cabin temperature, an engine, a suspension system, or a transmission system. The vehicle detector 112 is configured to capture the vehicular dynamics data and to transmit the data to the decision-making system 104.

The driver condition detector 114 comprises sensors to collect data about a driver's condition. The driver condition detector 114 may comprise at least one of a camera, for example, positioned in front of the driver to capture images of the head and the upper body of the driver, or a sensor, for example, located in a driver seat to detect the posture of the driver. For example, images of the driver's head tilt angle and the driver's eyes are captured. The driver condition detector 114 is configured to send the driver data, such as images of the driver, to the decision-making system 104.

The driver input detector 116 comprises sensors to collect data about the driver's operating input to the vehicle, which may comprise a driving input or a non-driving input. The driving input comprises an input that directly affects a movement of the vehicle on the road, for example, how much the driver presses on or releases an accelerator, how much the driver presses on or releases a brake pedal (even if the vehicle is stationary and the intention is to prevent the vehicle from moving), the turning of a steering wheel or a change of gear. The non-driving input comprises an input that does not directly affect a movement of the vehicle on the road, for example, an adjustment of the driver seat, the opening or closing of a storage compartment, an adjustment of a rear-view mirror or a wing mirror, an input to an information panel, an adjustment of an air-conditioning system, an adjustment of a sunroof or an adjustment of a wiper system. The driver input detector 116 is configured to transmit the driver's operating input data, including driving input data and non-driving input data, to the decision-making system 104.

The map module 118 may provide static map information and dynamic map information to the decision-making system 104. The map module 118 may comprise a static digital map stored in a memory device or a network, such as an intranet, an extranet or the internet. The static digital map may include information on physical objects such as buildings, roads, traffic lights, road signs and petrol stations. The map module 118 may regularly update the static digital map through a network, such as an intranet, an extranet or the internet. The map module 118 may update the static digital map by using a mobile network, such as the second generation mobile network, the third generation mobile network, the fourth generation mobile network or the fifth generation mobile network. Furthermore, the map module 118 may receive the dynamic map information from the internet, vehicle-to-environment (V-to-X) communication devices or the environment detector 110 of the vehicle. The dynamic map information may include information on roadworks, accidents and the weather. The map module 118 may collate the static map information and the dynamic map information to construct a real-world model, which is then transmitted to the decision-making system 104.

The localisation module 120 may comprise a satellite navigation system, for example, a global positioning system (GPS) based satellite navigation system. The GPS based satellite navigation system comprises a GPS receiver for receiving location and time information from satellites. The localisation module 120 may cooperate with the map module 118 to allow the vehicle to localise itself. The localisation module 120 is configured to send localisation data to the decision-making system 104.

The decision-making system 104 is configured to receive the data or information from the data collection system 102. The decision-making system 104 is then configured to analyse and to interpret the data and finally to draw certain conclusions. The decision-making system 104 may comprise at least one of the following modules: a data retriever 128, a critical event assessor 130, a driving strategy developer 132, a sleeping driver assessor 134, a driver evasion travel profile predictor 136, a travel profile assessor 138, a dangerous driver assessor 140, a driving pattern assessor 142, a predicted travel profile critical event assessor 146, an actual travel profile critical event assessor 148, an actual evasion travel profile developer 150, a historical driving pattern developer 152, an actual driving pattern developer 154 and a history storage module 144.

The data retriever 128 is configured to retrieve the data from the data collection system 102, so that the various modules comprised in the decision-making system 104 could analyse and interpret the data. The data retriever 128 may comprise a processor, a memory device and software programs for retrieving the data or information from the various modules in the data collection system 102. Each module in the decision-making system 104 may use a different subset of the data obtained from the data collection system 102.

The history storage module 144 is configured to store all the available historical data. The history storage module 144 may comprise a memory device, such as a hard disk, or devices configured to access the cloud. The history storage module 144 stores the data retrieved by the data retriever 128 from the data collection system 102. The history storage module 144 may store all data and information from the modules comprised in the decision-making system 104. The historical data stored in the history storage module 144 may, for example, be used by the driver evasion travel profile predictor 136 to predict a driver predicted evasion travel profile or by the historical driving pattern developer 152 to develop a driver historical driving pattern.

The critical event assessor 130 is configured to determine whether a critical event may occur. A critical event is a situation in which an animate object (for example, a driver of a vehicle, a passenger, another driver of another vehicle, a pedestrian, an animal, a tree or other living being) or an inanimate object (for example, a vehicle, another vehicle, a building, a lamp post or other non-living physical object) is endangered due to the direct action or the inaction of a vehicle or a driver. A critical event may result from the action of a vehicle or a driver, for example, a vehicle colliding with another vehicle travelling in front of the vehicle, a pedestrian or a wall. A critical event may also result from the inaction of a vehicle or a driver, for example, a driver of a vehicle stops the vehicle on a railway, and thus causes a train to collide with the stationary vehicle. In fact, a critical event is not limited to collision events. A critical event may, for instance, be a vehicle driving off a cliff.

A driver may cause a critical event for a variety of reasons. There are some possible non-malicious reasons why a driver may create a critical event; the driver may have poor driving skill, the driver may perform a sudden reflex action or the driver may be in a bad condition, for instance, the driver may be tired or sleepy, the driver may even have fallen asleep or lost con-sciousness at the wheel, the driver may have taken drowsy medication or drugs, or the driver may be inattentive or distracted. In addition, there are possible malicious reasons why a driver may cause a critical event to occur. The driver may want to hurt himself, for instance, in an attempt to commit suicide or to incur an injury in a fit of anger. The driver may also want to destroy a physical object or to hurt another person. For instance, the driver may attempt to crash the vehicle to hurt a passenger, to damage another vehicle or to intimidate a pedestrian.

Therefore, the critical event assessor 130 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102, and to determine whether a critical event may occur. The critical event assessor 130 may analyse the data from the map module 118 and the localisation module 120 to localise the vehicle. The critical event assessor 130 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle, and analyse the data from the vehicle detector 112 to determine whether there is possibility of a collision or other critical event occurring based on the vehicle's current travel profile, which includes factors such as speed, acceleration, position, trajectory and rotation rates. The critical event assessor 130 may then conclude whether a critical event is likely to occur. The critical event assessor 130 is also configured to calculate an estimated time period before the critical event would occur and a comfort evasion time period threshold value, from the data obtained from the environment detector 110, the vehicle detector 112, the map module 118 and the localisation module 120. The comfort evasion time period threshold value is a minimum time period necessary for the vehicle to avoid the potential critical event without requiring aggressive manoeuvres. The critical event assessor 130 is further configured to determine whether the estimated time period before the critical event would occur is higher or lower than the comfort evasion time period threshold value. The critical event assessor 130 may then send the information to the driving strategy developer 132 or the response system 106.

For instance, as a vehicle turns a sharp corner, although the map module 118 shows no obstacles after the sharp corner, the environment detector 110 detects a pedestrian crossing the road metres in front of the vehicle's current position indicated by the localisation module 120. The critical event assessor 130 analyses the vehicle's current travel profile, including factors such as trajectory, speed, acceleration, rotation rates on its longitudinal axis, transverse axis and yaw axes, to determine whether the vehicle would collide with the pedestrian, and to calculate an estimated time period before the collision with the pedestrian would occur and a comfort evasion time period threshold value. The critical event assessor 130 then determines whether the estimated time period before the critical event would occur is higher or lower than the comfort evasion time period threshold value. If the critical event assessor 130 determines that the vehicle might collide with the pedestrian, the critical event assessor 130 may send the information to the driving strategy developer 132 or the response system 106.

The driving strategy developer 132 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102. The driving strategy developer 132 may also receive information from the critical event assessor 130 on whether a critical event would occur and if so, whether an estimated time period before the critical event would occur is higher or lower than a comfort evasion time period threshold value. The driving strategy developer 132 is configured to analyse the data from the environment detector 110, the vehicle detector 112, the map module 118 and the localisation module 120. The driving strategy developer 132 may analyse the data from the map module 118 and the localisation module 120 to determine various possible evasive options available to prevent the critical event from occurring or to mitigate the effects of the critical event. For example, the driving strategy developer 132 may consider whether there is a neighbouring lane into which the vehicle may possibly move, or whether there is a pavement that the vehicle may mount. The driving strategy developer 132 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle, for example, whether there is another vehicle in the neighbouring lane that would collide with the vehicle should the vehicle swerve into the neighbouring lane, whether there is a pedestrian or a lamppost on the pavement, or whether the road is sandy or wet. In addition, the driving strategy developer 132 may analyse the data from the vehicle detector 112 to determine whether the vehicle is able to take each of the 5 possible evasive options based on the vehicle's current travel profile, which includes factors such as speed, acceleration, position, trajectory and rotation rates. For example, the driving strategy developer 132 may determine that the vehicle is taking a corner too fast to safely swerve into the neighbouring lane, 10 or that the vehicle is travelling too fast for the vehicle to stop on the sandy or wet road.

After analysing the available data, the driving strategy developer 132 is configured to develop an evasion strategy. An evasion strategy is a strategy that a driving assistant 162 comprised in the response system 106 could implement in order to prevent a critical event from occurring or to mitigate the effects of the critical event. An evasion strategy may include an evasion route trajectory, a speed profile plan, an acceleration profile plan, a steering wheel turning profile plan, a gear change plan, a brake activation profile plan or combinations thereof.

The driving strategy developer 132 may develop an emergency evasion strategy or a comfort evasion strategy. An emergency evasion strategy may involve aggressive manoeuvres such as turning the steering wheel drastically, swerving the vehicle suddenly, jamming on the brakes or combinations thereof. A comfort evasion strategy may involve gentle manoeuvres such as turning the steering wheel gradually, changing the vehicle's direction gradually, applying the brakes gradually or combinations thereof. The driving strategy developer 132 may then send the information on the evasion strategies developed to the response system 106, for the driving assistant 162 to implement.

One advantage of the driving strategy developer 132 is that it is able to develop at least one evasion strategy to prevent a critical event from occurring or to mitigate the effects of the critical event. Another advantage of the driving strategy developer 132 is that it is able to develop more than one evasion strategies, and the decision-making system may then decide which evasion strategy is suitable to be implemented in each different situation or which evasion strategy is the best evasion strategy to be implemented in a particular situation.

For instance, the driving strategy developer 132 receives information from the critical event assessor 130 that the vehicle may collide with a temporary structure 5 metres in front of the vehicle in about 3 seconds. The driving strategy developer 132 analyses the data from the environment detector 110, the map module 118 and the localisation module 120 to determine, for example, whether there is any neighbouring lane, whether the neighbouring lane has space for the vehicle to move into, whether the vehicle would collide into any other vehicle if it were to swerve into the neighbouring lane, or whether the ahead road is wet or sandy. The driving strategy developer 132 also analyses the data from the vehicle detector 112 to determine whether the vehicle is travelling too fast to safely stop before the temporary structure if it were to brake. If the driving strategy developer 132 determines that it is unable to avoid a collision, whether with the temporary structure, with another vehicle, or with a guard rail, the driving strategy developer 132 may choose the least harmful strategy, for example, to crash the vehicle into the guard rail and to stop the vehicle on an empty pavement. After developing the evasion strategy to mitigate the effects of this critical event, the driving strategy developer 132 then sends the information on the evasion strategies developed to the response system 106, for the driving assistant 162 to implement.

The sleeping driver assessor 139 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102 in order to determine whether the driver has fallen asleep. The sleeping driver assessor 134 is configured to analyse the data from the driver condition detector 114 and the driver input detector 116. The sleeping driver assessor 134 may analyse the images from the driver condition detector 114 to determine, for example, whether the driver's eyes are closed or whether the driver's head is tilted. The sleeping driver assessor 134 may analyse the data from the driver input detector 116 to determine whether the driver has provided any input, such as turning the steering wheel, changing gear or applying the brakes. The sleeping driver assessor 134 may also determine whether there is any change in the driver's input, for example, a change in the pressure applied on the accelerator or a change in the speed of the wipers.

One advantage of considering the data from the driver input detector 116, is that the sleeping driver assessor 134 may accurately and robustly determine whether a driver is asleep, unconscious or sleepy. For instance, a driver who is asleep would not change gear, apply the brakes or adjust the driver seat. Thus, the driver condition detector 114 may detect that the driver's eyes are open and the driver's head is straight, but the driver may in fact be daydreaming and not paying any attention to the road.

The sleeping driver assessor 134 may measure how much time has elapsed since the last input or change in input, and continuously compare the elapsed time with a first sleeping driver determination time threshold value. If the driver provides an input or a change in input before the elapsed time reaches the first sleeping driver determination time threshold value (for example, 15 seconds), then the sleeping driver assessor 134 may set a sleeping driver flag to negative. If the driver does not provide an input or a change in input before the elapsed time reaches the first sleeping driver determination time threshold value, the sleeping driver assessor 134 may continue measuring how much time has elapsed since the last input or change in input. Once the elapsed time reaches a second sleeping driver determination time threshold value (for example, 60 seconds) , the sleeping driver assessor 134 may set the sleeping driver flag to positive and send the positive sleeping driver flag information to the response system 106. However, even if the driver provides an input or a change in input before the elapsed time reaches the second sleeping driver determination time threshold value, the sleeping driver assessor 134 may increase a sleeping driver counter by 1. If the sleeping driver counter reaches a sleeping driver counter threshold value (for example, 5) within a predetermined time period (for example, 120 seconds), the sleeping driver assessor 134 may set the sleeping driver flag to positive and send the positive sleeping driver flag information to the response system 106.

One advantage of using the sleeping driver counter is that the 10 sleeping driver assessor 134 may detect whether a driver is gradually falling asleep or losing consciousness, because the sleepy driver would gradually provide fewer and fewer inputs.

The driver evasion travel profile predictor 136 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102 and the historical data stored in the history storage module 144, to predict a driver predicted evasion travel profile. The driver evasion travel profile predictor 136 is configured to analyse the data from the environment detector 110, the vehicle detector 112, the driver input detector 116, the map module 118, the localisation module 120 and the history storage module 144, to predict the driver predicted evasion travel profile. The driver evasion travel profile predictor 136 may analyse the data from the map module 118 and the localisation module 120 to localise the vehicle. The driver evasion travel profile predictor 136 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle. The driver evasion travel profile predictor 136 may further analyse the data from the vehicle detector 112 to determine the current state of the vehicle. Additionally, the driver evasion travel profile predictor 136 may analyse the data from the driver input detector 116, for example, any driving input or non-driving input, to determine the driver's intention or current condition.

One advantage of considering the data from the driver input detector 116, is that the driverevasiontravelprofilepredictor 136 may accurately predict the driver predicted evasion travel profile by taking into account the driver's current driving inputs or non-driving inputs. Hence, the driver evasion travel profile predictor 136 recognises that a driver in a particular situation may adopt different evasion travel profiles depending on the driver's condition or intention. For instance, a driver intently concentrating on driving may react in a different manner from the same driver attempting to adjust the air-conditioning system while driving.

In addition, the driver evasion travel profile predictor 136 may analyse historical data about a driver from the history storage module 144, to predict a driver predicted evasion travel profile. The driver evasion travel profile predictor 136 may analyse the historical data, such as the driver's historical driving data about the driver's previous trips. The historical data may include data about the driver's driving habits, how the driver drove in similar situations (for example, whether the driver braked or swerved), or whether the driver managed to prevent a critical event from occurring in similar situations. The driver evasion travel profile predictor 136 may also analyse historical data from at least one of the modules in the data collection system 102, such as the environment detector 110, the vehicle detector 112, the driver condition detector 114, the driver input detector 116, the map module 118 or the localisation module 120. The driver evasion travel profile predictor 136 may further analyse historical data from at least one of the modules in the decision-making system 104, such as the data retriever 128, the critical event assessor 130, the driving strategy developer 132, the sleeping driver assessor 134, the driver evasion travel profile predictor 136, the travel profile assessor 138, the dangerous driver assessor 140, the driving pattern assessor 142, the predicted travel profile critical event assessor 146, the actual travel profile critical event assessor 148, the actual evasion travel profile developer 150, the historical driving pattern developer 152 or the actual driving pattern developer 154. Additionally, the driver evasion travel profile predictor 136 may analyse historical data from the response system 106.

One advantage of considering the historical data from the history storage module 144, is that the driver evasion travel profile predictor 136 may accurately predict a driver predicted evasion travel profile by taking into account the driver's driving history. Hence, the driver evasion travel profile predictor 136 recognises that different drivers may adopt different evasion travel profiles in a particular situation.

The driver evasion travel profile predictor 136 may then combine the data from the environment detector 110, the vehicle detector 112, the driver input detector 116, the map module 118 and the localisation module 120 with the historical data from the history storage module 144 to predict the driver predicted evasion travel profile.

A driver predicted evasion travel profile is a travel profile that a driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event. A driver predicted evasion travel profile may include factors such as speed, acceleration, position, trajectory and rotation rates, that a driver is predicted to adopt in order to evade a critical event based on the current input data and the historical driving data about that driver. A driver predicted evasion travel profile is to be distinguished from a route that a driver plans to take from a stating location to a destination in order to avoid a traffic jam or an accident.

The driver evasion travel profile predictor 136 may use artificial intelligence, for example, an artificial neural network, to predict a driver predicted evasion travel profile. The driver evasion travel profile predictor 136 may input the data from the environment detector 110, the vehicle detector 112, the driver input detector 116, the map module 118, the localisation module 120 and the history storage module 144 into the artificial neural network to construct a driver predicted evasion travel profile.

The driver evasion travel profile predictor 136 may further analyse the data from the driver input detector 116 together with the historical data from the history storage module 144, to determine whether the manner that a driver is driving during the current trip is substantially different from the manner that the driver drove historically in previous trips. For instance, although historically a driver drives economically, but is driving aggressively for the current trip. In that case, the driver evasion travel profile predictor 136 may advantageously give the data about the current trip precedence over the data about the previous trips, when the driver evasion travel profile predictor 136 develops a driver predicted evasion travel profile.

The predicted travel profile critical event assessor 146 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102 and the data from the driver evasion travel profile predictor 136. The predicted travel profile critical event assessor 146 is configured to analyse the data from the environment detector 110, the map module 118, the localisation module 120 and the driver evasion travel profile predictor 136. The predicted travel profile critical event assessor 146 may analyse the data from the map module 118 and the localisation module 120 to localise the vehicle. The predicted travel profile critical event assessor 146 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle. The predicted travel profile critical event assessor 146 may then analyse the driver predicted evasion travel profile in the light of the data from the environment detector 110, the map module 118 and The localisation module 120, to determine whether a driver predicted evasion travel profile would create a critical event. If the driver predicted evasion travel profile would create a critical event, the predicted travel profile critical event assessor 146 may send this information to the response system 106.

The actual evasion travel profile developer 150 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102. The actual evasion travel profile de-veloper 150 is configured to analyse the data from the environment detector 110, the vehicle detector 112, the map module 118, the localisation module 120 and the driver input detector 116. The actual evasion travel profile developer 150 may analyse the data from the map module 118 and the localisation module 120 to localise the vehicle. The actual evasion travel profile developer 150 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle. The actual evasion travel profile developer 150 may then analyse the data from the vehicle detector 112, the map module 118, the localisation module 120 and the driver input detector 116 to determine an actual evasion travel profile, which includes factors such as speed, acceleration, position, trajectory and rotation rates, of the driver.

One advantage of considering the data from the driver input detector 116, is that the actual evasion travel profile developer 150 may accurately determine an actual evasion travel profile by taking into account the driver's current driving inputs or non-driving inputs.

The travel profile assessor 138 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102, the data from the driver evasion travel profile predictor 136 and the data from the actual evasion travel profile developer 150. The travel profile assessor 138 is configured to compare a driver predicted evasion travel profile obtained from the driver evasion travel profile predictor 136 with an actual evasion travel profile obtained from the actual evasion travel profile developer 150, in order to determine how much the actual evasion travel profile deviates from the driver predicted evasion travel profile. The travel profile assessor 138 may consider factors such as speed, acceleration, position, trajectory, rotation rates or combinations thereof. The travel profile assessor 138 may determine whether the deviation of the actual evasion travel profile from the driver predicted evasion travel profile meets a travel profiles deviation criterion.

The travel profile assessor 138 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102 and the data from the driver evasion travel profile predictor 136. The travel profile assessor 138 is configured to analyse the data from the environment detector 110, the vehicle detector 112, the map module 118, the localisation module 120, the driver input detector 116 and the driver evasion travel profile predictor 136. The travel profile assessor 138 may analyse the data from the map module 118 and the localisation module 120 to localise the vehicle. The travel profile assessor 138 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle. Thereafter, the travel profile assessor 138 may determine how much the actual evasion travel profile deviates from the driver predicted evasion travel profile. The travel profile assessor 138 may consider factors such as speed, acceleration, position, trajectory, rotation rates or combinations thereof. The travel profile assessor 138 may determine whether the deviation of the actual evasion travel profile from the driver predicted evasion travel profile meets a travel profiles deviation criterion. The travel profile assessor 138 may determine whether the actual evasion travel profile deviates from the driver predicted evasion travel profile by more than one travel profiles deviation criteria.

A travel profiles deviation criterion may comprise a deviation of an actual speed of the vehicle from a predicted speed of the vehicle, a deviation of an actual acceleration of the vehicle from a predicted acceleration of the vehicle, a deviation of an actual position of the vehicle from a predicted position of the vehicle, or a deviation of an actual steering angle of the steering wheel of the vehicle from a predicted steering angle of the steering wheel of the vehicle. For example, the travel profiles deviation criterion may be that an actual evasion travel profile speed is 20 km/h higher than a driver predicted evasion travel profile speed, or that an actual evasion travel profile acceleration is 10 m/s2 higher than a driver predicted evasion travel profile acceleration.

A travel profiles deviation criterion may also comprise a total period of time that the actual speed of the vehicle deviates from the predicted speed by above a speed deviation threshold value over a predetermined period of time is above a total speed deviation time threshold value, a number of times that the actual speed of the vehicle deviates from the predicted speed of the vehicle by above another speed deviation threshold value over a predetermined period of time is above a threshold value, an actual vehicle position deviates from a predicted vehicle position by a position threshold value, a number of times that an actual steering angle of the steering wheel of the vehicle deviates from a predicted steering angle of the steering wheel of the vehicle by a steering angle threshold value over a predetermined period of time is above another threshold value, or an integration of a steering angle deviation-time graph over a predetermined period of time is above an integrated steering angle deviation threshold value.

If the travel profile assessor 138 determines that a deviation of an actual evasion travel profile from a driver predicted evasion travel profile meets a travel profiles deviation criterion or travel profiles deviation criteria, the travel profile assessor 138 may send its conclusion to the actual travel profile critical event assessor 148, the dangerous driver assessor 140, the driving pattern assessor 142 or the response system 106.

The actual travel profile critical event assessor 148 is configured to receive information from the travel profile assessor 138 on whether a deviation of an actual evasion travel profile from a driver predicted evasion travel profile meets a travel profiles deviation criterion or travel profiles deviation criteria. The actual travel profile critical event assessor 148 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102, and to determine whether a critical event may occur. The actual travel profile critical event assessor 148 is configured to analyse the data from the environment detector 110, the vehicle detector 112, the driver input detector 116, the map module 118 and the localisation module 120. The actual travel profile critical event assessor 148 may analyse the data from the map module 118 and the localisation module 120 to localise the vehicle. The actual travel profile critical event assessor 148 may also analyse the data from the environment detector 110 to detect objects and the environment around the vehicle, and analyse the data from the vehicle detector 112 to determine whether there is possibility of a collision or other critical event occurring based on the vehicle's current travel profile. The actual travel profile critical event assessor 148 may further analyse the data from the driver input detector 116, for example, any driving input or non-driving input, to determine the driver's intention or current condition. The actual travel profile critical event assessor 148 may then analyse the actual evasion travel profile in the light of the data from the environment detector 110, the vehicle detector 112, the driver input detector 116, the map module 118 and the localisation module 120, to determine whether the actual evasion travel profile would create a critical event. If the actual evasion travel profile would create a critical event, the actual travel profile critical event assessor 148 may send its conclusion to the dangerous driver assessor 140, the driving pattern assessor 142 or the response system 106.

One advantage of considering the data from the driver input detector 116, is that the actual travel profile critical event assessor 148 may accurately assess whether an actual evasion travel profile would create a critical event by taking into account the driver's current driving inputs or non-driving inputs. Hence, the actual travel profile critical event assessor 148 recognises that a driver in a particular situation may adopt different evasion travel profiles depending on the driver's condition or intention.

The dangerous driver assessor 140 is configured to analyse the dataretrievedbythedataretriever128 from the data collection system 102. The dangerous driver assessor 140 may also receive information from the actual travel profile critical event assessor 148 on whether an actual evasion travel profile would create a critical event, or the travel profile assessor 138 on whether a deviation of an actual evasion travel profile from a driver predicted evasion travel profile meets a travel profiles deviation criterion or travel profiles deviation criteria. The dangerous driver assessor 140 is configured to analyse the data from the driver condition detector 114 and the driver input detector 116. The dangerous driver assessor 140 may analyse the images from the driver condition detector 114 to determine, for example, whether the driver's eyes are closed or whether the driver's head is tilted. The dangerous driver assessor 140 may also analyse the data from the driver input detector 116 to determine whether the driver has provided any input, such as turning the steering wheel, changing gear or applying the brakes. The dangerous driver assessor 140 may also determine whether there is any change in the driver's input, for example, a change in the pressure applied on the accelerator or a change in the speed of the wipers.

The dangerous driver assessor 140 may assess whether the driver is alert, for example, by assessing whether the driver's eyes are open and whether the driver is providing any input. If the driver is not alert, for example, because the driver is drowsy, the dangerous driver assessor 140 may set an unaware driver flag to positive and send the positive unaware driver flag information to the response system 106. If the driver is alert, the dangerous driver assessor 140 may assess whether the driver is distracted, for example, whether the driver is adjusting the air-conditioning system or whether the driver's eyes are looking at the road. If the driver is alert yet distracted, the dangerous driver assessor 140 may set a distracted driver flag to positive and send the positive distracted driver flag information to the response system 106. If the driver is alert and not distracted, for example, the driver is looking intently at a pedestrian and aggressively accelerating towards the pedestrian, then the dangerous driver assessor 140 may set a dangerous driver flag to positive and send the positive dangerous driver flag information to the response system 106 or the driving pattern assessor 142.

One advantage of considering the data from the driver input 5 detector 116, is that the dangerous driver assessor 140 may accurately and robustly determine whether the driver is alert, distracted or dangerous.

The historical driving pattern developer 152 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102 and the historical data stored in the history storage module 144, to develop a driver historical driving pattern. The historical driving pattern developer 152 may analyse a driver's historical driving data, such as historical data about the driver's previous trips, which include the driver's past driving habits, the driver's past driving pattern (for example, aggressive, calm or economical), or how risk-averse was the driver. The historical driving pattern developer 152 may also analyse the historical data from at least one of the modules in the data collection system 102, such as the environment detector 110, the vehicle detector 112, the driver condition detector 114, the driver input detector 116, the map module 118 or the localisation module 120. The historical driving pattern developer 152 may further analyse the historical data from at least one of the modules in the decision-making system 104, such as the data retriever 128, the critical event assessor 130, the driving strategy developer 132, the sleeping driver assessor 134, the driver evasion travel profile predictor 136, the travel profile assessor 138, the dangerous driver assessor 140, the driving pattern assessor 142, the predicted travel profile critical event assessor 146, the actual travel profile critical event assessor 148, the actual evasion travel profile developer 150, the historical driving pattern developer 152 or the actual driving pattern developer 154. Additionally, the historical driving pattern developer 152 may analyse the historical data from the response system 106.

The historical driving pattern developer 152 may analyse the historical data from the history storage module 144 to develop a driver historical driving pattern, which may, for example, be categorised as aggressive, calm or economical. The driver historical driving pattern may comprise a plurality of categories (for example, aggressive, calm or economical), a scale (for example, expressed as a percentage), or a combination thereof. The driver historical driving pattern may also comprise a historical manoeuvring pattern, such as a historical accel-eration pattern, a historical braking pattern, a historical driving speed pattern or a historical lateral acceleration pattern. The historical driving pattern developer 152 may use artificial intelligence, for example, an artificial neural network, to develop a driver historical driving pattern. The historical driving pattern developer 152 may input the historical data from the history storage module 144 into the artificial neural network to construct a driver historical driving pattern.

One advantage of the historical driving pattern developer 152 is that the vehicular control assistance system 100 is able to use the information from the historical driving pattern developer 152, such as a driver historical driving pattern developed, to understand a driver better and to assess whether the driver is, for example, maliciously attempting to create a critical event.

The actual driving pattern developer 154 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102 and the historical data stored in the history storage module 144, to develop an actual driving pattern of a driver. The actual driving pattern developer 154 is configured to analyse the data from the driver input detector 116 and the history storage module 144, to develop the actual driving pattern of the driver. The actual driving pattern developer 154 is configured to analyse the data from the driver input detector 116, for example, whether the driver is aggressively accelerating towards an object.

One advantage of considering the data from the driver input detector 116, is that the actual driving pattern developer 154 may accurately develop an actual driving pattern of a driver by taking into account the driver's current driving inputs or non-driving inputs.

The actual driving pattern developer 154 may analyse historical data about a driver's current trip from the history storage module 144, to develop an actual driving pattern of the driver. The actual driving pattern developer 154 may analyse the historical data about the driver's current trip, for example the driver's current driving behaviour, the driver's current driving pattern (for example, aggressive, calm or economical), orhowrisk-averse is the driver. The actual driving pattern developer 154 may also analyse the historical data from at least one of the modules in the data collection system 102, such as the environment detector 110, the vehicle detector 112, the driver condition detector 114, the driver input detector 116, the map module 118 or the localisation module 120. The actual driving pattern developer 154 may further analyse the historical data from at least one of the modules in the decision-making system 104, such as the data retriever 128, the critical event assessor 130, the driving strategy developer 132, the sleeping driver assessor 134, the driver evasion travel profile predictor 136, the travel profile assessor 138, the dangerous driver assessor 140, the driving pattern assessor 142, the predicted travel profile critical event assessor 146, the actual travel profile critical event assessor 148, the actual evasion travel profile developer 150, the historical driving pattern developer 152 or the actual driving pattern developer 154. Additionally, the actual driving pattern developer 154 may analyse the historical data from the response system 106.

One advantage of considering the historical data from the history storage module 144, is that the actual driving pattern developer 154 may accurately develop an actual driving pattern of the driver by taking into account the driver' s current driving behaviour for the current trip.

The actual driving pattern developer 154 may then combine the data from the driver input detector 116 with the historical data from the history storage module 144 to develop an actual driving pattern of the driver. The actual driving pattern may comprise a plurality of categories (for example, aggressive, calm or economical), a scale (for example, expressed as a percentage), or a combination thereof. The actual driving pattern may also comprise an actual manoeuvring pattern, such as an actual acceleration pattern, an actual braking pattern, an actual driving speed pattern or an actual lateral acceleration pattern. The actual driving pattern developer 154 may use artificial intelligence, for example, an artificial neural network, to develop the actual driving pattern. The historical driving pattern developer 152 may input the historical data from the history storage module 144 into the artificial neural network to construct an actual driving pattern.

One advantage of the actual driving pattern developer 154 is that the vehicular control assistance system 100 is able to use the information from the actual driving pattern developer 154, such as an actual driving pattern developed, to understand a driver better and to assess whether the driver is, for example, maliciously attempting to create a critical event.

The driving pattern assessor 142 is configured to analyse the data retrieved by the data retriever 128 from the data collection system 102, information from the historical driving pattern developer 152 on a driver historical driving pattern developed, and information from the actual driving pattern developer 154 on a driver historical driving pattern developed. The driving pattern assessor 142 may also receive information from the travel profile assessor 138 on whether a deviation of an actual evasion travel profile from a driver predicted evasion travel profile meets a travel profiles deviation criterion or travel profiles deviation criteria, information from the actual travel profile critical event assessor 148 on whether the actual evasion travel profile would create a critical event, or information from the dangerous driver assessor 140 on whether a dangerous driver flag is positive.

The driving pattern assessor 142 is configured to compare a driver historical driving pattern developed by the historical driving pattern developer 152 with an actual driving pattern developed by the actual driving pattern developer 154, in order to determine how much the driver historical driving pattern deviates from the actual driving pattern. The driving pattern assessor 142 may determine whether a deviation of the driver historical driving pattern from the actual driving pattern meets a driving pattern deviation criterion. A driving pattern deviation criterion may comprise the driver historical driving pattern falling into a different category from the actual driving pattern, or a de-viation of the driver historical driving pattern from the actual driving pattern by a predetermined driving pattern percentage threshold value. A driving pattern deviation criterion may also comprise a deviation of an actual acceleration pattern from a historical acceleration pattern, a deviation of an actual braking pattern from a historical braking pattern, a deviation of an actual driving speed pattern from a historical driving speed pattern, or a deviation of an actual lateral acceleration pattern from a historical lateral acceleration pattern. The driving pattern assessor 142 may determine whether the driver historical driving pattern deviates from the actual driving pattern by more than one driving pattern deviation criteria.

If the driving pattern assessor 142 determines that a deviation of a driver historical driving pattern from an actual driving pattern meets a driving pattern deviation criterion, the driving pattern assessor 142 may set a malicious driver flag to positive and send the positive malicious driver flag information to the response system 106. However, if the driving pattern assessor 142 determines that the deviation of the driver historical driving pattern from the actual driving pattern does not meet the driving pattern deviation criterion, the driving pattern assessor 142 may set an incompetent driver flag to positive and send the positive incompetent driver flag information to the response system 106.

For example, the driving pattern assessor 142 may receive information from the travel profile assessor 138 that an actual evasion travel profile deviates from a driver predicted evasion travel profile by a travel profiles deviation criterion. The driving pattern assessor:42 may also receive information from the actual travel profile critical event assessor 148 that the actual evasion travel profile would create a critical event by colliding with a stationary car 10 metres ahead of the vehicle.

The driving pattern assessor 142 may further receive information from the dangerous driver assessor 140 that its driver is alert, not distracted, and dangerous. The driving pattern assessor 142 then compares a driver historical driving pattern, which is determined by the historical driving pattern developer 152 to be calm, with a driver historical driving pattern, which is determined by the actual driving pattern developer 154 to be aggressive. Thus, the driving pattern assessor 142 may conclude that its driver is maliciously attempting to create a critical event and send this conclusion to the response system 106.

One advantage of the driving pattern assessor 142 is that it is able to assess whether a driver is maliciously attempting to create a critical event or is merely incompetent, for example, because the driver is a novice or is ill.

The response system 106 is configured to receive data or information from the decision-making system 104, and may perform an action in the light of the data or information to prevent a critical event from occurring or to mitigate the effects of the critical event. The response system 106 may comprise the driving assistant 162. The driving assistant 162 may control the vehicle at different levels of autonomy. The driving assistant 162 may produce a warning signal to warn the driver at a lowest level of autonomy, autonomously produce, increase, reduce or maintain a driving input to compensate a driver's input at an intermediate level of autonomy, or fully take overall steering, accelerating and braking maneuverers at a highest level of autonomy. The driving assistant 162 may produce at least one of an acoustic signal, a visual signal, a haptic signal or an olfactory signal in order to warn the driver at the lowest level of autonomy. The driving assistant 162 may produce, increase, reduce or maintain at least one of a steering maneuverer, an accelerating maneuverer 5 or a braking maneuverer in order to compensate the driver's input at the intermediate level of autonomy. The driving assistant 162 may take overall steering, accelerating and braking maneuverers to, for example, perform an emergency braking maneuverer or implement an evasion strategy at the highest level of autonomy. 10 The driving assistant 162 may receive the sleeping driver flag information from the sleeping driver assessor 134, the unaware driver flag information from the dangerous driver assessor 140, the distracted driver flag information from the dangerous driver assessor 140, the dangerous driver flag information from the dangerous driver assessor 140, the malicious driver flag information from the driving pattern assessor 142, and the incompetent driver flag information from the driving pattern assessor 142. The driving assistant 162 may perform an action if at least one of the flags is positive. The driving assistant 162 may produce a warning signal to warn the driver at a lowest level of autonomy, autonomously produce, increase, reduce or maintain a driving input to compensate a driver's input at an intermediate level of autonomy, or fully take overall steering, accelerating and braking maneuverers at a highest level of autonomy, if at least one of the flags is positive.

Figure 2 shows a diagram for a vehicular control assistance method 200 using the vehicular control assistance system 100.

At step 202, the vehicular control assistance method 200 starts by initialising the vehicular control assistance system 100. At step 204, the data retriever 128 retrieves the data from the data collection system 102 and the history storage module 144, so that the various modules comprised in the decision-making system 104 could analyse and interpret the data.

Then at step 206, the critical event assessor 130 determines whether a critical event may occur and determines whether an estimated time period before the critical event would occur is higher or lower than a comfort evasion time period threshold value. Figure 3 illustrates the process steps involved in step 206, when the critical event assessor 130 determines whether a critical event may occur and whether an estimated time period before the critical event would occur is higher or lower than a comfort evasion time period threshold value. At step 302, the critical event assessor 130 is initialised, and then the critical event assessor 130 analyses the data, at step 304. At step 306, the critical event assessor 130 determines whether a critical event may occur. If no critical event would occur, the step 206 process ends at step 312, and the vehicular control assistance method 200 proceeds to the next step 208. If a critical event may occur, the critical event assessor 130 proceeds to determine whether an estimated time period before the critical event would occur is higher or lower than a comfort evasion time period threshold value, at step 308. Next, at step 310, the critical event assessor 130 sends the information to the driving strategy developer 132 or the response system 106. Thereafter, the step 206 process ends at step 312, and the vehicular control assistance method 200 proceeds to the next step 208. At step 208, the driving strategy developer 132 develops at least one evasion strategy, for example, an emergency evasion strategy or a comfort evasion strategy.

At step 210, the sleeping driver assessor 134 determines whether the driver has fallen asleep. Figure 4 illustrates the process steps involved in step 210, when the sleeping driver assessor 134 determines whether the driver has fallen asleep. At step 402, the sleeping driver assessor 134 is initialised, and then the sleeping driver assessor 134 analyses the data, at step 404. At step 406, the sleeping driver assessor 134 measures how much time has elapsed since the driver last provided an input and determines whether the elapsed time is more than a first sleeping driver determination time threshold value. If the elapsed time is less than the first sleeping driver determination time threshold value, the step 210 process ends at step 416, and the vehicular control assistance method 200 proceeds to the next step 212. If the elapsed time is more than the first sleeping driver determination time threshold value, the sleeping driver assessor 134 determines whether the elapsed time reaches a second sleeping driver determination time threshold value, at step 408. If the elapsed time reaches the second sleeping driver determination time threshold value, the sleeping driver assessor 134 sets a sleeping driver flag to positive and sends the positive sleeping driver flag information to the response system 106, at step 410. Then, the step 210 process proceeds from step 410 to end at step 416, and the vehicular control assistance method 200 proceeds to the next step 212. Ifatstep40B, the elapsed time does not reach the second sleeping driver determination time threshold value, the sleeping driver assessor 134 increases a sleeping driver counter by 1, at step 412. At step 414, the sleeping driver assessor 134 determines whether the sleeping driver counter reaches a sleeping driver counter threshold value within a predetermined time period. If the sleeping driver counter does not reach the sleeping driver counter threshold value within the predetermined time period, the step 210 process proceeds from step 414 to end at step 416, and the vehicular control assistance method 200 proceeds to the next step 212. If the sleeping driver counter reaches the sleeping driver counter threshold value within the predetermined time period, the sleeping driver assessor 134 sets the sleeping driver flag to positive and sends the positive sleeping driver flag information to the response system 106, at step 410. Thereafter, the step 210 process proceeds from step 410 to end at step 416, and the vehicular control assistance method 200 proceeds to the next step 212.

At step 212, the driver evasion travel profile predictor 136 develops a driver predicted evasion travel profile. At step 214, the predicted travel profile critical event assessor 146 de-termines whether the driver predicted evasion travel profile would create a critical event. At step 216, the actual evasion travel profile developer 150 determines an actual evasion travel profile. At step 218, the travel profile assessor 138 determines whether a deviation of the actual evasion travel profile from the driver predicted evasion travel profile meets a travel profiles deviation criterion. At step 220, the actual travel profile critical event assessor 148 determines whether the actual evasion travel profile would create a critical event.

At step 222, the dangerous driver assessor 140 determines whether the driver is alert, distracted or dangerous. At step 224, the historical driving pattern developer 152 constructs a driver historical driving pattern. At step 226, the actual driving pattern developer 154 develops an actual driving pattern. At step 228, the driving pattern assessor 142 determines whether the deviation of the driver historical driving pattern from the actual driving pattern meets a driving pattern deviation criterion.

Subsequently, at step 230, the response system 106 performs an action to prevent a critical event from occurring or to mitigate the effects of the critical event. Finally, the vehicular control assistance method 200 ends at step 232. However, the vehicular control assistance method 200 may continuously loop from step 232 back to step 202.

Although the vehicular control assistance method 200 is il-lustrated in a sequential plurality of steps in Figure 2, it is to be understood that the vehicular control assistance method 200 has been described thus merely to aid understanding and the vehicular control assistance method 200 is not necessarily limited to the described order of steps. In fact, it is to be appreciated that each step in the vehicular control assistance method 200 is continuously repeated, and thus a subsequently described step maybe performed before an earlier described step.

Thus, in accordance with the vehicular control assistance method 200, it is now possible prevent a critical event from occurring or to mitigate the effects of the critical event, even if, for example, a driver is maliciously attempting to create the critical event or the driver is incompetent.

Although the invention has been described in considerable detail with reference to certain embodiments, other embodiments are possible.

For example, the data collection system 102, the decision-making system 104 and the response system 106 may comprise more or fewer modules than described.

Therefore, the spirit and scope of the appended claims should not be limited to the description of the embodiments contained herein.

All features disclosed in this specification (including the appended claims, abstract, and accompanying drawings) may be replaced by alternative features serving the same, equivalent, or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features.

Claims

PATENT CLAIMSWhat is claimed is: 1. A vehicular control assistance system (100) comprising: a data retriever (128) configured to retrieve data from a data collection system (102) of a vehicle; a historical driving pattern developer (152) configured to develop a driver historical driving pattern of a driver based on 10 historical driving data; an actual driving pattern developer (154) configured to develop an actual driving pattern of the driver based on driver input data comprised in the data retrieved by the data retriever (128) and current trip historical data comprised in the his-torical driving data; and a driving pattern assessor (142) configured to determine whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion.
2. The vehicular control assistance system (100) of claim 1, wherein the driver historical driving pattern and the actual driving pattern comprise a plurality of categories.
3. The vehicular control assistance system (100) as in anyone of the preceding claims, wherein the driving pattern deviation criterion comprises the historical driving pattern falling into a different category from the actual driving pattern.
4. The vehicular control assistance system (100) of claim 1, wherein the driver historical driving pattern and the actual driving pattern comprise a scale.
5. The vehicular control assistance system (100) as in claim 1 or claim 4, wherein the driving pattern deviation criterion comprises the deviation of the actual driving pattern from the historical driving pattern by more than a driving pattern percentage threshold value.
6. The vehicular control assistance system (100) as in anyone of the preceding claims, wherein the driving pattern deviation criterion comprises: a deviation of an actual acceleration pattern from a historical acceleration pattern of the driver; a deviation of an actual braking pattern from a historical braking pattern of the driver; a deviation of an actual driving speed pattern from a 10 historical driving speed pattern of the driver; or a deviation of an actual lateral acceleration pattern from a historical lateral acceleration pattern of the driver.
7. The vehicular control assistance system (100) as in anyone of the preceding claims, further comprising a driver evasion travel profile predictor (136) configured to develop a driver predicted evasion travel profile that the driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever (128) and the historical driving data.
8. The vehicular control assistance system (100) of claim 7, further comprising a predicted travel profile critical event 25 assessor (146) for determining whether the driver predicted evasion travel profile would create the critical event.
9. The vehicular control assistance system (100) as in anyone of the preceding claims, further comprising an actual evasion travel profile developer (150) configured to develop an actual evasion travel profile that the driver takes in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever (128).
10. The vehicular control assistance system (100) of claim 9, further comprising an actual travel profile critical event assessor (148) configured to determine whether the actual evasion travel profile would create the critical event.
11. The vehicular control assistance system (100) as in anyone 5 of claims 1-6, further comprising: a driver evasion travel profile predictor (136) configured to develop a driver predicted evasion travel profile that the driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever (128) and the historical driving data; an actual evasion travel profile developer (150) configured to develop an actual evasion travel profile that the driver that the driver takes in order to avoid the critical event or to mitigate the effects of the critical event, based on the driver input data; and a travel profile assessor (138) configured to determine whether a deviation of the driver predicted evasion travel profile from the actual evasion travel profile meets a travel 20 profiles deviation criterion.
12. The vehicular control assistance system (100) of claim 11, wherein the travel profile deviation criterion comprises: a deviation of an actual speed of the vehicle from a 25 predicted speed of the vehicle; a deviation of an actual acceleration of the vehicle from a predicted acceleration of the vehicle; a deviation of an actual position of the vehicle from a predicted position of the vehicle; or a deviation of an actual steering angle of a steering wheel of the vehicle from a predicted steering angle of the steering wheel of the vehicle.
13. The vehicular control assistance system (100) as in anyone 35 of claims 11-12, wherein the travel profile deviation criterion comprises: a total period of time that an actual speed of the vehicle deviates from a predicted speed by above a first speed deviation threshold value over a first predetermined period of time is above a total speed deviation time threshold value; a number of times that the actual speed of the vehicle deviates from the predicted speed of the vehicle by above a second 5 speed deviation threshold value over a second predetermined period of time is above a first threshold value; an actual vehicle position deviates from a predicted vehicle position by a position threshold value; a number of times that an actual steering angle of a steering wheel of the vehicle deviates from a predicted steering angle of the steering wheel of the vehicle by a steering angle threshold value over a third predetermined period of time is above a second threshold value; or an integration of a steering angle deviation-time graph over 15 a fourth predetermined period of time is above an integrated steering angle deviation threshold value.
14. The vehicular control assistance system (100) as in anyone of the preceding claims, further comprLsing a driving strategy developer (132) for developing an evasion strategy comprising manoeuvres for avoiding a critical event or to mitigate the effects of the critical event, based on the data retrieved by the data retriever (128).
15. The vehicular control assistance system (100) as in anyone of the preceding claims, further comprising a critical event assessor (130) configured to, based on the data retrieved by the data retriever (128): determine whether a critical event would occur; calculate an estimated time period before the critical event calculate a comfort evasion time period threshold value that indicates a minimum time period necessary for the vehicle to avoid the critical event without requiring aggressive manoeuvres; and determine whether the estimated time period before the critical event would occur is higher or lower than the comfort evasion time period threshold value.
16. The vehicular control assistance system (100) as in anyone of the preceding claims, further comprising a sleeping driver assessor (134) configured to determine whether the driver has fallen asleep based on the driver input data comprised in the data retrieved by the data retriever (128).
17. The vehicular control assistance system (100) as in anyone of the preceding claims, further comprising a driving assistant (162) configured to control the vehicle at a plurality of levels 10 of autonomy.
18. Amotor vehicle comprising the vehicular control assistance system (100) as in any one of the preceding claims.
19. A vehicular control assistance system (100) comprising: a data retriever (128) configured to retrieve data from a data collection system (102) of a vehicle; a historical driving pattern developer (152) configured to develop a driver historical driving pattern of a driver based on 20 historical driving data; an actual driving pattern developer (154) configured to develop an actual driving pattern of the driver based on driver input data and current trip historical data comprised in the historical driving data; a driving pattern assessor (142) configured to determine whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion; a driver evasion travel profile predictor (136) configured to develop a driver predicted evasion travel profile that the driver is predicted to take in order to avoid a critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever (128) and the historical driving data; a predicted travel profile critical event assessor (146) for determining whether the driver predicted evasion travel profile would create the critical event; an actual evasion travel profile developer (150) configured to develop an actual evasion travel profile that the driver takes in order to avoid the critical event or to mitigate the effects of the critical event, based on the driver input data comprised in the data retrieved by the data retriever (128); an actual travel profile critical event assessor (148) configured to determine whether the actual evasion travel profile would create the critical event; a travel profile assessor (138) configured to determine 10 whether a deviation of the driver predicted evasion travel profile from the actual evasion travel profile meets a travel profiles deviation criterion; wherein the travel profile deviation criterion comprises: a deviation of an actual speed of the vehicle from a predicted speed of the vehicle; a deviation of an actual acceleration of the vehicle from a predicted acceleration of the vehicle; a deviation of an actual position of the vehicle from a predicted position of the vehicle; a deviation of an actual steering angle of a steering wheel of the vehicle from a predicted steering angle of the steering wheel of the vehicle; a total period of time that an actual speed of the vehicle deviates from a predicted speed by above a first speed deviation threshold value over a first predetermined period of time is above a total speed deviation time threshold value; a number of times that the actual speed of the vehicle deviates from the predicted speed of the vehicle by above a second speed deviation threshold value over a second predetermined period of time is above a first threshold value; an actual vehicle position deviates from a predicted vehicle position by a position threshold value; a number of times that an actual steering angle of the steering wheel of the vehicle deviates from a predicted steering angle of the steering wheel of the vehicle by a steering angle threshold value over a third predetermined period of time is above a second threshold value; or an integration of a steering angle deviation-time graph over a fourth predetermined period of time is above an integrated steering angle deviation threshold value; a driving strategy developer (132) fordevelopinganevasion strategy comprising manoeuvres for avoiding the critical event or to mitigate the effects of the critical event, based on the data retrieved by the data retriever (128); a critical event assessor (130) configured to, based on the data retrieved by the data retriever (128): determine whether the critical event would occur; calculate an estimated time period before the critical event would occur; calculate a comfort evasion time period threshold value that indicates a minimum time period necessary for the vehicle to avoid the critical event without re quiring aggressive manoeuvres; and determine whether the estimated time period before the critical event would occur is higher or lower than the comfort evasion time period threshold value; a sleeping driver assessor (134) configured to determine whether the driver has fallen asleep based on the driver input data comprised in the data retrieved by the data retriever (128); 25 and a driving assistant (162) configured to control the vehicle at a plurality of levels of autonomy.
20. A vehicular control assistance method (200) for a vehicle 30 comprising the steps of: retrieving data from a data collection system (102) of the vehicle; developing a driver historical driving pattern of a driver based on historical driving data; developing an actual driving pattern of the driver based on driver input data comprised in the data retrieved from the data collection system (102) and current trip historical data comprised in the historical driving data; and determining whether a deviation of the actual driving pattern from the historical driving pattern meets a driving pattern deviation criterion.