TECHNICAL FIELD

[0001] The present disclosure relates to the field of Personalization, Advanced Driver Assistance System (ADAS) and Autonomous Driving (AD). In particular, the present disclosure provides a system with a unique feature which facilitates the ADAS or AD systems based personalized driving style mapped close to the driver natural driving style and a method thereof.

BACKGROUND

[0002] An Advanced Driver Assistance System (ADAS) and Autonomous Driving (AD) assists driver in driving safely, increases comfort of driving, road safety and more time for driver to focus on non-driving task. Therefore, ADAS/AD feature in vehicles is becoming increasingly important in terms of safety, comfort and free time for users. Moreover, number of ADAS or AD features are increasing day by day to provide the users more safe, comfortable and pleasant driving experience. However, most of the users have limited knowledge and proactive experience with AD functions. A survey shows that usage of AD features are much lower than the expectations, which could also be seen in graph 100 in FIG. 1 A. Automatic lane change (ALC) is one of the AD comfort feature which considers vehicle dynamic parameters for doing trajectory computation while during lane change. Stepwise functioning of the conventional ADAS systems for ALC is illustrated in FIG IB. As depicted in the FIG. IB, firstly at block 102, lane change initiation is triggered by driver of a vehicle implemented with the conventional ADAS system, then at block 104, adjacent lane status is checked. If it is found to be free at block 106, then at block 108, trajectory is computed based on speed, lane offset, and lane width. Otherwise, at block 118, lane change operation is stopped.

[0003] Further, at block 110, adjacent lane status is checked after the trajectory computation. If, at block 112, it is found to be free, then at block 114, lane change operation is completed along with the trajectory computation at the block 108. Else, at block 116, lane change operation is stopped. However, a user may not get expected perceived benefits and satisfaction while using AD features provided by the conventional ADAS systems as they might not suit the user’s driving style. Therefore, the AD features are required to be personalized in order to provide a better and comfortable experience to the user.

[0004] Various techniques have been implemented for personalization of the AD features, for instance, Patent document DE102014208311A1 discloses a Driver assistance system with an operating mode for fully automated vehicle guidance of a motor vehicle, in which a fully automated vehicle guidance is individualized by being adapted to the individual needs of a vehicle driver. In this system, the fully automated vehicle guidance takes place in dependence of a stored user profile, where a plurality of user profiles are deposited on the system and further by means of a selection device, one of the stored user profiles is selected, which is subsequently used for the fully automated vehicle guidance. The automated identification of a driver and associated user profile is selected based on the identified driver.

[0005] Patent document EP3750765A1 relates to methods, apparatuses and computer programs for generating a machine-learning model and for generating a control signal for operating a vehicle. The method for generating the machine-learning model comprises determining information about a driving behavior of a driver of the vehicle, and transforming the information about the driving behavior of the driver of the vehicle into a target function for the machine-learning model. The method also comprises generating the machine-learning model, which is trained using an imitation learning approach that is based on the target function.

[0006] Patent document FR3074123A1 relates to trained training models, methods, and apparatuses for evaluating a driving style of a driver of a moving road vehicle. Basically it is based on the determination of the different driving style variations of a driver of a road vehicle during a driving session by means of machine learning. For this, an automatic learning model could be created, which is trained to evaluate a recent driving style of a driver of a road vehicle. The model of trained automatic learning is obtained according to a two-phase automatic learning approach involving a non-supervised classification followed by a supervised classification.

[0007] While the cited references disclose various apparatuses and methods for facilitating personalization of the AD features, there is a possibility of providing more efficient solution that provides user-friendly personalization. OBJECTS OF THE PRESENT DISCLOSURE

[0008] A general object of the present disclosure is to provide a smart and efficient system and method for facilitating driving style based personalizable driver assistance for a vehicle.

[0009] An object of the present disclosure is to provide a system and method that provides high accuracy using Al model for personalization of driver style for Automatic Lane Change (ALC) AD function.

[0010] Another object of the present disclosure is to provide a system and method that is user-friendly and can be configured for L3/ L4/ L5 AD features.

SUMMARY

[0011] Aspects of the present disclosure relate to the field of Advanced Driver Assistance System (ADAS) and Autonomous Driving (AD). In particular, the present disclosure provides a driving style based personalizable driver assistance system and method thereof. The gap in adaption of AD functional usage can be reduced by making the driving experience personalized by a new and unique personalizable AD feature that could be also referred to as Drive Like Me feature.

[0012] An aspect of the present disclosure pertains to a driving style based personalizable driver assistance system for a vehicle. The proposed system includes an evaluating unit operatively coupled to a learning engine, where the evaluating unit comprises one or more processors coupled to a memory storing instructions executable by the one or more processors. The evaluating unit is configured to acquire first set of data associated with vehicular parameters and user parameters associated with the vehicle and obtain second set of data associated with lane change parameters provided by the user of the vehicle. The evaluating unit is further configured to extract, by filtering, at least one of the acquired vehicular parameters and user parameters and the obtained lane change parameters based on nature of trajectory to be covered. The evaluating unit is also configured to create, taking into consideration the extracted vehicular parameters, user parameters, and lane change parameters, one or more data clusters using dynamic time warping as a metric to separate the created one or more data clusters; and then build a profile for an specific driving style based on the created data clusters, wherein a classifier is configured to classify lane changes and behaviour of the user associated with corresponding profile.

[0013] In an aspect, the learning engine may be configured to generate distinct driving styles corresponding to the vehicular parameters and user parameters, and further, the system may be tested-and-trained with respect to distinct driving styles. Variability in the first and second set of data may be monitored for each of the driving styles, and correspondingly profile associated with each of the driving styles may be updated. Further, the evaluating unit may be configured to segregate the distinct driving styles into any of normal, defensive, and aggressive style.

[0014] In one aspect, the system may be configured to plot a plurality of trajectories taking into consideration the distinct driving styles, wherein each of the plurality of trajectories may include one or more lanes. In another aspect, the system may also be configured to generate mean trajectories corresponding to the created one or more data clusters, wherein each of the one or more data clusters pertain to distinct lane completion times and lateral distances covered during a lane change.

[0015] In an aspect, the learning unit may include Long Short Term Memory (LSTM) based neural network architecture. The learning unit may also be equipped with Artificial Intelligence (Al). In another aspect, the system may also be configured to create the one or more data clusters through K-means clustering technique, and data corresponding to the one or more created data clusters may be labelled. Further, the classifier may be updated, by the Al equipped learning engine, taking into consideration the one or more labelled data clusters.

[0016] In yet another aspect, the first set of data may include speed, acceleration, jerk, roll, pitch, yaw, car position and angle relative to lane center, distance ahead, threshold of time to collision (TTC), and number of surrounding vehicles; and wherein, the second set of data may include direction of lane change, time to touch a lane, and lane change completion time.

[0017] An aspect of the present disclosure pertains to a method for facilitating driving style based personalizable driver assistance for a vehicle. The proposed method includes the steps of acquiring, through an evaluating unit operatively coupled to a learning engine, a first set of data associated with vehicular parameters and user parameters associated with the vehicle; obtaining, through the evaluating unit, a second set of data associated with lane change parameters provided by the user of the vehicle; extracting, through the evaluating unit, by filtering, at least one of the acquired vehicular parameters and user parameters and the obtained lane change parameters based on nature of trajectory to be covered. The method also includes: creating, through the evaluating unit, one or more data clusters taking into consideration the extracted vehicular parameters, user parameters, and lane change parameters. The one or more created data clusters may be separated using dynamic time warping as a metric. The method further includes: building, through the evaluating unit, a profile for an specific driving style based on the created data clusters, wherein a classifier is configured to classify lane changes and behaviour of the user associated with corresponding profile.

[0018] In an aspect, the method may include the steps of testing-and-training the evaluating unit and the learning engine with respect to distinct driving styles, wherein, the method may include monitoring variability in the first and second set of data for each of the driving styles, and correspondingly updating profile associated with each of the driving styles; and further the method may include the step of segregating the distinct driving styles into any of normal, defensive, and aggressive style.

[0019] Various objects, features, aspects and advantages of the inventive subject matter will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which like numerals represent like components.

BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are included to provide a further understanding of the present disclosure, and are incorporated in and constitute a part of this specification. The drawings illustrate exemplary embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0021] FIG. 1A illustrates a graphical representation including various AD features associated with conventional ADAS systems.

[0022] FIG. IB illustrates a flow chart depicting stepwise functioning of the conventional ADAS systems.

[0023] FIG. 2 illustrates an exemplary network architecture of the proposed driving style based personalizable driver assistance system for a vehicle, in order to elaborate its overall working, in accordance with an embodiment of the present disclosure.

[0024] FIG. 3 illustrates a block diagram representing functional units of evaluating unit of the proposed system, in accordance with an embodiment of the present disclosure.

[0025] FIG. 4 illustrates an exemplary diagram depicting ALC logic and corresponding functioning of the proposed system, in accordance with an exemplary embodiment of the present disclosure.

[0026] FIG. 5A illustrates an exemplary diagram depicting steps involved in driving style based personalizable driver assistance, in accordance with an exemplary embodiment of the present disclosure. [0027] FIG. 5B illustrates an exemplary diagram depicting deployment of the Al model in driving style based personalizable driver assistance, in accordance with an exemplary embodiment of the present disclosure.

[0028] FIG. 6 illustrates an exemplary graphical representation depicting data plotted with respect to three modes of driving, in accordance with an exemplary embodiment of the present disclosure.

[0029] FIGs. 7A and 7B illustrate exemplary graphical representation depicting lane change trajectory data corresponding to distinct driving styles, in accordance with an exemplary embodiment of the present disclosure.

[0030] FIG. 8 illustrates an exemplary graphical representation depicting feature specific data clustering, in accordance with an exemplary embodiment of the present disclosure.

[0031] FIG. 9 illustrates an exemplary graphical representation depicting feature specific supervised Al model building, in accordance with an exemplary embodiment of the present disclosure.

[0032] FIG. 10 illustrates an exemplary graphical representation depicting personalized profiles and corresponding lane change applications, in accordance with an exemplary embodiment of the present disclosure.

[0033] FIG. 11 illustrates a flow diagram representing steps of the proposed method for facilitating driving style based personalizable driver assistance for a vehicle, in accordance with an embodiment of the present disclosure.

DETAILED DESCRIPTION

[0034] The following is a detailed description of embodiments of the disclosure depicted in the accompanying drawings. The embodiments are in such details as to clearly communicate the disclosure. However, the amount of detail offered is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosures as defined by the appended claims.

[0035] Embodiments explained herein relate to the field of Advanced Driver Assistance System (ADAS). In particular, the present disclosure provides a driving style based personalizable driver assistance system and method thereof.

[0036] Referring to FIG. 2, the proposed driving style based personalizable driver assistance system 200 (interchangeably, referred to as system 200, hereinafter) can be implemented within a vehicle, whereby the system 200 can be equipped with Artificial Intelligence (Al) that can identify a driving style, and can correspondingly personalizable driver assistance.

[0037] In an embodiment, the system 200 can include a monitoring unit 204, which can include one or more sensors, and can be configured to monitor real-time parameters associated with the vehicle, which may include vehicular parameters and user parameters. Further, the monitoring unit 204 can transmit a first set of data that may pertain to the parameters and user parameters associated with the vehicle.

[0038] In an exemplary embodiment, the vehicular parameters and user parameters such as, but not limited to, speed, acceleration, jerk, roll, pitch, yaw, car position and angle relative to lane center, distance ahead, threshold of time to collision (TTC), number of surrounding vehicles, time taken by the user in performing any of the above-mentioned tasks. In an exemplary embodiment, the one or more sensors can include, but not limited to, odometer, speedometer, accelerometer, gyroscope, proximity sensor, touch sensor, and navigation module 202 like GPS, GNSS, GLONASS, and the like.

[0039] In another embodiment, the system 200 can include an evaluating unit 206 that can be in communication with the monitoring unit 204, where the evaluating unit 206 can be configured through one or more processors coupled to a memory storing instructions executable by the one or more processor, Moreover, the evaluating unit 206 can be operatively coupled to a learning engine 208 that may include Long Short Term Memory (LSTM) based neural network architecture, and may also be equipped with Artificial Intelligence (Al).

[0040] In one embodiment, the evaluating unit 206 can acquire the first set of data being transmitted from the monitoring unit 204. In other embodiment, the evaluating unit 206 can obtain second set of data pertaining to lane change parameters provided by the user of the vehicle such as direction of lane change, related timestamp, and the like. Further, in an instance, the second set of data can be monitored and transmitted to the evaluating unit 206 by the monitoring unit 204. In another instance, the evaluating unit 206 can obtain the second set of data from a user device, such as smartphone, tablet, laptop, GUI-interface module, and the like, which may be operatively coupled to the system 200.

[0041] Further, the evaluating unit 206 can extract at least one of the acquired vehicular parameters and user parameters and the obtained lane change parameters. In an exemplary embodiment, the evaluating unit 206 can extract said parameters by filtering all the parameters and selecting relevant parameter, which may be selected taking into consideration nature of trajectory to be covered by the vehicle. In an exemplary embodiment, the nature of trajectory can be pre-determined by the evaluating unit 206 by communicating through the navigation module 202.

[0042] In an embodiment, the evaluating unit 206 can create one or more data clusters by taking into consideration the extracted vehicular parameters, user parameters, and lane change parameters. Further, the evaluating unit 206 can separate the created one or more data clusters using dynamic time warping as a metric.

[0043] In another embodiment, the evaluating unit 206 can build a profile for a specific driving style based on the created data clusters, wherein a classifier can be configured to classify lane changes and behaviour of the user associated with corresponding profile. In an implementation, the learning engine 208 can be configured to generate distinct driving styles corresponding to the vehicular parameters and user parameters.

[0044] According to an embodiment, the system 200 can include a display unit (not shown) that may be operatively coupled to the evaluating unit 206, and can be configured to display output obtained from the evaluating unit 206, such as driving style, lane changes, corresponding profile, and the like. Further, the display unit can be in form of a LED display board, LCD display board, GUI module, and the like. In another embodiment, a personal laptop, smartphone, tablet, or any other such mobile computing device can be coupled to the system 200, which can also act as the display unit.

[0045] According to an embodiment, the evaluating unit 206 can be in communication with the monitoring unit 204 and the display unit, through a network 210. Further, the network 210 can be a wireless network, a wired network or a combination thereof that can be implemented as one of the different types of networks, such as Intranet, Local Area Network (LAN), Wide Area Network (WAN), Internet, and the like.

[0046] Further, the network 210 can either be a dedicated network or a shared network. The shared network can represent an association of different types of networks that can use variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), Wireless Application Protocol (WAP), and the like.

[0047] In an embodiment, the system 200 can be implemented using any or a combination of hardware components and software components such as a cloud, a server 212, a computing system, a computing device, a network device and the like. Further, the evaluating unit 206 can interact with other components of the system 200, through a website or an application that can reside in the proposed system 200. In an implementation, the system 200 can be accessed by website or application that can be configured with any operating system, including but not limited to, Android™, iOS™, and the like.

[0048] Further, the proposed system 200 can be implemented in any vehicle, including but not limited to, car, truck, and bus.

[0049] Referring to FIG. 3, a block diagram 300 representing exemplary functional units of the evaluating unit 206 can include one or more processor(s) 302 that can be implemented as microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that manipulate data based on operational instructions. Among other capabilities, the processor(s) 302 are configured to fetch and execute computer-readable instructions stored in a memory 304 of the evaluating unit 206. The memory 304 can store one or more computer-readable instructions or routines, which may be fetched and executed to create or share the data units over a network service. The memory 304 can include any non-transitory storage device including, for example, volatile memory such as RAM, or non-volatile memory such as EPROM, flash memory, and the like.

[0050] In an embodiment, the evaluating unit 206 can also include an interface(s) 306. The interface(s) 306 may include a variety of interfaces, for example, interfaces for data input and output devices, referred to as I/O devices, storage devices, and the like. The interface(s) 306 may facilitate communication of the evaluating unit 206 with various devices coupled to the evaluating unit 206. The interface(s) 306 may also provide a communication pathway for one or more components of the evaluating unit 206. Examples of such components include, but are not limited to, existing processing engine(s) 308 and database 310.

[0051] In an embodiment, the processing engine(s) 308 can be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processing engine(s) 308. The database 310 can include data that is either stored or generated as a result of functionalities implemented by any of the components of the processing engine(s) 308. In an embodiment, the processing engine(s) 308 can include a data clustering unit 312, a profile building unit 314, and other unit(s) 316. The other unit(s) 316 can implement functionalities that supplement applications or functions performed by the evaluating unit 206 or the processing engine(s) 308.

[0052] According to an embodiment, the data clustering unit 312 can create one or more data clusters taking into consideration any or a combination of the extracted vehicular parameters, user parameters, and lane change parameters. The data clustering unit 312 can use dynamic time warping as a metric to separate the created one or more data clusters. Further, other unit(s) 316 along with the data clustering unit 312 can generate mean trajectories corresponding to the created one or more data clusters, wherein each of the data clusters can pertain to distinct lane completion times and lateral distances covered during a lane change.

[0053] In an implementation, the data clustering unit 312 can create the one or more data clusters through K-means clustering technique, and further data corresponding to the one or more created data clusters can be labelled. Furthermore, the classifier can be updated, by the Al equipped learning engine, taking into consideration the one or more labelled data clusters. [0054] In an implementation, the data clustering unit 312 can generate mean trajectories corresponding to the created one or more data clusters, wherein each of the one or more data clusters may pertain to distinct lane completion times and lateral distances covered during a lane change.

[0055] According to an embodiment, the profile building unit 314 can build a profile for an specific driving style based on the created data clusters, wherein the classifier can be configured to classify lane changes and behaviour of the user associated with corresponding profile. In an embodiment, the learning engine 208 can generate distinct driving styles corresponding to the vehicular parameters and user parameters; and further, the system 200 can be tested-and-trained with respect to distinct driving styles, where variability in the first and second set of data can be monitored for each of the driving styles, and correspondingly profile associated with each of the driving styles can be updated. Moreover, the profile building unit 314 can segregate the distinct driving styles into any of normal, defensive, and aggressive style. [0056] In one embodiment, a plurality of trajectories can be plotted taking into consideration the distinct driving styles, wherein each of the plurality of trajectories can include one or more lanes.

[0057] Therefore, the proposed system 200 facilitates personalization of AD system to driving style of a user within pre-defined safety margins as driving styles are already generated by the Al-equipped learning engine where user parameters (also, referred to as human factors, herein) are also considered during personalization. Moreover, the proposed idea may help to increase the usages of AD features by the users which can have direct positive impact on the sale and brand value of corresponding company.

[0058] Referring to FIG. 4, block diagram 400 represents ALC logic and corresponding functioning of the proposed system 200. In an embodiment, at block 402, vehicular parameters (also, referred to as vehicle dynamic parameters) can include speed, acceleration, jerk, roll, pitch, yaw, car position with respect to a lane, car angle relative with respect to curvature of the lane, steering angle, steering rate, distance from ahead vehicle, and the like. Further, relevant vehicle dynamic parameters can be fed to the evaluating 206 along with the learning engine 208 that includes supervised Al model, which can generate driving styles at block 404. Further, recommendation of the supervised Al model can be fed to ALC (Automatic Lane Change) module, which may run, at block 406, ALC algorithms on the recommendation of the supervised Al model, and thereby facilitating personalized action, i.e., based on a specific profile based on corresponding driving style.

[0059] Referring to FIG. 5A, block diagram 500 depicts steps involved in driving style based personalizable driver assistance. It involves five steps, wherein first step (Step 1) pertains to identification of required data used for AD feature specific personalization, second step (Step 2) pertains to system modification for collecting required data, and third step (Step 3) pertains to feature specific data extraction and featuring. Further, it involves a fourth step (Step 4) that pertains to feature specific data clustering, a fifth step (Step 5) that involves feature specific supervised Al model building and sixth (Step 6) that involves Deployment of Al model in the cloud and integration of personalizable AD features. These can be further elaborated.

• Step 1: Identification of required data used for AD feature specific personalization [0060] According to an embodiment, data related to vehicular, user, and lane change parameters can be obtained and then a survey can be carried out to check variability of the data corresponding to distinct driving styles. In an exemplary embodiment, statistical analysis approach has been used to identify the distinction, where said approach can include Min, Max, Mean, Standard Deviation (SD), and other such techniques.

[0061] In one example, parameters such as speed, acceleration, jerk, roll, pitch, yaw, car position relative to lane center, car angle relative to lane curvature, distance ahead (before and after lane change), TTC (before and after lane change), and number of surrounding vehicles (before and after lane change) can be appropriately found through the statistical analysis approach. Example of one such parameters (speed) is represented in FIG. 6, where graph 600 represents variation in speed when a test driver drove in 3 modes including Defensive, Aggressive and Normal.

• Step 2: System modification for collecting required data

[0062] In an embodiment, as the proposed system 200 attempts to personalize the ALC feature based on driving styles, hence logging of lane change related parameters is required. In an exemplary embodiment, system modifications can be done to log lane change parameters including, but not limited to, direction of lane change (left or right), time to touch the lane, and lane completion time. Values obtained for these lane change parameters may help to validate the model accuracy needed for personalizing the AD features.

[0063] In an example, a format of data logged with respect to (w.r.t.) lane change can be prepared taking into consideration parameters like time stamp, time to touch left lane, time to touch right lane, and lane change duration, where all the parameters are measured in seconds.

• Step 3: Feature specific data extraction and featuring

[0064] All the data related to lane change parameters and corresponding trajectory can be plotted by taking into consideration all types of driving styles, as illustrated in graphs 700 in FIG. 7 A and FIG. 7B. In an embodiment, data related to automatic lane change, as illustrated by black highlighted curves, can also be showcased to check the visual impression of all the manual lane changes w.r.t automatic one.

[0065] In another embodiment, manual lane change data can further be filtered based on nature of trajectory completion. The idea is to consider vehicle dynamic related data whenever there is lane change and use it for building the Al model.

[0066] In an exemplary embodiment, only left lane change related data can be filtered for further analysis purposes, which is illustrated in the FIG. 7B.

• Step 4: Feature specific data clustering

[0067] According to an embodiment, while creating data clusters, important lane change parameters (also, referred to as features relevant to lane change, herein) can be identified and selected for clustering, and corresponding data can be processed and filtered to remove outliers that correspond to data points noticeably different from rest of the data. The outliers may represent errors in measurement, bad data collection, or simply variables that are not considered when collecting the data.

[0068] In an exemplary embodiment, Time series KMeans clustering can be performed using dynamic time warping as a metric to separate the data clusters. The K-means clustering can be defined as a form of clustering where unlabeled data can be grouped into a number of groups, with the number of groups represented by the variable K.

[0069] In an embodiment, mean trajectories of the clusters can be observed through graph 800 of FIG. 8. In the FIG. 8, all the three clusters i.e. Cluster 1, Cluster 2, and Cluster 3 may represent mean trajectories of the clusters, where each cluster represents distinct lane completion times and lateral distances covered during lane change. In an example, average lane change completion time (in seconds) and lateral distance covered (in metres) could be jotted down for each of the three clusters, which may include supervised and unsupervised clusters along with real-time Automatic Lane Change (ALC).

• Step 5: Feature specific supervised Al model building

[0070] In an embodiment, the labelled data from the clustering can be used to build a classifier for driver lane changes. Further, for classifying time series data, Stacked LSTM can be used as the neural network architecture.

[0071] In another embodiment, only useful features relevant to lane change can be used for classification. This can be done using statistical analysis approach and corresponding tools. In an exemplary embodiment, plot 900 in FIG. 9 may illustrate high accuracy obtained in a test. In an example, clusters, both supervised as well as unsupervised along with ALC, can be taken into consideration as per time series analysis and then accuracy of prediction of a profile related to Al driver model corresponding to each cluster can be computed.

[0072] It can be observed that in the plot 900 that 89% driver model training accuracy can be achieved. Hence, following results prove that the model is able to classify the driver behavior. Further, the accuracy of the system can be further increased by properly tuning the model and increasing dataset size.

• Step 6: Al model deployment in cloud and integration of personizable Al feature [0073] In an embodiment, the Al equipped learning engine (also, referred to as Al model, herein) can be tested and trained with respect to the vehicular parameters and user parameters. In one embodiment, the personalizable AD features can be integrated into the Al model, which can be deployed in a cloud or server. In an implementation, as illustrated in FIG. 5B, the system 200 can be configured to collect data from a class of automatic vehicle (AV) community, at block 510, including various reference vehicles, and feed it into cloud based Al model, at block 514. Further, the system 200 may obtain a request pertaining to activation of personalizable features (also, referred to as ‘drive like me’ features, herein) from the build profile associated with a vehicle 512, and in response the system 200 may extract various vehicular perimeters and user parameters associated with the vehicle 512 and correspondingly, the system 200 may alter driving style settings.

[0074] In another exemplary embodiment, graph 1000 in FIG. 10 depicts ALC personalized profiles and corresponding lane change applications. The proposed system 200 involves training the Al model with vehicular parameters and user parameters. Moreover, the proposed system 200 can further be enhanced to train the Al model with only user parameters with corresponding embodiments being well within the scope of the invention. The separation of vehicular parameters and user parameters can help to segregate the AD personalization features which could be of great use to automotive-OEMs.

[0075] Referring to FIG. 11, the proposed method 1100 (also, referred to as method 1100, herein) facilitates driving style based personalizable driver assistance for a vehicle. In an embodiment, the method 1100 can include acquiring, at step 1102, through an evaluating unit operatively coupled to a learning engine, first set of data pertaining to vehicular parameters and user parameters associated with the vehicle; and obtaining, at step 1104, through the evaluating unit, second set of data pertaining to lane change parameters provided by the user of the vehicle. [0076] Further, the method 1100 can include extracting, at step 1106, through the evaluating unit, by filtering, at least one of the acquired vehicular parameters and user parameters and the obtained lane change parameters based on nature of trajectory to be covered [0077] Furthermore, the method 1100 can include creating, at step 1108, through the evaluating unit, one or more data clusters taking into consideration the extracted vehicular parameters, user parameters, and lane change parameters, wherein the one or more created data clusters are separated using dynamic time warping as a metric; and then building, at step 1110, through the evaluating unit, a profile for an specific driving style based on the created data clusters, wherein a classifier is configured to classify lane changes and behaviour of the user associated with corresponding profile.

[0078] In another embodiment, the method 1100 can include the steps of testing-and- training the evaluating unit and the learning engine with respect to distinct driving styles, monitoring variability in the first and second set of data for each of the driving styles, and correspondingly updating profile associated with each of the driving styles. The method 1100 can also include the step of segregating the distinct driving styles into any of normal, defensive, and aggressive style.

[0079] In yet another embodiment, the method 1100 can include the step of plotting a plurality of trajectories taking into consideration the distinct driving styles, wherein each of the plurality of trajectories includes one or more lanes. Therefore, the process of AD personalization for different driving styles using Al modelling can be established. Further, the proposed idea can be scaled up for real time implementation and for other AD functions.

[0080] In an embodiment, the proposed invention/ idea facilitates personalization of AD system and functions based on natural driving style of the driver with enhanced accuracy using Al model for personalization of driver style for ALC function. The process of AD personalization for different driving styles can be established using Al modelling. Further, the proposed idea can be scaled up for real time implementation and for other AD functions. Moreover, the proposed idea may help to increase the usages of AD features by the users which can have direct positive impact on the sale and brand value of corresponding company.

[0081] It should be appreciated that the proposed system and method can be implemented in a vehicle for personalization of Level 3 (L3) (conditional automation), L4 (high automation), L5 (full automation), and other such AD features through driving style classification using Al. [0082] While the foregoing describes various embodiments of the invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof. The scope of the invention is determined by the claims that follow. The invention is not limited to the described embodiments, versions or examples, which are included to enable a person having ordinary skill in the art to make and use the invention when combined with information and knowledge available to the person having ordinary skill in the art.

ADVANTAGES OF THE PRESENT DISCLOSURE

[0083] The present disclosure provides a smart and efficient system and method for facilitating personalization of automated driving systems and functions based on natural driving style of the driver.

[0084] The present disclosure provides a systematic method for data collection, data analysis, and using Al model driving style classification to map driver natural driving style, for personalization of driver style for AD function (illustrated for ALC function), and can be adopted for personalization for all relevant Level3/ Lelvel4/ Levels AD features.