Method and processing device for moderating a driver's confidence in automation

Abstract

A method (2) for moderating a driver's automation confidence comprises the steps of: quantifying (21) driver cooperation based on driver actions performed during an observation period; quantifying (22) the driver's automation experience as a function of longitudinal and / or lateral vehicle control automation during the observation period; quantifying (23) the vehicle occupant's automation confidence based on the quantified driver cooperation and automation experience; checking (24) whether the quantified automation confidence meets a predetermined trigger criterion; and triggering (25) a moderation action to moderate the driver's automation confidence if the quantified automation confidence meets the trigger criterion.

Description

[0001] The invention relates to a method for moderating a driver's confidence in the automation of a vehicle. Furthermore, the invention relates to a processing device and a computer program for executing such a method, as well as a computer-readable storage medium on which such a computer program is stored.

[0002] Advanced driver assistance systems (ADAS) are now widespread. For example, semi-automated driving functions according to SAE Level 2 can independently perform certain dynamic driving tasks such as steering, accelerating, and braking. However, the driver remains responsible and must be ready to take control at any time. Typical applications include adaptive cruise control (ACC) or lane keeping assist, which can also be combined into a comprehensive longitudinal and lateral driving function.

[0003] As the performance and functionality of ADAS increase, it becomes ever more difficult for the driver to distinguish between the individual functions. At the same time, research shows that in predominantly automated driving without active driver involvement, there is a certain probability that the driver will develop a very high level of trust in the automation, even to the point of so-called "overtrust" (i.e., excessive trust in the automation), and consequently no longer adequately monitor the automated driving function. This can lead to misuse of the ADAS, because sufficient driver cooperation is necessary for safety reasons, especially with partially automated driving systems.

[0004] To counteract this effect, measures for moderating, i.e., specifically regulating, trust in automation itself are known. These measures can include, for example, specific cues directed at the driver, such as an Attention Request (ATR), a Hands-On Request (HOR), or a Take-Over Request (TOR).

[0005] A method described in the applicant's patent specification EP 4 132 830 B1 provides for the issuance of a moderation signal after a predetermined number of semi-automated functions have been performed, in order to prevent the development of excessive confidence in the automation. The number of functions is predetermined based on the performance of the ADAS and, if applicable, on other factors such as the driver's risk profile or the output of a random number generator.

[0006] It is an object of the present invention to provide an improved method for moderating a driver's confidence in automation, in order to ensure sufficient driver cooperation, particularly when using advanced semi-automated driving functions.

[0007] The problem is solved by the subject matter of the independent claims. Advantageous embodiments are specified in the dependent claims.

[0008] According to a first aspect of the present invention, a computer-implemented method for moderating a driver's confidence in one or more automated driving functions of a vehicle is proposed.

[0009] The vehicle in question may be, in particular, a motor vehicle. The term "motor vehicle" here refers specifically to a land vehicle that is moved by mechanical power and is not bound to railway tracks. A motor vehicle in this sense can be, for example, a passenger car, a motorcycle, or a tractor.

[0010] The one or more automated driving functions can be, in particular, semi-automated driving functions according to SAE Level 2.

[0011] A first step of the procedure is to quantify driver cooperation based on actions taken by the driver during an observation period - for example, since the start of the journey or over several journeys.

[0012] Driver cooperation can be quantified, in particular, by considering the driver's manual interventions in the longitudinal and / or lateral control of the vehicle. These manual interventions can include, for example, braking, accelerating, and / or steering. "Manual" interventions, therefore, do not only refer to steering movements, but also include, for example, the driver's use of the vehicle pedals with their feet or other direct actions by the driver to perform dynamic driving tasks, which, however, do not necessarily have to be performed with their hands.

[0013] The advantage here is that the detection of whether the driver is accelerating, braking, applying steering (counter)torque (steering with or oversteering / countersteering), or making similar manual interventions can be performed with high integrity. The method is therefore fundamentally suitable for implementation with defined integrity (e.g., according to ASIL) and can thus contribute positively to functional safety.

[0014] Alternatively or in addition to a dependence on manual interventions by the driver in the longitudinal and / or lateral control of the vehicle, driver cooperation can also be quantified depending on other actions that contribute to vehicle controllability or controllability of the driving situation.

[0015] This can include, for example, one or more of the following actions, which may indicate driver cooperation: activating driver assistance functions via a display; adjusting display settings, such as a head-up display (HUD); changing a driving mode; operating a horn, turn signal, hazard warning lights, or windshield wipers; adjusting vehicle lights, such as switching them on or off or activating high beams; adjusting a mirror, windshield heating or ventilation, or the defog mode of an air conditioning system; adjusting the driver's seat to a driving position. In general, this can include all operating and confirmation actions that have a direct or indirect influence on the vehicle dynamics or the driver's monitoring capabilities.

[0016] Such activities by the driver can be seen as cooperative behavior and can be positively factored into the quantification of driver cooperation.

[0017] A further step is to quantify the driver's experience of automation. This is determined based on the actual and objectively verifiable automation of the vehicle's longitudinal and / or lateral control during the observation period.

[0018] Building on this, the driver's trust in automation is quantified in a further step.

[0019] Quantifying automation confidence is based on quantified driver cooperation and quantified automation experience. Specifically, this can be done in such a way that quantified automation confidence—in line with the empirical findings mentioned earlier—increases with increasing quantified automation experience and decreases with increasing quantified driver cooperation. In other words, the greater the quantified automation experience, the greater the quantified automation confidence tends to be, and the greater the quantified driver cooperation, the lower the quantified automation confidence tends to be.

[0020] In a further step, it is checked whether the quantified automation confidence meets a predetermined trigger criterion for initiating a moderation measure. This may involve, in particular, a comparison of the quantified automation confidence with a previously defined threshold. Specifically, it may be stipulated that the trigger criterion is met if the quantified automation confidence exceeds a predetermined confidence threshold.

[0021] If the above-described test shows that the quantified automation confidence meets the trigger criterion, a moderation measure is triggered in a further step to moderate the driver's automation confidence.

[0022] According to one embodiment, the moderation measure comprises one or more of the following: issuing a message to the driver, for example, regarding their lack of cooperation; deactivating and / or downgrading an automated driving function, such as from a higher SAE level function to a lower SAE level function or manual driving, or indicating that such deactivation and / or downgrading is imminent if cooperation remains insufficient; or (possibly temporarily) preventing the activation of an automated driving function, such as functions above a certain SAE level, e.g., SAE level 2.

[0023] Alternatively or additionally, the moderation measure can include reducing the performance of the automated driving function(s), thereby forcing more driver cooperation. For example, a random number generator could be used to decide whether the automated driving function will control the vehicle towards an object or not, so that in the latter case, driver cooperation in the form of manual intervention would be required.

[0024] Furthermore, the moderation measure can include a change in the settings of one or more automated driving functions, thereby reducing the comfort aspect and increasing the warning aspect. Specifically, a driving dynamics reaction, such as deceleration to an object in front as part of longitudinal control, could occur later and be more abrupt, thus being perceived as less comfortable.

[0025] The instruction to the driver may also include a request to pay attention (ATR), to place hands on the steering wheel (HOR), or to take manual control of the vehicle (TOR).

[0026] It is also conceivable that information about a driver's insufficient cooperation could be passed on to a motor vehicle insurance company with the driver's consent and taken into account within the framework of an insurance tariff that financially incentivizes cooperative driver behavior. In such a case, the notification to the driver could specifically include a warning about the negative impact of their potentially insufficient cooperation on the insurance premium.

[0027] Such information can be conveyed to the driver in a manner known per se, for example, by optical, acoustic and / or haptic means.

[0028] According to one embodiment, quantifying the experience of automation can be based on influencing factors such as the duration during which one or more automated driving functions are activated, and / or the degree of automation (e.g., an SAE level) of one or more automated driving functions activated during the observation period. Such influencing factors can be evaluated, for example, using a state machine.

[0029] Additionally or alternatively, the quantification of the automation experience can also be carried out depending on individual actual influences of one or more automated driving functions on the vehicle guidance, such as on the basis of a number, intensity and / or duration of automatic interventions in the longitudinal and / or lateral guidance of the vehicle during the observation period.

[0030] According to some embodiments, the aforementioned dependence on the duration and / or intensity of the automatic interventions can be specified in such a way that the quantification of the automation experience is based on one or more of the following influencing factors: a duration of automatic lateral control; a dynamics of automatic lateral control, which can be determined in particular as a function of a steering angle velocity occurring during automatic lateral control; a duration of automatic longitudinal control; or a dynamics of automatic longitudinal control, which can be determined in particular as a function of accelerations occurring during automatic longitudinal control.

[0031] The term acceleration should be understood to include deceleration (as negative accelerations).

[0032] With regard to the aforementioned durations, it should be noted that these may, for example, be a total duration resulting from the sum of the individual durations of several automatic interventions or longitudinal or lateral controls (e.g., actual longitudinal or lateral control interventions) during the observation period.

[0033] The term "automatic interventions" can encompass both interventions within the scope of an automated driving function (such as ACC) and interventions within the scope of an automated parking function. This can also include, for example, automatic acceleration at a green light, automatic standstill protection, or generally situations in which the vehicle drives or maneuvers without driver input.

[0034] Furthermore, the frequency and / or duration of detailed displays related to automated driving functions, such as the representation of the vehicle's surroundings, can also be used as an indicator of the driver's experience of automation. In particular, it is possible to consider how often and / or for how long the driver actually views these displays.

[0035] According to another embodiment, quantifying the automation experience includes determining an automation metric that increases with one or more of the aforementioned influencing factors, i.e., the greater the respective influencing factor, the greater the automation metric tends to be.

[0036] Quantifying driver cooperation can be done, for example, based on one or more of the following influencing factors: the number and / or duration of recorded pedal actuations; recorded pedal actuation speeds (dynamics); the number and / or duration of manual steering inputs, especially steering wheel inputs; steering angle velocities and / or steering torques occurring during manual steering inputs; recorded gear shift inputs (especially their number), i.e., manual operation of a gear selector to select forward, reverse, or neutral, or a gear ratio of a vehicle transmission. For example, the aforementioned parameters, such as pedal actuation or steering angle velocities, can also be incorporated as values ​​integrated over the respective duration of the input. This makes it possible to consider the dynamics of the interventions.

[0037] Alternatively or additionally, hands-on detection carried out in a known manner (e.g., using recorded steering torques or a capacitive mat in the steering wheel rim) can also be used as an influencing factor for quantifying driver cooperation.

[0038] The durations mentioned above may also be durations of individual pedal and / or steering actuations (possibly also individual integrals of quantities recorded during the respective actuations, such as pedal actuation or steering angle velocities) summed over the observation period, similar to what was explained above in connection with the durations of automatic interventions as possible influencing factors in quantifying the experience of automation.

[0039] According to further training, driver cooperation can also be quantified based on, for example, head poses, gaze directions, or other driver characteristics captured by a driver-focused camera. The captured head poses and / or gaze directions can be considered in relation to the expected direction of gaze for an attentive driver in a given situation. For instance, if an attentive driver in a driving situation were to look through the windshield at the traffic situation (possibly also at specific areas in the vehicle's surroundings) or at specific displays with information regarding an active automated driving function, then the quantified driver cooperation score can be positively influenced if it is determined that the driver is indeed looking in the corresponding direction (and vice versa).

[0040] Quantifying driver cooperation can involve determining a (driver) cooperation metric that increases with the number, intensity and / or (e.g., a cumulative duration over the observation period) of the driver's manual interventions.

[0041] In particular, the cooperation measure can increase with one or more of the aforementioned possible influencing factors.

[0042] The increase in the cooperation measure with the aforementioned influencing factors can be implemented in the form of a counter, which is incremented, for example, by a predetermined unit or by an amount with each recorded pedal, steering, and / or shift operation. This amount can, in turn, increase with a recorded pedal actuation speed, a recorded steering angle speed, or a recorded steering torque.

[0043] Quantifying automation confidence can involve determining an (automation) confidence measure as a function of the cooperation measure and the automation measure, such that the resulting confidence measure is higher the higher the automation measure is, and lower the higher the cooperation measure is. This means that the confidence measure is calculated so that it increases with increasing automation measure (and conversely decreases with decreasing automation measure) and decreases with increasing cooperation measure (and conversely increases with decreasing cooperation measure).

[0044] These dependencies are consistent with the empirical facts already mentioned in the introduction: the more the driver intervenes manually in the vehicle control, the lower his confidence in automation, and conversely, confidence in automation increases in predominantly automated journeys without active driver involvement.

[0045] For example, determining the confidence measure may involve calculating a difference by subtracting the cooperation measure from the automation measure.

[0046] The vehicle can also issue warnings or perform automatic interventions through active safety functions that do not necessarily require driver input. Examples include lane departure warnings, speed limit warnings, or active front protection. Provided that these functions are designed / implemented in a way that serves as a warning and is not primarily intended for convenience, such warnings or interventions by active safety functions have a moderating effect, even though they are also perceived as automation. Therefore, further training may stipulate that confidence in automation is influenced by vehicle-generated warnings and / or interventions by one or more active safety functions.For example, such warnings or interventions by active safety systems can be taken into account by reducing the confidence score for automation. To this end, a corresponding measure can be determined for warnings and / or interventions by active safety functions experienced by the driver, which increases with each vehicle-side warning and / or intervention by an active safety function. This measure can, for example, be negatively factored into the calculation of the confidence score, i.e., subtracted from the difference between the automation score and the cooperation score.

[0047] A warning or active safety function can also be generated outside the vehicle, such as through acoustic / haptic feedback from the lane, horns from other road users, a speed limit warning, or a motorway exit warning. For moderation, the experience is crucial; that is, if the vehicle can recognize such external warning and safety functions, these can be taken into account when determining automation confidence. For this purpose, the aforementioned metric can consider such external moderation events in addition to vehicle-side warnings or interventions by active safety functions, or a separate metric can be calculated for this purpose, which, for example, can also be negatively factored into the calculation of the confidence metric, analogous to the metric mentioned above.It is conceivable that in the future there will be an increasing number of infrastructure-based feedback systems that notify the vehicle of a violation. Such information can also be considered as external moderation events.

[0048] Checking whether the quantified automation confidence meets the predetermined trigger criterion can involve comparing the confidence measure with a previously defined confidence threshold. Specifically, it can be stipulated that the trigger criterion is met, and consequently a moderation measure is initiated, if the result of the check is that the confidence measure exceeds the confidence threshold.

[0049] It is also within the scope of the invention that the moderation measure(s) and / or the trigger criterion, such as, in particular, the aforementioned trust threshold, can be individually adapted to the driver, for example, by an automatic setting based on a stored driver-specific moderation profile after the driver has been identified. The development of trust in automation also depends on the driver's personality. If the journey is personalized, a driver-specific moderation requirement or a driver-specific trust threshold can be defined.

[0050] For example, a particular driver might intervene regularly even during a long-term, highly automated driving experience. In this case, the moderation threshold (i.e., the trust threshold) could be raised. Conversely, there are drivers who quickly become overly reliant on automation and may even misuse it, which can manifest, for example, as a frequent activation of active safety functions. In this case, the moderation measures could be adjusted, for instance, so that the driver is warned only once and the function is then deactivated. Individualization can therefore be achieved by adjusting the trust threshold and / or the moderation measures.

[0051] A corresponding moderation profile can be exchanged via a backend, for example, so that a driver can take their moderation profile with them even when changing vehicles or using a rental car.

[0052] Furthermore, it is within the scope of the invention that the proposed method for moderating the driver's confidence in automation can be applied via a remote software update (RSU). This is particularly advantageous given that moderation becomes increasingly important as the performance of the automated driving functions in question improves.

[0053] For example, the automation trust moderation function may initially be deactivated because the underlying automated driving functions are not yet sophisticated enough to pose a risk of excessive automation trust. Over time, however, the performance of the automated driving functions may improve, for instance, through crowd-based machine learning to learn trajectories, intersection characteristics, etc. In this case, automation trust moderation can be activated or adjusted via the RSU to become effective.

[0054] In this way, the moderation of automation trust can be flexibly adapted to the maturity level of the automated driving functions. As long as the automation is not yet very sophisticated, there is no risk of excessive automation trust, so moderation is not necessary. However, as soon as the automation improves and the risk of building up excessive automation trust increases, moderation can be activated via RSU. Applying the proposed method via a remote software update thus enables flexible and needs-based adaptation to the current state of the art of automated driving functions, which further increases the safety of vehicle operation.

[0055] Overall, the proposed method enables targeted moderation of the driver's confidence in automation to ensure sufficient driver cooperation when using advanced semi-automated driving functions. Continuous monitoring and quantification of driver cooperation and the experience of automation allows for the early detection of excessive confidence in automation, which can then be corrected through appropriate moderation measures, thereby increasing the safety of vehicle operation.

[0056] Unlike previous approaches that considered driver moderation only within the specific scope of a single driving function, the present method can capture any perceptible automation of longitudinal and / or lateral control, thus mapping the driver's overall automation experience. If, as a result of the automatic interventions in vehicle control, the driver has the overall impression that the vehicle is driving in a "highly automated" manner, their confidence in the automation increases accordingly. Conversely, the "balance" of this confidence can decrease again if the driver has to intervene in vehicle control, intervenes voluntarily, or experiences uncomfortable situations, such as a warning or an intervention by an active safety function.This holistic approach to moderating automation trust makes it possible to strike a balance between the driver's experience of automation and their required cooperation.

[0057] A second aspect of the invention is a (data) processing device configured to execute a method according to the first aspect of the invention. Accordingly, the preceding and following explanations of the method according to the invention, as well as its possible embodiments, can be understood analogously for the processing device according to the invention, and vice versa.

[0058] The processing device can have at least one processor and be configured to carry out the method according to the first aspect of the invention by means of the at least one processor.

[0059] According to some embodiments, the processing unit can also be a spatially distributed processing unit (for example, across several processors or microcontrollers spaced apart from each other).

[0060] For example, the processing unit can be a control unit or part of a control unit of the vehicle. In particular, it can be a control unit for controlling one or more automated driving functions.

[0061] A third aspect of the invention is a computer program comprising instructions that, when executed by a processing unit (such as a processing unit according to the second aspect of the invention), cause it to execute a method according to the first aspect of the invention. The computer program may be divided into several separate subprograms, each of which can be executed by different, possibly spatially separated (sub-)processing units (such as several separate processors).

[0062] A processing device according to the second aspect of the invention can be configured (i.e. programmed) to execute a computer program according to the third aspect of the invention.

[0063] A fourth aspect of the invention is a computer-readable storage medium comprising instructions that, when executed by a processing device, cause it to execute a method according to the first aspect of the invention. In other words, a computer program according to the third aspect of the invention can be stored on the computer-readable storage medium.

[0064] The invention will now be explained in more detail with reference to exemplary embodiments and the accompanying drawings. Fig. Figure 1 illustrates, by way of example and schematically, an arrangement for moderating a driver's trust in automation. Fig. Figure 2 shows a block diagram of a procedure for moderating a driver's confidence in automation.

[0065] The in Fig. The arrangement shown in Figure 1 for moderating a driver's confidence in automation comprises, in particular, a (data) processing device 10 on which the Fig. The illustrated procedure 2 for moderating automation trust is executable. This processing unit 10 can simultaneously be a control unit of the vehicle, which controls one or more automated driving functions.

[0066] The processing unit 10 is connected to various sensors and components via signal technology in order to collect data relevant for the process 2.

[0067] An interior camera 11 captures information about the driver's head position, eye condition and gaze direction, which provides information about his involvement in driving and can be used to quantify driver cooperation.

[0068] A steering wheel 12 with sensors records a steering angle, a steering angle speed, a steering torque and a hands-on / off detection, which can also be included in the quantification of driver cooperation.

[0069] Similarly, pedal actions and speeds, which are detected by a pedal assembly 13 or by sensors arranged thereon, serve to quantify driver cooperation.

[0070] A data storage device 14 holds an individual driver profile, for example with a driver-specific trust threshold for one or more moderation measures, including driver-specific ones.

[0071] An optical display 15 serves to provide information to the driver as part of the moderation measure. Display 15 represents a variety of conceivable output interfaces, such as a loudspeaker for acoustic instructions or one or more vibration elements in the driver's seat or steering wheel rim for haptically perceptible instructions.

[0072] Procedure 2 for moderating automation trust comprises the following steps: In step 21, driver cooperation is quantified by determining a cooperation metric DCO based on various parameters that characterize manual interventions by the driver in the longitudinal and / or lateral control of the vehicle. These parameters are determined using the sensors and components described above. Fig. 1 recorded.

[0073] Information from the steering wheel 12 with integrated sensors, such as steering angle, steering angle speed and steering torque, is used to quantify the driver's manual steering interventions.

[0074] Pedal actions and speeds, which are recorded by the pedal assembly 13, also serve to record manual driver interventions in longitudinal guidance and are thus used to determine the cooperation measure DCO.

[0075] The more and more intensively the driver intervenes in vehicle control via the steering wheel and pedals, the higher the resulting cooperation metric DCO. This metric therefore increases with the number, intensity, and duration of manual driver interventions.

[0076] Additionally, data from the interior camera 11 regarding the driver's head position and gaze direction, which provide information about his attention and involvement in driving operations, can be included in the calculation of the cooperation measure DCO.

[0077] In step 22, the driver's experience of automation is quantified. This involves considering factors such as the duration of activation of automated driving functions, their degree of automation, and the number and dynamics of automatic longitudinal and lateral control interventions. The processing unit 10 can obtain such information, for example, directly from a control module for the relevant automated driving functions, which can also be executed on the processing unit (i.e., the control unit for the automated driving functions) 10. From this, an automation measure AG is derived, which increases with the degree of vehicle automation perceptible to the driver.

[0078] The longer and more comprehensively automated driving functions are active, and the higher their degree of automation, the greater the derived automation metric AG. For example, activating adaptive cruise control over a longer distance could contribute more to an increase in the automation metric AG than short-term interventions by the lane keeping assist system.

[0079] Based on the cooperation measure DCO and the automation measure AG, a confidence measure AV is then determined in step 23 as a measure of the driver's confidence in automation. This confidence measure is higher the higher the automation measure, but lower the higher the cooperation measure.

[0080] For example, the confidence level can be determined as the difference between the automation level AG and the cooperation level DCO: AV=AG−DCO

[0081] Next, in step 24, it is checked whether the determined confidence measure AV exceeds a predetermined, driver-specific adjustable confidence threshold AV_lim: AV>AV_lim?

[0082] If this is the case, a moderation measure is triggered in step 25. This moderation measure can take various forms, such as issuing a message to the driver via the visual display 15, deactivating or downgrading an automated driving function, or preventing the activation of such a function.

[0083] After completion of the moderation measure, or if the check in step 24 shows that the quantified automation confidence AV does not meet the trigger criterion AV > AV_lim, step 26 is intended to restart procedure 2 in order to ensure continuous monitoring and adjustment of the automation confidence.

[0084] In summary, the arrangement 1 with the method 2 serves to continuously monitor and moderate the driver's confidence in the automation system in order to ensure appropriate interaction between the driver and the vehicle automation. The various sensors and components, such as the interior camera 11, steering wheel 12, pedals 13, and data storage 14, provide the necessary information, which is processed in the processing unit 10. Method 2 first quantifies the driver's cooperation (step 21), the experience of automation (step 22), and consequently, the driver's confidence in the automation system (step 23). If the determined confidence in the automation system, i.e., the confidence measure AV, exceeds a confidence threshold AV_lim that can be individually set for the driver (test step 24), a moderation measure is triggered (step 25), and the method then starts again (step 26). QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] EP 4 132 830 B1

[0005]

Claims

[1] Method (2) for moderating a driver's confidence in automation of a vehicle, comprising the steps: - Quantifying (21) driver cooperation based on driver actions performed during an observation period; - Quantifying (22) the driver's experience of automation depending on the automation of the longitudinal and / or lateral control of the vehicle during the observation period; - Quantifying (23) the vehicle occupant's confidence in automation based on quantified driver cooperation and quantified automation experience; - Check (24) whether the quantified automation confidence meets a predetermined trigger criterion; - Triggering (25) a moderation action to moderate the driver’s confidence in automation if the quantified confidence in automation meets the trigger criterion. [2] Method (2) according to claim 1, wherein the quantification (21) of the driver cooperation is carried out depending on manual interventions by the driver in a longitudinal and / or lateral guidance of the vehicle during an observation period. [3] Method (2) according to any of the preceding claims, wherein the moderation measure comprises one or more of the following measures: - a notification; - deactivating and / or downgrading an automated driving function; - preventing the activation of an automated driving function. [4] Method (2) according to one of the preceding claims, wherein the quantification (22) of the automation experience is based on one or more of the following influencing factors: ▪ a duration during which one or more automated driving functions are activated in the observation period; ▪ a level of automation consisting of one or more automated driving functions activated during the observation period. [5] Method (2) according to one of the preceding claims, wherein the quantification (22) of the automation experience is based on one or more of the following influencing factors: a number and / or an intensity and / or a duration of automatic interventions by one or more automated driving functions in the longitudinal and / or lateral guidance of the vehicle during the observation period. [6] Method (2) according to claim 5, wherein the quantification (22) of the automation experience is based on one or more of the following influencing factors: ▪ the duration of an automatic lateral control; ▪ a dynamic of an automatic lateral control, in particular determined as a function of a steering angle velocity occurring during the automatic lateral control; ▪ the duration of an automatic longitudinal control; ▪ a dynamics of an automatic longitudinal control, in particular determined as a function of accelerations occurring during the automatic longitudinal control. [7] Method (2) according to any one of claims 4 to 6, wherein the quantification (22) of the automation experience comprises determining an automation measure (AG) as a function of one or more of the aforementioned influencing factors, wherein the automation measure (AG) increases with one or more of the influencing factors. [8] Method (2) according to one of the preceding claims, wherein the quantification (21) of driver cooperation is carried out depending on one or more of the following influencing factors: ▪ a number and / or duration of recorded pedal acts; ▪ recorded pedal actuation speeds; ▪ a number and / or duration of recorded manual steering operations; ▪ steering angle velocities and / or steering torques that occurred during manual steering operations; ▪ recorded manual switching operations; ▪ recorded head poses and / or gaze directions of the driver. [9] Method (2) according to any of the preceding claims, wherein the quantification (21) of driver cooperation comprises determining a cooperation measure (DCO), wherein the cooperation measure (DCO) increases with a number and / or an intensity and / or a duration of the manual interventions of the driver. [10] Method (2) according to claims 8 and 9, wherein the cooperation measure (DCO) is determined as a function of one or more of the influencing factors mentioned in claim 8. [11] Method (2) according to claim 7 and one of claims 9 and 10, wherein the quantification (23) of the automation confidence comprises determining a confidence measure (AV) as a function of the cooperation measure (DCO) and the automation measure (AV), wherein the confidence measure (AV) ▪ the larger the automation index (AV) is; and ▪ the smaller it is, the larger the cooperation measure (DCO) is. [12] Method (2) according to claim 11, wherein checking (24) whether the quantified automation confidence meets the predetermined trigger criterion comprises checking whether the confidence measure (AV) exceeds a predetermined confidence threshold (AV_lim). [13] Method (2) according to one of the preceding claims, wherein the triggering criterion and / or the moderation measure can be individually adapted to the driver. [14] Processing apparatus (10) which is configured to carry out a method (2) according to any of the preceding claims. [15] Computer program comprising instructions which, when the computer program is executed by a processing device (10), cause it to execute a method (2) according to any one of claims 1 to 13.

Images