Vehicle Battery Monitoring

Abstract

Vehicle battery voltage data is received from a telematics control unit (TCU) of a vehicle dining multiple driving cycles, and analysed to determine a state of health (SOH) and state of charge (SOC) of the battery. The TCU also provides data relating to the state of an ignition switch of the vehicle for use in analysing the voltage data. A starting probability factor for the vehicle is determined and monitored. Information about the ability of the battery to start the vehicle is output to the user.

Claims

1-29. (canceled)

30. A method of obtaining information in relation to the state of a vehicle battery for provision to a user, wherein the battery is used to power cranking of an engine of the vehicle, the method comprising: obtaining data indicative of a voltage of the battery at a plurality of different times during each one of a plurality of driving cycles of the vehicle; obtaining data indicative of a timing of each one of the plurality of driving cycles of the vehicle; obtaining a set of one or more driving cycle related parameters using at least a portion of the data indicative of the timing of the plurality of driving cycles of the vehicle; obtaining a set of one or more voltage related parameters using at least a portion of the data indicative of the voltage of the battery at a plurality of different times during each one of the driving cycles; determining, for one or more times, a starting probability factor for the battery indicative of a likelihood that the battery will be capable of starting the engine of the vehicle at an applicable time, wherein the starting probability factor is determined using at least a portion of the obtained set of driving cycle related parameters and at least a portion of the set of one or more obtained voltage parameters; and using the starting probability factor or factors determined for at least a portion of the one or more times to generate information based on the starting probability factor of the battery for the applicable time for output to a user.

31. The method of claim 30, wherein the data indicative of the timing of one or more driving cycles of the vehicle is obtained based on the at least a portion of the obtained data indicative of the voltage of the battery at a plurality of different times during each one of the plurality of driving cycles, and obtained data indicative of the timing of one or more transitions of the vehicle between ignition states of the vehicle.

32. The method of claim 31, wherein the data indicative of the timing of the one or more transitions of the vehicle between different ignition states is received from a telematics control unit (TCU) of the vehicle.

33. The method of claim 31, wherein the data indicative of the transition of the vehicle between ignition states is obtained using data indicative of a state of an ignition switch of the vehicle.

34. The method of claim 30, wherein the data indicative of the voltage of the battery at different times during the plurality of driving cycles of the vehicle is data received from a telematics control unit (TCU) of the vehicle.

35. The method of claim 30, comprising receiving a feed of data indicative of the voltage of the battery of the vehicle with respect to time from a telematics control unit (TCU) of the vehicle, and storing the data in a database of vehicle battery voltage data.

36. The method of claim 30, further comprising obtaining data indicative of the voltage of the battery via one or more of: at a plurality of different times between driving cycles of the vehicle; during an ignition off state of the vehicle.

37. The method of claim 30, further comprising obtaining data indicative of the voltage of the battery at different sampling rates depending upon a detected ignition state of the vehicle, and sampling data indicative of the voltage of the battery at a first rate when the vehicle is in an ignition off state, and at a second, higher rate, in an ignition on state of the vehicle.

38. The method of claim 30, wherein the ignition on or off state of the vehicle is determined via one or more of: based upon a transition in the state of an ignition switch of the vehicle; and using a telematics control unit (TCU) of the vehicle.

39. The method of claim 30, wherein the starting probability factor further takes into account temperature, wherein the temperature used in determining the starting probability factor is determined at least in part using weather data applicable to the time to which the starting probability factor relates, wherein the method comprises using data indicative of the position of the vehicle to select weather data applicable to the position of the vehicle.

40. The method of claim 39, wherein the data indicative of the position of vehicle is received from the vehicle and comprises live positional data.

41. The method of claim 39, further comprising determining a correction to the temperature at the location of the vehicle as indicated by the weather data for use in determining the starting probability factor, the method comprising using one or more temperature measurements from one or more temperature sensors associated with the vehicle to determine the correction to the temperature.

42. The method of claim 39, further comprising using engine usage history in determining a correction to the temperature at the location of the vehicle as indicated by the weather data for use in determining the starting probability factor.

43. (canceled)

44. The method of claim 30, wherein the one or more voltage related parameters include an open circuit voltage.

45. The method of claim 30, further comprising using at least a portion of the set of one or more voltage related parameters and at least a portion of the set of one or more driving cycle related parameters to determine a state of charge (SOC) and a state of health (SOH) of the battery, wherein the starting probability factor for a given time is based on the applicable SOC and SOH of the battery.

46. The method of claim 45, wherein the obtained set of one or more driving cycle parameters includes an average duration of a driving cycle and an average interval between cycles, and the SOC of the battery is determined based on the obtained average duration of a driving cycle and the average interval between driving cycles.

47. The method of claim 46, wherein the set of one or more voltage related parameters includes a difference between a first two minima in the voltage after initiation of a cranking mode in a driving cycle, wherein said difference is additionally used in determining the SOC of the battery.

48. The method of claim 30, wherein the information generated is a warning, advice or prompt to the user selected based on the applicable starting probability factor.

49. A non-transitory computer readable medium having a computer program product embodied thereon and comprising computer readable instructions executable by a processor direct the performance of operations comprising: obtaining data indicative of a voltage of a battery at a plurality of different times during each one of a plurality of driving cycles of an associated vehicle; obtaining data indicative of a timing of each one of the plurality of driving cycles of the vehicle; obtaining a set of one or more driving cycle related parameters using at least a portion of the data indicative of the timing of the plurality of driving cycles of the vehicle; obtaining a set of one or more voltage related parameters using at least a portion of the data indicative of the voltage of the battery at a plurality of different times during each one of the driving cycles; determining, for one or more times, a starting probability factor for the battery indicative of a likelihood that the battery will be capable of starting the engine of the vehicle at an applicable time, wherein the starting probability factor is determined using at least a portion of the obtained set of driving cycle related parameters and at least a portion of the set of one or more obtained voltage parameters; and using the starting probability factor or factors determined for at least a portion of the one or more times to generate information based on the starting probability factor of the battery for the applicable time for output to a user.

50. A system comprising a server device configured to obtain information in relation to the state of a vehicle battery for provision to a user, wherein the battery is used to power cranking of an engine of the vehicle, the server device configured to: obtain data indicative of a voltage of a battery at a plurality of different times during each one of a plurality of driving cycles of an associated vehicle; obtain data indicative of a timing of each one of the plurality of driving cycles of the vehicle; obtain a set of one or more driving cycle related parameters using at least a portion of the data indicative of the timing of the plurality of driving cycles of the vehicle; obtain a set of one or more voltage related parameters using at least a portion of the data indicative of the voltage of the battery at a plurality of different times during each one of the driving cycles; determine, for one or more times, a starting probability factor for the battery indicative of a likelihood that the battery will be capable of starting the engine of the vehicle at an applicable time, wherein the starting probability factor is determined using at least a portion of the obtained set of driving cycle related parameters and at least a portion of the set of one or more obtained voltage parameters; and use the starting probability factor or factors determined for at least a portion of the one or more times to generate information based on the starting probability factor of the battery for the applicable time for output to a user; and the system further comprising a positioning system of the vehicle configured to provide data indicative of a position of the vehicle to the server device, the server device further configured to determine the temperature expected at a position of the vehicle at the applicable time, using the data indicative of a position of the vehicle to select weather data applicable to the position of the vehicle.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Some preferred embodiments of the present invention will now be described, by way of example only, and with reference to the accompanying drawings, in which:

[0093] FIG. 1A shows voltage with respect to time before, during and after engine start-up, indicating various elements that can be detected from voltage measurements during engine cranking;

[0094] FIG. 1B illustrates the way in which the regions indicated in FIG. 1A correspond to positions of the ignition switch;

[0095] FIG. 2 illustrates the behaviour of the voltage during start-up in more detail;

[0096] FIG. 3 illustrates the voltage behaviour while the vehicle is not running;

[0097] FIG. 4 is a flow chart illustrating a way in which the algorithm may operate in accordance with an embodiment of the invention;

[0098] FIG. 5 illustrates the way in which SOH may depend upon the number of cycles experience by the battery and the battery age;

[0099] FIGS. 6A, B and C illustrate the way in which battery age, number of starting cycles, and SOH may be categorised respectively;

[0100] FIG. 7 illustrates an example of information which may be output to a user using a desktop device;

[0101] FIG. 8 illustrates another example of a display for a user in relation to a mobile device;

[0102] FIG. 9 illustrates the voltage behaviour during cranking for two start-up cycles.

DETAILED DESCRIPTION

[0103] In embodiments at least, the present invention provides a method which enables a starting probability of a vehicle battery to be remotely monitored by a server, and a user alerted when there may be an issue with the ability of the battery to start the vehicle. The method is performed by analysing voltage and ignition state data received from a TCU of a vehicle.

[0104] In accordance with one exemplary embodiment of the invention, the voltage of a vehicle battery is measured with the aid of the analogue to digital converter (ADC) of a Telematics Control Unit (TCU). While the main purpose of a TCU device is to retrieve various controller area network (CAN) information such as odometer or error messages, it has been found that its applications can be extended with the aid of an algorithm as described herein, with no need of extra hardware and only an additional payload to the data communication.

[0105] The TCU hardware is connected directly to the battery to enable voltage measurements for use in the algorithm described herein to be obtained. A connection from the TCU to the hard-wired ignition switch is also provided. The latter connection is already used in some previous applications.

[0106] The voltage of the battery is sampled by the ADC of the TCU at different sampling rates, depending whether the car is in an ignition off or ignition on state. The voltage measurements are transmitted to a server for use as an input to the algorithm described below.

[0107] The driving cycle of a vehicle may be detected with the aid of the hard-wired ignition switch and the voltage readings. The TCU installation follows a so-called 3-wire installation (+CAN connection). 3-wire stands for Power, Ground and Ignition. Once connected to these three wires of the vehicle CAN, the TCU can determine whether the vehicle is in an ignition on or ignition off state.

[0108] When the vehicle is in an ignition off state, the sampling rate will be low, e.g. hourly. The TCU maintains a rolling average of the last five voltage measurements, and stores the most recent such average value. When the next transition to an ignition on state is detected, the most recent average voltage value from the ignition off state is transmitted to the server. It will be appreciated that if the ignition off state exceeds a predetermined duration e.g. 2 hours, the TCU may enter a standby mode until the next transition to an ignition on state is detected. In the standby mode, voltage measurements continue to be taken and stored, but certain functionality of the TCU is disabled. For example, position data e.g. GPS data may not be taken, as it may be assumed that the vehicle is stationary. A modem of the TCU may be disabled, such that the voltage measurements are stored locally without transmission to the server. When the vehicle is in an ignition off state prior to entering a standby mode, data may be stored for later transmission in a similar manner to that during standby, or may be transmitted to the server.

[0109] Upon detection of a transition to an ignition on state, the TCU starts to sample voltage at a higher sampling rate, and to transmit voltage measurements to the server. This ensures that data of an appropriate resolution is captured to enable sufficient information about the subsequent cranking mode and drive mode to be obtained. At The TCU continues to sample voltage data and transmit it to the server at the higher sampling rate until the next transition to an ignition off state is detected, at which time the sampling rate decreases once more, and voltage data is stored for subsequent transmission (upon initiation of a further ignition on state).

[0110] The rate at which voltage measurements are taken when the ignition on state is initiated is drastically increased compared to that in the ignition off state, due to the relatively short time that cranking takes. A resolution of 10 ms would be ideal in order to detect local and global peaks, i.e. to detect start-up timings and revolutions. However, in practice the resolution achievable is limited by hardware, as the TCU should not be overloaded, as this may interfere with its ability to perform other conventional functions of the TCU e.g. downloading of CAN signals. It has been found that a data rate of 40 milliseconds per sample may provide a balance between providing a useful data resolution and avoiding overloading of the TCU.

[0111] By way of example, voltage measurements may be taken hourly in the ignition off state, and at 50 Hz during an ignition on state. Of course, more complex arrangements may be envisaged in which transitions between modes of the vehicle are detected by reference to a position of an ignition switch e.g.

[0112] accessories activated mode, cranking mode, drive mode. The voltage sampling rate may differ between these modes.

[0113] FIG. 1A shows the various elements that can be detected from voltage measurements during engine cranking. The open circuit voltage (OCV) (1) can be observed at the leftmost side of the figure. This figure illustrates the different voltage states which may be identified, and how they relate to the mode of vehicle e.g. a state of an ignition switch. In the region labelled 1, the vehicle is in a “locked” mode, in which the battery supplies energy only to the anti-theft and other low power consumption specific loads. The vehicle is in an ignition off state. In the left hand region 2, the ignition has transitioned to an ignition on state. All of the electrical consumers in the car are being activated (AC/seat heating/radio etc.). This may be referred as an “accessories activated” mode. Region 3 corresponds to the period when the key is in the ignition position, i.e. the engine is starting/cranking. This corresponds to the cranking mode of the vehicle. In region 4 engine start-up has finished, and the vehicle is in drive mode. FIG. 1A labels various features of the voltage which may be used in accordance with the invention. These include the Open Circuit Voltage (OCV) steady state voltage in the locked mode 1, the voltage before cranking occurs in region 2, the first and second minima in voltage after initiation of start up in region 3, the start up duration corresponding to the duration of region 3, the end of engine start up, and the start of alternator charging.

[0114] FIG. 1B illustrates the way in which these states correspond to positions of the ignition switch. In the region 1, the switch is in the “lock” position. In region 2 the switch is in the “acc” i.e. accessories activated position. In region 3 the switch is in the “start” position. In region 4 the switch is in the “on” position (i.e. for driving).

[0115] As mentioned above, it is not necessary to detect the positions of the ignition switch. By capturing voltage data e.g. as illustrated in FIG. 1A, the transition between modes within an ignition on state of the vehicle may be identified. Thus, in embodiments, only a transition between ignition off and ignition on states of the vehicle is detected, in order to adjust the sampling rate appropriately to ensure that the resulting data is of a sufficient resolution to enable certain features of the voltage during the various modes within the ignition on state to be identified.

[0116] The behaviour of the voltage during start-up can also be analysed. This is shown in FIG. 2. The initial voltage drop should be instantaneous, and should be the lowest value reached during cranking. The voltage drop may be taken to define the start of the engine cranking mode. As the engine RPM increases, the load on the electromotor decreases and the battery voltage recovers. There might be some local peaks corresponding to the high mechanical resistance induced by the compression process of the 4-stroke engines.

[0117] In the example of FIGS. 1 and 2, the cranking time is approximately 1 second. However, this time i.e. the duration of the cranking mode can vary from a couple hundreds of milliseconds to 2.5 seconds, or even more. A calculation of the cranking time may be performed. After the engine is running on itself i.e. in the drive mode, depending on several factors, such as SOC, temperature, Engine Control Unit (ECU) management etc., the alternator may start charging the battery.

[0118] The voltage measurement may be calibrated. Ideally, the battery voltage would be measured at the battery terminal. However, this would mean extra hardware would need to be implemented. It is recognised that the voltage measurements obtained using the ADC of a TCU might be lower than the actual voltage drop over the battery, due to the series voltage drop over the wires or diodes/fuses. Calibration may be done by comparing the voltage measured by the TCU's ADC and the one measured by an oscilloscope connected to the batteries terminals directly.

[0119] The voltage behaviour while the vehicle is not running i.e. in an ignition off state will now be discussed by reference to FIG. 3. In FIG. 3, the typical evolution of the battery voltage is shown when the vehicle is not running. V.sub.0 represents the battery steady state OCV after the end of a driving cycle. As soon as the engine is turned off, it is expected for the voltage to drop linearly to V.sub.1, as the surface charge induced by the alternator wears off. Afterwards, at t.sub.1, the voltage is expected to vary, but to steadily decrease overall. V.sub.2 is the voltage before the next driving cycle, i.e. before the driver starts the car again. The time difference between t.sub.0 and t.sub.2 can range from a couple of minutes to even months. Depending on the environment conditions, the steady state of the battery may be reached during this interval. If so, it may be taken as a reference for the SOC. Thus, a reference voltage corresponding to the steady state of the battery may be determined using this data relating to the voltage of the battery between driving cycles.

[0120] In accordance with the invention, the voltage measurements obtained by the TCU, and data indicative of the ignition state of the vehicle obtained by the TCU are transmitted to a server for use in calculating the starting probability factor (SPF) for the battery. The ignition state data, and the voltage data during the ignition on state of the vehicle, is transmitted to the server in real-time, enabling the SPF for the battery to be monitored i.e. continually updated i.e. to provide a live “SPF”. The SPF may be calculated as a continual function in terms of a percentage probability that the vehicle battery will be able to start the vehicle. Alternatively, it is envisaged that the SPF may be calculated on a demand basis i.e. when a user requests information about the state of the vehicle battery. The algorithm may then process the data available for the vehicle at the given time.

[0121] One way in which an algorithm may operate to implement an embodiment of the invention will now be described by reference to FIG. 4.

[0122] FIG. 4 illustrates a way in which an algorithm may operate using input data from a database 10. The database 10 includes stored voltage and ignition state data obtained from the TCU of a vehicle over time.

[0123] The algorithm starts at step 1.

[0124] At step 3 data from the database 10 is processed to obtain two battery parameters, SOH and SOC upon which the SPF is based.

[0125] A state of health (SOH) for the battery may be determined based on the total number of cycles that the battery sustained since it was installed in the vehicle, as well as the battery age. For convenience, the user will be asked to input if the battery was replaced, its manufacturing date and the battery type (e.g. typically wet cell or AGM). The SOH of a battery is indicative of its absolute capacity for energy storage, and will be affected by the energy loss over time while the vehicle is not driving.

[0126] A state of charge (SOC) for the battery will also be determined based on an average duration of a driving cycle and average interval between driving cycles. The SOC also takes into account a difference between the first two minima in the voltage encountered after initiation of start up i.e. cranking mode, as shown in FIG. 1A. If steady state conditions apply as mentioned above, the SOC will be calibrated with the aid of the applicable OCV. The applicable OCV will be the OCV associated with the most recent locked state of the vehicle where steady state voltage was attained. The SOC of a battery is indicative of the amount of energy available in the battery.

[0127] It will be appreciated that the average driving cycle duration and the average interval between driving cycles may be determined based upon the voltage data received by the server from the TCU, together with the ignition state data. Such data is stored in the database 10. The combination of the voltage data and the ignition state data enables the timing of driving cycles to be determined (and the timing of individual modes within the driving cycles i.e. the start and end of cranking mode), and thus the cycle duration and interval between cycles may be obtained. As new data is received from the TCU, the average values for these parameters may be updated. Thus, they are based on the historical data received from the TCU and the most recent data received.

[0128] In step S the operation of the alternator is checked. If there is no evidence of alternator activity during the previous driving cycles, the algorithm would immediately generate a warning for output to a user—step 7. Without the alternator, a vehicle is prone to stall while driving, if it starts at all. The operation of the alternator may be evaluated by checking historical data of the OCV of the battery after startup, while the engine is running. The OCV should indicate more than 14V when the alternator is enabled.

[0129] Provided the alternator checkup is passed, the algorithm will proceed to the next stage—step 9.

[0130] As mentioned above, the SPF may be calculated continually, or upon request by a user. In either case, when the SPF for a given time is determined, the algorithm generates an output e.g. message, such as a warning, alert or prompt for the user dependent upon the value of the SPF. Some exemplary output for given levels of SPF will now be described. The output may be provided to a user on a mobile device or other device e.g. PC of the user. The output may be provided via an app. The output may be provided automatically if certain conditions are met e.g. if the SPF falls below a given threshold, or in response to a request by a user. Alternatively or additionally, the output may be provided to a user via a display associated with the vehicle, e.g. of an ADAC thereof. It will be appreciated that the output may be provided to more than one user. The users may be users of the vehicle, or anyone with an interest in the performance of the battery e.g. a fleet manager, someone responsible for maintaining the vehicle etc. The SPF may be in respect of a current or future time.

[0131] The starting probability factor (SPF) reflects the likelihood that the battery will be able to start the vehicle, and is dependent upon both SOC and SOH of the battery. The starting probability factor is lowest i.e. 0% where both SOH and SOC are low i.e. the battery is decayed and discharged—condition 12. The SPF is highest i.e. 100% where both SOH and SOC are high i.e. the battery is healthy and charged—condition 14.

[0132] If the battery is discharged and decayed (old with many start-up cycles)—condition 12, i.e. where SOC and SOH are both low, the algorithm will generate data to advise the user to replace the battery (step 20) and will generate a warning that the current battery is highly likely to fail while cranking (step 22).

[0133] If the battery is healthy and charged (condition 14), a message may be generated indicating that startup is expected to be fine.

[0134] Of course, the SPF may have any value between these extremes. The action taken by the algorithm for intermediate values of SPF will now be described by reference to two possible intermediate values, corresponding to conditions 16 and 18, in which the battery is healthy and discharged, and the battery is decayed and charged respectively.

[0135] In condition 16, where the SOH is good, but the SOC is not, the algorithm will prompt the user to either check for any parasitic loads straining the battery (step 26), and/or will provide advice on how to recharge the battery (step 24) e.g. to try to use longer driving cycles and/or to externally recharge the battery. The latter advice will be given if the average driving cycles are short and/or infrequent. The existence of any parasitic loads would be associated with a “steep” OCV drop while offline (e.g. the drop as shown in FIG. 3).

[0136] After the user has taken any suggested action, the SOC is recalculated—step 28. If the SOC is still extremely low, the algorithm would give a failure warning step 30. Regardless of how healthy the battery is, if the SOC is low, there will not be any potential energy stored in the battery to be supplied. If the SOC is mildly low, the system will check for the forecasted temperature—step 32. A negative (degrees Celsius) ambient temperature might be troublesome for the cranking process. The cranking probability increases with temperature. If the temperature according to weather data is below a predetermined threshold e.g. is zero degrees Celsius or below, a warning may be provided to the user that startup is unlikely to succeed—step 34 (“potential issue with startup). If the temperature is above the threshold, then a message may be provided that startup should be fine—step 36 (“startup ok”).

[0137] In condition 18, if the SOC is good, but the SOH is not, i.e. the battery is decayed and charged, the algorithm may also advise a battery replacement—step 38. However, this advice may not be provided in all circumstances. If the user keeps the battery charged and the following season is not a cold one, the expected battery cycle life may be exceeded. The algorithm will therefore next consider SOH of the battery—step 40. If the SOH is average, there ought to be no issues with the cranking process, and the algorithm will indicate that startup should be ok—step 42 (“startup ok”). If SOH is low, the algorithm follows step 32, and then step 34 or 36 as appropriate, as described in relation to condition 16.

[0138] It will be appreciated that the SPF may have any value within a continuous range between 0% and 100%, and need not be categorised into one of the four conditions 12, 14, 16 and 18 described. For intermediate conditions of the battery, the algorithm is able to determine suitable action using a fuzzy logic approach based on the described conditions. For example, where SOH and SOC are both average, the algorithm may identify a divided approach based on the action taken in relation to conditions 16 and 18 described above.

[0139] It will be appreciated that other battery parameters may be determined e.g. cranking amps, which is indicative of the speed at which energy may be dissipated by the battery.

[0140] In the example given above, the SOH of the battery is defined as a function of the number of starting cycles and the age of the battery. FIG. 5 illustrates the way in which SOH depends upon these parameters. For convenience the weight factor is evenly distributed between the number of cycles and the age of the battery.

[0141] FIGS. 6A, B and C illustrate the way in which battery age, number of starting cycles, and SOH may be categorised respectively.

[0142] The SOC may be defined analogously to the SOH, with an added factor quantifying the voltage difference between the first two initial minimum peaks.

[0143] Other data may be used to improve the accuracy of the SOH and SOC determinations. For example, one or more of battery rating and vehicle fuel and engine type, mileage and service history may be taken into account. These may fine tune the operation of the algorithm. In particular, additional factors may be taken into account to provide a more accurate reflection of SOH, which may depend on factors such as temperature, capacity and the depth of discharge during cranking.

[0144] The way in which temperature data may be taken into account will now be described. This data may be used in refining estimates of SOH, SOC and hence SPF, and also in determining the action to be taken by the user as illustrated in FIG. 4.

[0145] The server receives a feed of data indicative of the position of the vehicle with respect to time. This may be GPS data or any other suitable data received from a positioning system of the vehicle. The position data may or may not be supplied by the TCU device. When determining the SOH of the vehicle, the server obtains weather data indicative of the temperature expected at the applicable time at the position of the vehicle. This may be obtained from any suitable weather service e.g. from another server. The temperature obtained according to the weather forecast may then be adjusted based upon readings obtained from temperature sensors associated with the vehicle. These may include an oil temperature, water temperature, and temperature within the TCU housing where the TCU includes a temperature sensor. This may help to compensate for any difference between the temperature of the environment surrounding the vehicle, and the actual temperature of the vehicle e.g. due to the precise position of the vehicle, such as in a garage, and the thermal inertia of the vehicle hood. Other weather data may also be obtained and used to try to achieve a more accurate value for the e.g. battery parameters e.g. air pressure, sun coverage. The air pressure may affect oxygen density. A lower oxygen density is associated with lower combustion energy. Engine usage data may also be used in helping to adjust a temperature obtained based on weather data. For example, the server may be arranged to store trip history data. If the vehicle has recently been driven, e.g. if the preceding driving cycle was not long ago, the vehicle battery may be expected to still be at an elevated temperature in comparison to the case in which the vehicle had been parked e.g. for 24 hours.

[0146] It will be appreciated that data is accessed from multiple sources in order to implement embodiments of the invention e.g. vehicle TCU, vehicle GPS unit, weather service. The data obtained from different sources may be analysed to compare timestamps, and GPS coordinates as appropriate to correlate the data.

[0147] FIG. 7 illustrates an example of information which may be output to a user in the form of a start-up failure warning e.g. as in step 22 or 30 of FIG. 4. This may be provided to the user where the user accesses an app to provide real-time battery information, or where the user has requested an alert in the case of possible startup failure or in the event of cold weather conditions etc. FIG. 7 shows the information as it may be output via a desktop type device. As shown on FIG. 8, the information may equally be provided via a mobile device.

[0148] Parasitic loads refer to devices which are using electricity when the car is not running and the doors are locked. Such devices might include mis-grounded light bulbs, lights mistakenly left on, or a device connected to the lighter socket of the car. The present invention provides the ability to detect such loads and alert the user to take action before the battery is drained to too great an extent.

[0149] Some more detailed examples of the way in which SOH and SOC for a battery may be defined will now be described.

[0150] A generic formula for the SOH may be defined as follows:

SOH=1−f(N,age,SOC.sub.average,temp.sub.average)

[0151] The SOH is dependent on the age of the battery (age), the number of starting cycles it has provided (N), the average SOC (SOC.sub.average) and the average temperature (temp.sub.average).

[0152] The number of cycles can also vary, depending on the battery type or percentage of the capacity required for start-up. For simplicity a factor K.sub.n will quantify all of these factors and will result in a smaller or higher N.

[00001]$N=K_n\times n_{\mathrm{rated}},\mathrm{where}{K}_n=\{\begin{array}{c}1,\mathrm{for}\mathrm{wetcell}\\ 0.7,\mathrm{for}\mathrm{AGM}\end{array}$

[0153] For convenience, the age of the battery can be considered a linear factor, whereas for the average SOC and temperature, an analogous rationale to the number of cycles is applicable. The decay rate of the battery is proportional to the environment temperature and inversely proportional to the SOC. However, for the purposes of the present invention, these constants will be set to one and furthermore, the SOH will be oversimplified to the following formula.

SOH=1−f(N,age)

[0154] The SOC is defined as a function of the simple moving average of the last 10 driving cycles duration, the ignition ON/OFF time ratio delimited by these cycles and the difference between the first two minimum voltage peaks during the cranking process. The ignition “ON” time corresponds to time spent in the accessories activated, cranking and drive modes of the vehicle i.e. the time other than in an ignition off mode.

[00002]$\stackrel{\_}{\mathrm{cycles}}=\frac{1}{10}\underset{i=1}{\overset{10}{.Math.}}{\mathrm{cycle}}_i$

[0155] The SOC can be further calibrated based on the steady state OCV when this is achieved.

SOC=f(cycles,ON/locked.sub.average,OCV,ΔVoltage)

[0156] To enhance the SOC calculation, the difference between the first two minimum peaks during cranking are compared. As can be observed in FIG. 9, which illustrates the startup voltage for different SOCs, the second minimum peak is significantly lower when the battery is drained i.e. when SOC is low than when the SOC is high.

[0157] When determining any parameter which is an average over a number of cycles, e.g. average duration of cycle, or interval between cycles, the average may be determined as a rolling average over a predetermined number of cycles e.g. 10 cycles. This may help to reduce the amount of data to be processed. Thus, older data need not necessarily be stored.

[0158] A flat or faulty car battery is one of the most commonly encountered reasons for vehicle downtime, year-round. Because of this, the electromotor is not able to start-up the combustion engine of the car, especially in cold climates. The cause of this malfunction can range from manufacturing defects, physical damage, electrical loads, faulty alternator and more. Moreover, the number of electronics embedded in modern cars have increased significantly and this trend will continue, stressing a higher load on the car's electrical system. Therefore, considering the telematics technology evolution, as well as the decrease in their price and the need of diminishing the vehicle's downtime, a system that is able to continuously estimate the battery state of health (SOH) is desirable. As a result, a user, whether a vehicle owner, or a fleet manager may have a better insight into the technical status of the vehicles. As the information may become available at a telematics service platform, this information can be used to plan for maintenance more efficiently, ensuring the replacement of the battery in time.