WO2016005382A1

WO2016005382A1 - End-of-journey vehicle systems

Info

Publication number: WO2016005382A1
Application number: PCT/EP2015/065465
Authority: WO
Inventors: Adam GRZYWACZEWSKI; Joan BARCELO LLADO
Original assignee: Jaguar Land Rover Limited
Priority date: 2014-07-08
Filing date: 2015-07-07
Publication date: 2016-01-14
Also published as: GB2528064B; EP3166806A1; GB201412109D0; GB2528064A

Abstract

An end-of-journey (EOJ) system (100) for use in a vehicle comprises a collection means (50, 52) to receive, and collect into time limited sets, measurements of vehicle parameters (42-48) during a journey of a vehicle in which the system is fitted. Comparison means (56) compare said sets with recorded sets of comparable parameters that are indicative of an imminent EOJ event. Prediction means (56) issue a prediction of an EOJ event and a level of confidence with which a predicted EOJ event will occur in a time frame or distance travelled. In response to said prediction, and the level of confidence accompanying said prediction, initiation means (60) initiate an EOJ mode of one or more vehicle systems. A vehicle incorporating such a system comprises one or more vehicle systems that have an end-of-journey (EOJ) mode of operation. The EOJ predictor (56) includes a learning algorithm that records parameters that lead to an EOJ event for future predictions.

Description

End-of-Journey Vehicle Systems

TECHNICAL FIELD

This invention relates to vehicle systems and their shut down or modification at the end of a journey.

BACKGROUND

When a journey in a vehicle terminates and a driver (and any passenger(s)) is to leave the vehicle, various systems of the vehicle enter a shutdown or other end-of-journey (EOJ) mode. An obvious instance of this is stopping the engine. Some vehicles have systems that detect that the vehicle has come to a halt, and stop the engine immediately, but in this instance, the vehicle remains in a condition of readiness to restart the engine should the driver depress the accelerator, or remove his or her foot from the brake pedal. This is an energy saving mode "in-journey", with which the present invention is not concerned. However, it would be desirable for some systems to enter an EOJ mode in-journey (i.e. before the journey ends). For example, vehicles, when making parking manoeuvres, would frequently benefit from surround lighting to facilitate observation of obstacles to be avoided. Parking occurs, of course, only at the end of a journey. Switches may be provided to actuate these as and when needed. Likewise, vehicles that have elevated suspension systems (for example, many off-road vehicles have suspension that is elevated to provide ground clearance) also often have low ride-height modes, which makes entrance to and egress from a vehicle easier to achieve for drivers and passengers. However, air suspension vehicles often take several seconds to settle into low ride-height mode when a vehicle is "switched off".

It is an object of the present invention to provide a vehicle with improved end of journey systems, or at least to provide a system that addresses the above issues and mitigates some of the problems discussed. SUMMARY OF THE INVENTION

In accordance with the present invention there is provided a vehicle comprising:

one or more vehicle systems that have an end-of-journey (EOJ) mode of operation; an EOJ predictor, adapted to output an in-journey prediction of an EOJ event occurring within a specific time frame or distance of travel; and an EOJ controller which, when the EOJ predictor predicts the occurrence of an EOJ event, initiates said EOJ mode of one or more of said vehicle systems.

In current automotive systems, there are no means of predicting that the driver is near to finishing his/her journey unless he/she has introduced manually his/her destination to a navigation system. However, for many journeys, the driver does not introduce the destination into the navigation system. Hence current automotive systems cannot estimate when the driver is going to finish his/her journey in advance. Knowledge that an end-of-journey (EOJ) event is about to happen can be exploited in order to substantially enhance the driving experience during the last leg of the journey by automating several actions related to the end-of-journey process. For example, as mentioned above, when the driver is engaged with parking manoeuvres, he or she would frequently benefit from automatic activation of i) front, rear and side-cameras (if present in the vehicle) and ii), surround lighting to facilitate observation of obstacles.

An EOJ event may be defined as occurring when a driver operates a driver's door handle of the vehicle, or when the engine is switched off, followed by the door handle being operated. However, other means of defining an EOJ event are possible. For example, a vehicle ignition key being turned to an off-position and/or removed from its socket (if it has one) may be employed alone as confirmation of an EOJ event. For proximity keys, the EOJ event may be determined when the driver (assuming he or she has the proximity key) leaves the vicinity of the vehicle, so that the vehicle no longer detects its presence (although this is invariably preceded by a door-opening event). A number of different events may be employed to confirm the EOJ event, and the present invention is not limited to a particular parameter for defining it.

The EOJ mode of operation depends on the system in question:

• If the vehicle is equipped with an air suspension system, an EOJ mode may comprise a lowering of the suspension. If this is effected in the moments before a vehicle comes to a complete halt, egress from the vehicle may be facilitated. Indeed, in luxury vehicles, the vehicle settling onto its lowest suspension setting as it arrives at a destination could be described as "elegant arrival", which may be a desirable result provided by the present invention.

· To facilitate parking and final positioning of a vehicle at the end of a journey, the lighting system of the vehicle may have an EOJ mode, where lights are arranged to be switched on during hours of darkness and provide all round illumination of the immediate surroundings of the vehicle. Whether there is a detector to detect low levels of background illumination (i.e. darkness) or whether the system relies on the condition of the vehicle driving lights when the EOJ event approaches, or whether the relevant lights are switched on in the EOJ mode in any event, is not relevant.

· More prosaically, there may be vehicle systems that have an EOJ mode of less immediate and obvious benefit to the driver. For example, hybrid vehicles may have an electric engine (EE) for low speed operation and powered by a suitable energy storage device (for example batteries) and an internal combustion engine (ICE) that drives the vehicle at higher speeds and/or charges the energy storage device. An EOJ mode of operation for such hybrid vehicles may involve deliberately deselecting operation of the ICE, in order to exploit the charge on the EE energy storage device before the EOJ. This saves fuel because the ICE is not used and the EE energy storage device will lose charge in any event if the vehicle is unused for a significant length of time after the EOJ event. In addition, even this mode may have ergonomic benefits for the driver, also contributing to the "elegant arrival" motif: that is, gliding silently and without fumes to a final parking position. That might be at a location close to home where sleeping occupants may appreciate not being disturbed by an operating ICE, for example.

• Similarly, heating, ventilation and air conditioning (HVAC) systems can be shut down before the end of a journey, without any appreciable loss of comfort for the vehicle occupants, and yet saving energy by not being, or not remaining, implemented as a vehicle approaches its EOJ event.

• Other EOJ modes will exist and/or be developed.

In order to trigger the features above mentioned (and many more), the vehicle system is provided with a EOJ predictor algorithm. Being able to predict the end of the journey is a very valuable piece of information within the automotive context. Knowing that a driver is about to finish the journey allows the vehicle system to prepare for the event and activate a number of convenient features automatically. Predicting when the journey is about to finish, in sufficient advance of the finish, is a non-trivial problem. Moreover, the earlier the prediction is required, the more complex the problem becomes. One of the approaches for understanding that the journey is close to its end is the parking manoeuvre detection. Hence, detecting that the user is currently engaged in a parking process would indicate with high level of confidence about the driver intention to finish the journey.

This disclosure details a system that is capable of learning how individuals park at certain destinations, by observing characteristic indicators and providing the in-vehicle computer system with a prediction of whether the driver is currently engaged in a parking manoeuvre and, therefore, finishing the journey.

Predicting an EOJ event is not trivial. A vehicle simply coming to a halt is not a safe predictor (despite being a necessary precursor of an EOJ event). In any event, coming to a halt is often too late to be of use. Nevertheless, there are a number of parameters that can be employed and in one embodiment the EOJ predictor is provided with one or a combination of the following vehicle parameters or contextual data on the basis of which to predict an EOJ event:

• Speed: Speed is a clear indicator of the intention of the driver to park. The parking manoeuvres typically occur at low speeds (<1 OKm/h).

• Acceleration and braking: Parking manoeuvres follow certain patterns on acceleration and breaking. Detecting those patterns can be used as clear indicators of a parking manoeuvre event.

• Steering wheel angle: The sequence of movements of the steering wheel during the parking typically differs from the movements during normal driving. An example can be observed in Figure 1 (see below).

• Seatbelt unbuckling (driver and passengers). The unbuckling event can be used as an indicator that the user is getting ready to leave the car.

• Gear mode: The parking event is typically preceded for a certain changes of gears, forward and reverse. Also, higher gears will discard parking events.

• Directional Indicators: These indicators should anticipate some certain manoeuvres, such as parking or stopping events.

• Others: such as pedal pressure, direction indicators, inside and outside cameras and radar.

In addition GPS and map data may be employed.

• Current x,y coordinates: Knowing the current position of the vehicle can help to increase the confidence that steering angle changes are a parking manoeuvre.

• Historical x,y coordinates: A location area where repeatedly EOJ events have occurred in the past may be a strong indicative of an EOJ event with a high degree of confidence.

• Map Information: Having information about the surroundings of the vehicle can increase the accuracy of the system. For instance, knowing that the car is entering a parking area (e.g., shopping centre, airport, etc.) will increase the probability of an end-of-journey event. Conversely, being on a motorway will dramatically decrease the probability of an EOJ event although the vehicle might be stationary (e.g., traffic jam situation). Note that this data may be combined with current x,y coordinates of the vehicle.

• Destination Prediction: Based on historical data, this input will give information on the probability of an EOJ event based on already-recorded destinations.

Thus, one or more of said vehicle parameters may be monitored as they change during a journey and a set of said measurements over a predetermined period of time may be collected and compared with stored sets of corresponding data to produce an output comprising a prediction of an EOJ event within a given time frame or distance to be travelled by the vehicle and a level of confidence that the EOJ event will occur in that time frame or distance.

For example, over a period of, say, two minutes, the steering angle of the steering wheels may be measured and recorded as a set of steering angles at specific time points (e.g. each separated by one second). This set may be compared with stored sets of steering angle measurements made over corresponding time periods and that resulted in an end of journey event. A degree of similarity may be determined and hence a probability that the current set of steering angle measurements is the precursor of an end of journey event. However, such a procedure would be intensive in terms of processing and storage. In one embodiment, pre-processing of a subset of the set of said measurements is effected comprising a Fourier transform of said subset in respect of a period of time over which said subset of measurements is collected to produce a set of the coefficients of the frequency components of said subset, said set of coefficients comprising the set compared with said stored set that also comprises coefficients of frequency.

That is, instead of storing the actual measurements that are made, a set of (a relatively large number of) measurements is pre-processed to produce a set of (a relatively small number of) coefficients of frequencies of change in the measurements actually made. The frequencies of change of measurement will range between zero change and a maximum rate of change. A particular pattern of frequencies and the extents or coefficients of certain frequencies may be highly indicative of an end of journey event.

It is to be understood that the set of measurements made, pre-processed and compared with stored sets will be updated periodically. Indeed, a current set of coefficients may include later measurements from a set of measurements already made and pre-processed, with older measurements from the set being progressively discarded. In this way, a current set of coefficients may be from a set of measurements that overlap with a previous set of measurements.

The parameters above cannot be directly used efficiently without a degree of pre-processing, including any one or more of:

• ln-vehicle signal extraction;

• Signal downsampling;

• Signal scaling and normalization;

• Noise reduction; and

· Simplification of signal representation (e.g., domain transformation)

Vehicle steering angle movements can be particularly indicative of a parking event, which usually is also highly indicative of an EOJ event. If steering angle movements (from 0° to 720°, being generally the number of degrees of steering wheel rotation from lock to lock) are plotted against time over a journey there is an infinitely varying signal produced. By performing a Fourier transform, such a signal can be reduced to a sum of an infinite number of sine waves, each with a specific coefficient. Over a short period of time during a parking event, the coefficients of some specific sine waves have a typical and dominant value. These values for specified sine waves can be stored in a memory of the EOJ predictor and used to compare with a signal received during a journey.

Likewise, other parameters may be stored as indicators of an EOJ event. For example, vehicle location, especially in association with certain other parameters, may be highly indicative of an EOJ event. In that respect, a precise vehicle location, such as a precise parking bay in a car park, is not required. Instead, a location area, repeatedly visited, in combination with a characteristic sequence of steering angle changes, may be indicative of an EOJ event with a high degree of confidence. Such an event may occur where a driver parks daily in a large area car park, but in different parking bays.

In one embodiment, default EOJ indicators may be provided as factory-fitted. These include major coefficients and sine waves of steering angle movements that precede typical EOJ events, seat-belt unbuckling etc. Such factory-fitted parameter maps give initial (but not personalised) EOJ event predictions. In one embodiment, however, the EOJ predictor also includes a learning algorithm. Thus different parameters of vehicle systems are recorded in real time and predictions of an EOJ event made. However, if an EOJ event is subsequently determined, or not, as the case may be, the particular set of parameters may be stored and used for future comparisons of real time events for the purpose of making better and more confident predictions.

In one embodiment, the output (ie the prediction) of the EOJ predictor is of a probability or level of confidence of an EOJ event occurring within a specific timeframe. Indeed, the output may be multiple probabilities of an EOJ event occurring in different time frames. Alternatively, or additionally, the output may be one or more probabilities of an EOJ event within one or more distances from the current location.

Entering an EOJ mode of operation following a prediction of an EOJ, but when an EOJ event does not, in fact occur, may have different consequences for some vehicle systems. This is a false-positive prediction. Indeed, a false positive may have more serious consequences than a false negative, which is a failure to predict an EOJ event with a sufficient level of confidence to initiate EOJ modes. For example, premature lowering of the vehicle suspension system when an EOJ event is predicted may lead to bumps in the road not being adequately cushioned. At a minimum, this may cause discomfort to vehicle occupants and, in a worst case scenario, could lead to damage to the vehicle. This is more serious than early turning off of the HVAC system, for example.

Thus, in one embodiment, the EOJ controller requires a different level of confidence (ie probability) of an EOJ event occurring for different vehicle systems before the EOJ mode is operated for a given system.

The EOJ controller may continue to monitor the vehicle, after initiation of an EOJ mode of different vehicle systems, to determine whether the EOJ event does in fact occurred, as predicted, within the relevant time frame or distance. The EOJ controller may disable the EOJ mode and return the relevant system to normal, in-journey (IJ), operation in the event of failure to detect an EOJ event. Again, the need to restart IJ operation in the event of a false- positive prediction that is acted upon by the EOJ controller varies from vehicle system to vehicle system. In one embodiment, the EOJ predictor continuously updates its predictions. By this is meant, at least, that a new prediction is made by the EOJ predictor at intervals during a journey. In the event of a false-positive (ie a probability of an EOJ within a specific time frame or distance which is not realised in practice), and before resuming normal IJ operation, the EOJ controller may act on an updated prediction, either in the same way as during normal IJ operation, or by requiring a greater level of confidence of an EOJ event, before acting upon the false-positive determination and resuming IJ operation. For example, the fact that the ICE has not been operated may not need to be reversed if there is still plenty of charge in the EE energy storage device and the probability of an EOJ event occurring in the near future remains sufficiently high. This may be determined not only by the probability, but also by the number of occasions a false positive has been recorded and, indeed, the remaining charge on the EE energy storage device. Similarly for HVAC systems. However, for air suspension or vehicle lighting, continued running of the vehicle in EOJ mode may be undesirable. In the case of the suspension, an intermediate position may be maintained or initiated after a false positive if the probability of an EOJ event remains high.

In another respect, the present invention provides an end-of-journey (EOJ) system for use in a vehicle, the system comprising:

collection means to receive, and collect into time limited sets, measurements of vehicle parameters during a journey of a vehicle in which the system is fitted;

comparison means to compare said sets with stored sets of parameters that are indicative of a probable and EOJ event within a time frame and/or distance of travel;

prediction means to issue a prediction of an EOJ event within a time frame or distance travelled; and

means to initiate an EOJ mode of one or more vehicle systems on the basis of said prediction. Said prediction may comprise a level of confidence with which an EOJ event will occur and said means to initiate an EOJ mode of one or more vehicle systems depends on said level of confidence.

The collection means may comprise a collection section provided on a controller having a processor and electronic memory. For example, the collection section may receive measurements via an input that communicates with a sensor, and the time limited sets may be stored on the electronic memory.

The comparison means may comprise a comparison section operable to compare said sets with stored sets of parameters that are indicative of a probable and EOJ event and provided on a controller having a processor and electronic memory. The comparison section may communicate with the collection section, and may also include or communicate with electronic memory having said stored sets of parameters saved thereon. Instructions for comparing said sets with stored sets of parameters that are indicative of a probable and EOJ event may be stored in the electronic memory. The prediction means may comprise a prediction section provided on a controller having a processor and electronic memory. The prediction means may communicate with the comparison means, and may issue said prediction in dependence on the comparison of aid sets with stored sets of parameters that are indicative of a probable and EOJ event. The means to initiate an EOJ mode may comprise an initiation section, which section may be provided on a controller having a processor and electronic memory, which controller may be in communication with said one or more vehicle subsystems and said prediction means.

It will be understood that some or all of the collection section, the comparison section, the prediction section and the initiation section may be provided in a common controller having a processor and electronic memory. The electronic memory may be a non-transitory computer readable media.

In an embodiment said prediction comprises a level of confidence with which an EOJ event will occur and said means to initiate an EOJ mode of one or more vehicle systems depends on said level of confidence.

Optionally, the prediction is of multiple probabilities of an EOJ event occurring in different time frames or distances.

In an embodiment the means to initiate an EOJ mode are operable to initiate an EOJ mode in a plurality of vehicle systems, wherein entering an EOJ mode of operation following a prediction of an EOJ requires a different level of confidence in the prediction depending on the vehicle system having the EOJ mode.

According to another aspect of the present invention there is provided a vehicle comprising an end-of-journey system as claimed in any preceding claim.

In an embodiment an EOJ event is defined as occurring when a driver operates a driver's door handle of the vehicle, and, optionally, when the engine of the vehicle is disabled. Optionally, an EOJ mode of operation is for an air suspension system, and comprises a lowering of the suspension.

The EOJ mode of operation may be for the lighting system of the vehicle, and may comprise lights of the lighting system arranged during darkness to be switched on and provide all round illumination of the immediate surroundings of the vehicle.

In an embodiment the vehicle is a hybrid vehicle having both a driving electric motor of the vehicle powered by an electric engine (EE) energy storage device, and an internal combustion engine (ICE) that drives the vehicle at higher speeds and/or charges the EE energy storage device, and in which an EOJ mode of operation comprises deliberately deselecting operation of the ICE and/or charging of the EE energy storage device.

Optionally, an EOJ mode of operation of a heating, ventilation and air conditioning (HVAC) system of the vehicle comprises shutting down the HVAC system before the end of a journey.

Optionally, the EOJ system continues to monitor the vehicle after initiation of an EOJ mode of one or more vehicle systems to determine whether the EOJ event occurs within the relevant time frame or distance, and wherein, in the event that the EOJ event does not occur, performs one of:

a) terminating the EOJ mode in at least one of the systems in which it has been initiated; and

b) maintaining the EOJ mode in at least one of the systems in which it has been initiated or maintained, provided that the EOJ predictor continues to output a prediction of an EOJ event occurring within a specific time frame or distance of travel. Further optionally, the EOJ system may maintain the EOJ mode in at least one of the systems in which it has been initiated or maintained only when:

a) the prediction of an EOJ event occurring is with a greater level of confidence than the confidence with which the EOJ event was previously predicted and on the basis of which the EOJ mode was initiated or, in the event it has been maintained, last maintained; and/or

b) the consequence of maintaining the EOJ mode if an EOJ event does not occur on the basis of a current prediction does not exceed a predetermined threshold.

In an embodiment the EOJ predictor continuously updates its predictions. In an embodiment the EOJ predictor is provided with one or more of the following vehicle parameters or contextual data on the basis of which to predict an EOJ event:

a) vehicle speed;

b) vehicle acceleration and braking;

c) steering angle;

d) seatbelt unbuckling;

e) gear mode;

f) directional indicators.

Optionally, the EOJ predictor is provided with more than one of said vehicle parameters contextual data on the basis of which to predict an EOJ event.

In an embodiment the EOJ predictor is additionally provided with one or more of the following vehicle parameters or contextual data on the basis of which to predict an EOJ event:

g) x,y current position coordinates;

h) map information at current position; and

i) destination prediction. Optionally, one or more of said vehicle parameters may be monitored as they change during a journey and a set of said measurements over a predetermined period of time may be collected and compared with stored sets of corresponding data to produce an output comprising a prediction of an EOJ event within a given time frame or distance to be travelled by the vehicle and a level of confidence that the EOJ event will occur in that time frame or distance. Further optionally pre-processing of a subset of the set of said measurements may be effected comprising a Fourier transform of said subset in respect of a period of time over which said subset of measurements is collected to produce a set of the coefficients of the frequency components of said subset, said set of coefficients comprising the set compared with said stored set that also comprises coefficients of frequency.

Said subset of the set of values on which said pre-processing is effected may comprise steering angle measurements.

In an embodiment the EOJ predictor includes a learning algorithm, whereby sets of parameters of measurements made of different vehicle systems during a journey that lead to an EOJ event are recorded in real time and comparisons are made between current sets of comparable parameters and said recorded sets wherein, on the basis of which comparisons, predictions of an EOJ event are made.

According to a further aspect of the present invention there is provided a method of controlling one or more vehicle systems that have an end-of-journey (EOJ) mode of operation, the method comprising:

receiving measurements of vehicle parameters during a journey of the vehicle;

collecting said measurements into time limited sets;

making a comparison between said sets and stored sets of parameters that are indicative of a probable and EOJ event within a time frame and/or distance of travel;

making a prediction of an EOJ event responsive to said comparison; and initiating said EOJ mode responsive to said prediction.

Within the scope of this application it is expressly envisaged that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and/or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. Features described in connection with one embodiment are applicable to all embodiments, unless such features are incompatible.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which:

Figure 1 a plot of steering angle of a vehicle against time;

Figure 2 is a block diagram showing signal processing of a signal like that shown in Figure 1 to identify Fourier coefficients;

Figure 3 is a block diagram showing production of coefficients from a discrete signal, such as buckling/unbuckling safety belts in a vehicle; and

Figure 4 is a block diagram of a system in accordance with an embodiment of the present invention

DETAILED DESCRIPTION

In Figure 1 , a signal 3 is produced from a steering rack sensor of steering angle plotted against time during a journey. The plot shows the steering movements made by the driver. The plot is plus or minus units of measure (that may or may not be the same as degrees of rotation of a steering wheel) around a zero position being a straight-ahead position for the vehicle, plotted against time during a journey. It is evident that region 1 of the plot demonstrates substantial back and forth swings of steering angle over a short space of time that is highly characteristic of the steering changes made during parking of a vehicle. Parking of a vehicle is highly predictive of the end of a journey, but as demonstrated earlier in the signal,

A similar plot may be drawn of numerous vehicle characteristics, although the shape of such plots will be very different, depending on the parameter detected. For example, a plot of vehicle speed, or acceleration/deceleration, or gear selection may be made.

In Figure 2, the processing of how such a plot may be handled is illustrated. Here, s1 is the value of the parameter being measured at a particular time, and is stored in register 10. A time signal 2 is supplied to the register which shifts the value in register 10 to the next register 12, whereupon the value is relabelled s2 and a new value of the current parameter s1 is captured in register 10. After N time periods, registers 10 to 14 are filled with values s1 , s2, s3 ... sN. How many measurements are made depends on the parameter sensed and how rapidly it changes. That is, N, and the time period 2, are selectable.

Once all the registers are filled, the values are supplied to multiplexer 18 and a fast Fourier transform performed on the data at 20 to establish the pattern of change between values - that is, the frequency components of the captured array. These are demultiplexed at 22 providing M outputs (f1 , f2, f3 ... fM) in registers 24. These are the coefficients of the frequency components of the input signal s1 to sN. These frequency components obviously change continuously but some patterns of them are characteristic of an end-of-journey (EOJ) event. In block 26 known coefficients are stored and compared against the coefficients output from registers 24. When a match between experienced coefficients and EOJ characteristic coefficients is detected, this will be predictive of an EOJ event. When a match occurs the probability of an EOJ event happening within a timeframe can be calculated. This might be on the basis of how close the experienced coefficients are to an EOJ characteristic set of coefficients (that is, small differences might still indicate a likelihood of an EOJ event, but with less confidence than an exact match) or even an exact match of one set may not be absolutely certain of an EOJ event, or have less probability of being an EOJ event than another set of coefficients. Given that the set of coefficients is continuously updated (potentially with each time signal 2, depending on how short the time signal period is and how many values (N) have to be multiplexed and the speed of operation of the multiplexer 18, Fast Fourier Transform (FFT) 20 and demuliplexer 22), the probability that an EOJ event is about to occur is also continuously updated. Thus, at point 4 of the steering angle plot 3, it might already be feasible to predict with levels of confidence that an EOJ event is about to happen. However, point 4 is not greatly different from point 5, for example, which in the journey represented by the plot 3 was clearly not an EOJ event. Consequently, using steering angle alone might not give sufficient confidence of an EOJ event.

However, it is not necessary to rely solely on steering angle. Indeed, a similar plot could be employed of gear position, or speed, and compared with the plot of steering angle. For example, at point 4, reverse gear may have been selected, whereas it is not selected at point 5. Selection of reverse gear could be employed as an indicator of an EOJ event and produce its own distinct characteristic.

With reference to Figure 2, the important numbers from a practical perspective are N and M. N is the number of time samples collected in each snapshot. The bigger value that N has, the longer are the manoeuvres that are detected. There is a trade-off here in terms of memory and processing power required. M is the number of frequency coefficients that are output. The larger the value of M, the higher is the resolution. Again, however, there is the same trade-off in terms of computing power and memory. A and B are integers. They are not significant and primarily indicate that M and N are a number power of 2, because and FFT algorithm is optimal for inputs/outputs of the size 2^X. However, other values of M and N can be considered and the algorithm will work as well. A is greater than B because one of the goals of the FFT is to achieve dimensionality reduction without loss of too much (or indeed any critical) information. A frequency transformation achieves that, so that M can be much smaller than N without losing important information. Turning to Figure 3, a similar arrangement to that of Figure 2 is illustrated which could be the selection of reverse gear. In fact, here it is illustrated as a buckling/unbuckling event of a seat belt of the vehicle. For example, an unbuckled seat belt may have a logical value 0, whereas a buckled seat belt may have a logical value 1 . The value of, for example, the driver's seat belt, may be input to register 30 as input s1 . When a time signal 2' is received, the value in s1 is shifted to s2 and the current value of the seat belt input as new s1 . The combined signal s1 ,s2 is then analysed against the simple logic:

F(s1 ,s2) = s2 AND (NOT sl ) which gives four possible outcomes, one of which indicates an unbuckling event.

If an unbuckling event occurs when the probability determined by other factors suggests an EOJ event is imminent, the probability that this is a correct prediction would be increased. The output 32 could become one of the parameters in the set 26.

The same methodology could be employed with the seat belt of each passenger seat of the vehicle.

Map data may also be employed to assist prediction of an EOJ event. For example, the probability score that the dynamic parameters mentioned above return could be increased by a given percentage if the vehicle enters a car park, or shopping centre or filling station, or a cul de sac. Entry of the vehicle in any of these locations is not enough to predict an EOJ on its own, but in association with other data, the vehicle being in this vicinity may well add confidence that an EOJ event is imminent.

In respect of the above, a new vehicle, incorporating the system of the present invention, could be provided with many (but still only a limited number of) sets of criteria that are always predictive of an EOJ event. When an EOJ event is predicted with a level of confidence, then the vehicle may enter EOJ modes for a number of its systems as described further below.

However, there may also be, for a given vehicle and a given driver, many parameter criteria that experience can determine are also predictive of an EOJ event. These are criteria that could not be said in advance to be predictive, but just happen to be so from the habits and experience of the driver and vehicle concerned. A primary example of this is position data of the vehicle when the vehicle approaches a location where an EOJ event has occurred in the past. Indeed, there are two primary situations regarding past use:

1 . User travels on a certain route repeatedly and parks always on the same parking space or in parking spaces in close proximity;

2. User travels on a certain route repeatedly and parks regularly in different places which are separated by a substantial (walking) distance.

Thus following a certain route, or indeed simply programming a route using satellite navigation, enables a destination prediction to be made, well in advance of reaching an EOJ event, but again enabling combination with other parameters to predict an EOJ event in a specific time frame.

Thus, turning to Figure 4, a system 100 embodies one example of the present invention. Here, in-vehicle parameters 42, GPS and map data parameters 44, and Destination

Prediction data 46, along with other parameters 48 that may be selected, are combined as contextual data in a multiplexer 50 to provide a pre-processor 52 with data from which EOJ identifier sets 54 are produced and output. The pre-processor carries out several functions including:

• In-vehicle signal extraction

· Signal down-sampling

• Signal scaling

• Noise reduction

• Simplification of signal representation, e.g., domain transformation A first set 54^ is provided to learning module 56 where the EOJ identifiers are compared with existing EOJ identifiers and a real time EOJ prediction 58 ensues, which is supplied to an In- vehicle computer system 60.

The prediction comprises two elements: a time frame (or distance) to the EOJ event, and a level of confidence associated with the prediction.

Depending on the level of confidence, the In-vehicle computer system 60 may or may not act on the prediction. The system 60 may instigate one or more of the following EOJ modes:

• Lowering the air suspension of the vehicle to facilitate egress from the vehicle after EOJ, and/or to provide an "elegant arrival" at the destination.

• Activation of parking cameras during parking manoeuvres. • Automatic initiation of autonomous parking.

• Automatic initiation of parking assistant feature.

• Change in the mapping of the steering wheel, making it softer during parking

manoeuvres.

· Automatic change of the configuration of the mirrors.

• Deactivation of an internal combustion engine and activation of the electrical engine for hybrid cars:

o to make the parking manoeuvre softer and give driver more power and control in difficult terrain, again also enhancing "elegant arrival" at the destination o to avoid unnecessary charging of vehicle electric engine energy storage

device

• Deactivation of HVAC systems

• Activation of vehicle systems that prepare the vehicle for going to sleep (e.g. saving changes or data logs)

For example, a high level of confidence may be required before a vehicle's air suspension is lowered, because of the potentially serious consequences of premature lowering when an EOJ event does not occur. Moreover, the air suspension may only be lowered when the EOJ is imminent, rather than when an EOJ likely to happen in a time from in excess of one minute or thereabouts. On the other hand, HVAC systems can be disabled when the degree of confidence is quite low, and possibly some significant time in the future (e.g. three or four or more minutes away).

Also, when an EOJ event does not occur within the time frame predicted, the In-vehicle computer system 60 may disable the EOJ mode of any of the systems it has placed in that mode and return them to ln-journey modes. However, that may conflict with a continuing prediction of an EOJ event. The decision to act on a prediction may depend on previous predictions that turned out to be false positives, as well as current conditions of the vehicle. For example, the HVAC system may have been disabled several minutes previously but still the journey has not ended. A new prediction that the journey will end may arrive. On the one hand, the computer system may wish to restart the HVAC systems (return them to ln-journey modes), because the conditions of the vehicle are deteriorating. It may also wish to do so because the level of confidence that an EOJ event will happen, as currently predicted by the Learning Module 56, is no better than or different from the previous false-positive prediction. That is, the In-vehicle computer system 60 may monitor whether or not an EOJ event actually occurs within the predicted time frame and adjust any of the EOJ modes accordingly, and potentially in dependence upon the continuing state of the EOJ prediction.

The second set 54₂ of EOJ identifiers is supplied to a delay block 62 which holds the identifiers for a period of time. This is to enable the actuality or otherwise of an EOJ event to be associated with the identifiers. Thus contextual data 66 indicative of an EOJ event is supplied to reference signal generator 68 which, if the contextual data 66 confirms an EOJ event returns a reference signal 70 (which conveniently may be a binary 1 or 0) that is multiplexed with the delayed EOJ identifiers in multiplexer 72 and supplied to learning module 56 for storage and use subsequently in predicting (or not) a future EOJ when the same set of identifiers is received (as set 54^ on a future journey. In that way, the learning module builds up a database of identifiers that have resulted in an EOJ improving the accuracy with which future predictions are made. Throughout the description and claims of this specification, the words "comprise" and "contain" and variations of them mean "including but not limited to", and they are not intended to (and do not) exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

Features, integers, characteristics, compounds, chemical moieties or groups described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.

Claims

1 . An end-of-journey (EOJ) system for use in a vehicle, the system comprising:

comparison means to compare said sets with stored sets of parameters that are indicative of a probable EOJ event within a time frame and/or distance of travel;

means to initiate an EOJ mode of one or more vehicle systems on the basis of said prediction.

2. An end-of-journey system as claimed in claim 1 , wherein said prediction comprises a level of confidence with which an EOJ event will occur and said means to initiate an EOJ mode of one or more vehicle systems depends on said level of confidence.

3. An end-of-journey system as claimed in claim 1 or claim 2, wherein the prediction is of multiple probabilities of an EOJ event occurring in different time frames or distances.

4. An end-of-journey system as claimed in any preceding claim, wherein the means to initiate an EOJ mode are operable to initiate an EOJ mode in a plurality of vehicle systems, wherein entering an EOJ mode of operation following a prediction of an EOJ requires a different level of confidence in the prediction depending on the vehicle system having the EOJ mode. 5. A vehicle comprising an end-of-journey system as claimed in any preceding claim.

6. A vehicle as claimed in claim 5, wherein an EOJ event is defined as occurring when a driver operates a driver's door handle of the vehicle, and, optionally, when the engine of the vehicle is disabled.

7. A vehicle as claimed in claim 5 or 6, wherein an EOJ mode of operation is for an air suspension system, and comprises a lowering of the suspension.

8. A vehicle as claimed in any one of claims 5-7, wherein an EOJ mode of operation is for the lighting system of the vehicle, and comprises lights of the lighting system arranged during darkness to be switched on and provide all round illumination of the immediate surroundings of the vehicle.

9. A vehicle as claimed in any one of claims 5-8, which vehicle is a hybrid vehicle having both a driving electric motor of the vehicle powered by an electric engine (EE) energy storage device, and an internal combustion engine (ICE) that drives the vehicle at higher speeds and/or charges the EE energy storage device, and in which an EOJ mode of operation comprises deliberately deselecting operation of the ICE and/or charging of the EE energy storage device.

10. A vehicle as claimed in any one of claims 5-9, wherein an EOJ mode of operation of a heating, ventilation and air conditioning (HVAC) system of the vehicle comprises shutting down the HVAC system before the end of a journey. 1 1 . A vehicle as claimed in any one of claims 5-10, wherein the EOJ system continues to monitor the vehicle after initiation of an EOJ mode of one or more vehicle systems to determine whether the EOJ event occurs within the relevant time frame or distance, and wherein, in the event that the EOJ event does not occur, performs one of:

b) maintaining the EOJ mode in at least one of the systems in which it has been initiated or maintained, provided that the EOJ predictor continues to output a prediction of an EOJ event occurring within a specific time frame or distance of travel.

A vehicle as claimed in claim 1 1 , wherein the EOJ system maintains the EOJ mode least one of the systems in which it has been initiated or maintained only when:

b) the consequence of maintaining the EOJ mode if an EOJ event does not occur on the basis of a current prediction does not exceed a predetermined threshold. 13. A vehicle as claimed in any preceding claim, wherein the EOJ predictor continuously updates its predictions.

14. A vehicle as claimed in any one of claims 5-13, wherein the EOJ predictor is provided with one or more of the following vehicle parameters or contextual data on the basis of which to predict an EOJ event:

a) vehicle speed;

b) vehicle acceleration and braking;

c) steering angle;

d) seatbelt unbuckling;

e) gear mode;

f) directional indicators.

15. A vehicle as claimed in claim 14, wherein the EOJ predictor is provided with more than one of said vehicle parameters or contextual data on the basis of which to predict an EOJ event.

16. A vehicle as claimed in claim 14 or claim 15, wherein the EOJ predictor is additionally provided with one or more of the following vehicle parameters or contextual data on the basis of which to predict an EOJ event:

g) x,y current position coordinates;

h) map information at current position; and

i) destination prediction.

17. A vehicle as claimed in any of claims 14-16, wherein one or more of said vehicle parameters are monitored as they change during a journey and a set of said measurements over a predetermined period of time is collected and compared with stored sets of corresponding data to produce an output comprising a prediction of an EOJ event within a given time frame or distance to be travelled by the vehicle and a level of confidence that the EOJ event will occur in that time frame or distance. 18. A vehicle as claimed in claim 17, wherein pre-processing of a subset of the set of said measurements is effected comprising a Fourier transform of said subset in respect of a period of time over which said subset of measurements is collected to produce a set of the coefficients of the frequency components of said subset, said set of coefficients comprising the set compared with said stored set that also comprises coefficients of frequency.

19. A vehicle as claimed in claim 18, wherein said subset of the set of values on which said pre-processing is effected comprises steering angle measurements. 20. A vehicle as claimed in any one of claims 5-19, wherein the EOJ predictor includes a learning algorithm, whereby sets of parameters of measurements made of different vehicle systems during a journey that lead to an EOJ event are recorded in real time and comparisons are made between current sets of comparable parameters and said recorded sets wherein, on the basis of which comparisons, predictions of an EOJ event are made.

21 . A method of controlling one or more vehicle systems that have an end-of-journey (EOJ) mode of operation, the method comprising:

receiving measurements of vehicle parameters during a journey of the vehicle; collecting said measurements into time limited sets;

making a prediction of an EOJ event responsive to said comparison; and initiating said EOJ mode responsive to said prediction 22. A vehicle comprising:

one or more vehicle systems that have an end-of-journey (EOJ) mode of operation; an EOJ predictor, adapted to output an in-journey prediction of an EOJ event occurring within a specific time frame or distance of travel; and

an EOJ controller which, when the EOJ predictor predicts the occurrence of an EOJ event, initiates said EOJ mode of one or more of said vehicle systems.

23. A vehicle as claimed in claim 22, wherein an EOJ event is defined as occurring when a driver operates a driver's door handle of the vehicle, and, optionally, when the engine of the vehicle is disabled.

24. A vehicle as claimed in claim 22 or 23, wherein an EOJ mode of operation is for an air suspension system, and comprises a lowering of the suspension.

25. A vehicle as claimed in any one of claims 22-24, wherein an EOJ mode of operation is for the lighting system of the vehicle, and comprises lights of the lighting system arranged during darkness to be switched on and provide all round illumination of the immediate surroundings of the vehicle.

26. A vehicle as claimed in any one of claims 22-25, which vehicle is a hybrid vehicle having both a driving electric motor of the vehicle powered by an electric engine (EE) energy storage device, and an internal combustion engine (ICE) that drives the vehicle at higher speeds and/or charges the EE energy storage device, and in which an EOJ mode of operation comprises deliberately deselecting operation of the ICE and/or charging of the EE energy storage device.

27. A vehicle as claimed in any one of claims 22-26, wherein an EOJ mode of operation of a heating, ventilation and air conditioning (HVAC) system of the vehicle comprises shutting down the HVAC system before the end of a journey.

28. A vehicle as claimed in any one of claims 22-27, wherein the output of the EOJ predictor is a probability of an EOJ event occurring within a specific timeframe or distance to be travelled.

29. A vehicle as claimed in claim 28, wherein the prediction is of multiple probabilities of an EOJ event occurring in different time frames or distances. 30. A vehicle as claimed in claim 28 or 29, wherein entering an EOJ mode of operation following a prediction of an EOJ, requires a different level of confidence of the prediction depending on the vehicle system having the EOJ mode.

31 . A vehicle as claimed in any preceding claim, wherein the EOJ controller continues to monitor the vehicle, after initiation of an EOJ mode of one or more vehicle systems, to determine whether the EOJ event occurs within the relevant time frame or distance, and wherein, in the event that the EOJ event does not occur, performs one of:

a) terminating the EOJ mode in at least one of the systems in which it has been

initiated; and

32. A vehicle as claimed in claim 31 , wherein the EOJ controller maintains the EOJ mode in at least one of the systems in which it has been initiated or maintained only when: a) the prediction of an EOJ event occurring is with a greater level of confidence than the confidence with which the EOJ event was previously predicted and on the basis of which the EOJ mode was initiated or, in the event it has been maintained, last maintained; and/or

33. A vehicle as claimed in any preceding claim, wherein the EOJ predictor is provided with one or more of the following vehicle parameters or contextual data on the basis of which to predict an EOJ event:

a) vehicle speed;

b) vehicle acceleration and braking;

c) steering angle;

d) seatbelt unbuckling;

e) gear mode;

f) directional indicators.

34. A vehicle as claimed in claim 33, wherein one or more of said vehicle parameters are monitored as they change during a journey and a set of said measurements over a predetermined period of time is collected and compared with stored sets of corresponding data to produce an output comprising a prediction of an EOJ event within a given time frame or distance to be travelled by the vehicle and a level of confidence that the EOJ event will occur in that time frame or distance.

35. A vehicle as claimed in claim 34, wherein pre-processing of a subset of the set of said measurements is effected comprising a Fourier transform of said subset in respect of a period of time over which said subset of measurements is collected to produce a set of the coefficients of the frequency components of said subset, said set of coefficients comprising the set compared with said stored set that also comprises coefficients of frequency.

36. A vehicle as claimed in claim 35, wherein said subset of the set of values on which said pre-processing is effected comprises steering angle measurements.

37. A vehicle as claimed in any one of claims 22-36, wherein the EOJ predictor includes a learning algorithm, whereby sets of parameters of measurements made of different vehicle systems during a journey that lead to an EOJ event are recorded in real time and comparisons are made between current sets of comparable parameters and said recorded sets wherein, on the basis of which comparisons, predictions of an EOJ event are made.

38. An EOJ system for a vehicle, substantially as described herein and with reference to any of the accompanying drawings.

39. A vehicle, substantially as described herein and with reference to any of the accompanying drawings.