Systems and methods for predicting driving behaviors of drivers by transforming trip data into image representation

Abstract

Method and system for predicting driving behaviors of a driver by transforming trip data into an image representation are disclosed. For example, the method includes receiving trip data of one or more trips of a driver, dividing the trip data into a plurality of trip data segments based on a predetermined time period, wherein each trip data segment corresponds to a portion of the one or more trips, transforming the plurality of trip data segments into the image representation, and determining predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model.

Description

SYSTEMS AND METHODS FOR PREDICTING DRIVINGBEHAVIORS OF DRIVERS BY TRANSFORMING TRIP DATA INTOIMAGE REPRESENTATIONCROSS REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to U.S. Patent Application No. 18 / 375,908, filed on October 2, 2023, which application is herewith incorporated by reference in its entirety.FIELD OF THE DISCLOSURE

[0002] Some embodiments of the present disclosure are directed to determining predicted driving behaviors of a driver by transforming trip data into an image representation. More particularly, certain embodiments of the present disclosure provide methods and systems for determining predicted driving behaviors of a driver by transforming at least a portion of trip data of the driver into the image representation to be used to train a predictive model. Merely by way of example, the present disclosure has been applied to transforming trip data of a driver into an image representation that represents at least a portion of the trip data, training a predictive model using the transformed image representation of the trip data, and determining predicted driving behaviors of the driver using the trained predictive model. But it would be recognized that the present disclosure has much broader range of applicability.BACKGROUND OF THE DISCLOSURE

[0003] Driving behaviors of drivers may be predicted based on trip data collected by one or more sensors of the mobile devices and / or vehicles. However, in some cases, the trip data may include enormous amount of data from a number of sensors that may not be easily interpretable or understandable by a predictive model to extract unique features that represent driving behaviors of a driver. The limitations of current computing resources (processors and memory) also lead to inability for predictive models to fully extract all the data to input and / or train the models. Hence it is highly desirable to develop more accurate techniques for transforming or translating trip data into a format that can be effectively and accurately interpreted and understood by a predictive model.BRIEF SUMMARY OF THE DISCLOSURE

[0004] Some embodiments of the present disclosure are directed to determining predicted driving behaviors of a driver by transforming trip data into an image representation. More particularly, certain embodiments of the present disclosure provide methods and systems for determining predicted driving behaviors of a driver by transforming at least a portion of trip data of the driver into an image representation to be used to train a predictive model. Merely by way of example, the present disclosure has been applied to transforming trip data of a driver into an image representation that represents at least a portion of the trip data, training a predictive model using the transformed image representation of the trip data, and determining predicted driving behaviors of the driver using the trained predictive model. But it would be recognized that the present disclosure has much broader range of applicability.

[0005] According to some embodiments, a method for predicting driving behaviors of a driver by transforming trip data into an image representation includes receiving trip data of one or more trips of a driver and dividing the trip data into a plurality of trip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the one or more trips. The method further includes transforming the plurality of trip data segments into the image representation, and determining predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model.

[0006] According to certain embodiments, a computing device for predicting driving behaviors of a driver by transforming trip data into an image representation includes a processor and a memory having a plurality of instructions stored thereon that, when executed by the processor. The instructions, when executed, cause the one or more processors to receive trip data of one or more trips of a driver and divide the trip data into a plurality of trip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the one or more trips. Also, the instructions, when executed, cause the one or more processors to transform the plurality of trip data segments into the image representation, and determine predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model.

[0007] According to some embodiments, a non-transitory computer-readable medium stores instructions for predicting driving behaviors of a driver by transforming trip data into an image representation. The instructions are executed by one or more processors of a computing device. The non-transitory computer-readable medium includes instructions rreceiving trip data of one or more trips of a driver and divide the trip data into a plurality oftrip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the one or more trips. Also, the non-transitory computer-readable medium includes instructions to transform the plurality of trip data segments into the image representation and determine predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model.

[0008] According to some embodiments, the invention may also relate to a computer- implemented method, performed by one or more processors, for predicting driving behaviors. The computer implemented method when executed on the one or more processors can perform: receiving a trip data of one or more trips of a driver from one or more sensors. The computer implemented method when executed on the one or more processors can also perform: dividing the trip data into one or more trip data segments based on a predetermined time period, each trip data segment corresponding to at least a portion of the one or more trips. The computer implemented method when executed on the one or more processors, additionally can perform: transforming the one or more trip data segments into a corresponding image representation. The computer implemented method when executed on the one or more processors can also perform: predicting driving behaviors of the driver based on the corresponding image representation of the one or more trips using a prediction model.

[0009] According to some embodiments, a non-transitory computer-readable medium storing instructions for predicting driving behaviors of a driver, wherein instructions, when executed by one or more processors of a computing device, cause the computing device to: receive trip data of one or more trips of the driver from one or more sensors. A non-transitory computer-readable medium storing instructions for predicting driving behaviors of a driver, wherein instructions, when executed by one or more processors of a computing device, also cause the computing device to: divide the trip data into one or more trip data segments based on a predetermined time period, each trip data segment corresponding to at least a portion of the one or more trips. A non-transitory computer-readable medium storing instructions for predicting driving behaviors of a driver, wherein instructions, when executed by one or more processors of a computing device, further cause the computing device to: transform the one or more trip data segments into a corresponding image representation. A non-transitory computer-readable medium storing instructions for predicting driving behaviors of a driver, wherein instructions, when executed by one or more processors of a computing device, additionally cause the computing device to: predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using a prediction model.

[0010] According to some embodiments, the non-transitory computer-readable medium storing instructions for predicting the driving behaviors, wherein transforming the reference trip data into the training image representation comprises to: divide, for each reference trip of the reference trips, the reference trip data into a one or more reference trip data segments based on the predetermined time period, each reference trip data segment corresponding to at least the portion of the reference trips. The non-transitory computer-readable medium storing instructions for predicting the driving behaviors, wherein transforming the reference trip data into the training image representation comprises to: generate, for each reference trip data segment of the one or more reference trip data segments, the graphical representation representing relative positions of a corresponding driver during the predetermined time period by extracting the location information from a corresponding reference trip data segment. The non-transitory computer-readable medium storing instructions for predicting the driving behaviors, wherein transforming the reference trip data into the training image representation comprises to: add the depth to each point of the graphical representation, wherein the depth includes the one or more channels that represent the sensor data associated with the corresponding reference trip data segment. The non-transitory computer-readable medium storing instructions for predicting the driving behaviors, wherein transforming the reference trip data into the training image representation comprises to: generate the corresponding image representation for each reference trip data segment of the one or more reference trip data segments, wherein the training image representation comprising the n- dimensional graphical representation with the n number of the sensor data associated with each reference trip data segment of the one or more reference trip data segments.

[0011] Depending upon the embodiment, one or more benefits may be achieved. These benefits and various additional objects, features and advantages of the present disclosure can be fully appreciated with reference to the detailed description and accompanying drawings that follow.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a simplified diagram showing a method for determining predicted driving behaviors of a driver based on trip data of the driver using a prediction model according to certain embodiments of the present disclosure.

[0013] Figures 2 is a simplified diagram showing a method for training a prediction model for predicting driving behaviors of a driver according to certain embodiments of the present disclosure.

[0014] Figure 3 is an exemplary diagram illustrating transformation of trip data into graphical representations of an image representation of trip data according to certain embodiments of the present disclosure.

[0015] Figure 4 is an exemplary diagram illustrating using an image representation of the trip data to predict driving behavior of a driver using a trained prediction model according to certain embodiments of the present disclosure.

[0016] Figure 5 is a simplified diagram showing a system for determining predicted driving behaviors of a driver based on trip data of the driver using a prediction model according to certain embodiments of the present disclosure.DETAILED DESCRIPTION OF THE DISCLOSURE

[0017] Some embodiments of the present disclosure are directed to determining predicted driving behaviors of a driver by transforming trip data into an image representation. As used herein, trip data may include various sensors (e.g., vehicle, mobile devices, external imaging devices) that collect data such as telematics, vehicle data from OBD ports, ADAS data, images, contextual data like weather and traffic and the like. More particularly, certain embodiments of the present disclosure provide methods and systems for determining predicted driving behaviors of a driver by transforming at least a portion of trip data of the driver into the image representation to be used to train a predictive model. Merely by way of example, the present disclosure has been applied to transforming trip data of a driver into the image representation that represents at least a portion of the trip data, training a predictive model using the transformed image representation of the trip data, and determining predicted driving behaviors of the driver using the trained predictive model. But it would be recognized that the present disclosure has much broader range of applicability.I. ONE OR MORE METHODS FOR DETERMINING PREDICTED DRIVING BEHAVIORS OF A DRIVER BASED ON TRIP DATA OF THE DRIVERUSING A PREDICTION MODEL ACCORDING TO CERTAINEMBODIMENTS

[0018] Figure 1 is a simplified diagram showing a method 100 for determining predicted driving behaviors of a driver based on trip data of the driver using a prediction model according to certain embodiments of the present disclosure. One of ordinary skill in the art would recognize many variations, alternatives, and modifications. In some embodiments, the method 100 is performed by a computing device (e.g., a server 506).

[0019] The method 100 includes process 102 for receiving trip data of one or more trips of a driver, process 104 for dividing the trip data into a plurality of trip data segments based on a predetermined time period, process 106 for transforming each trip data segment into an image representation, and process 114 for determining predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model.

[0020] Specifically, at the process 102, the trip data includes telematics data, vehicle data, ADAS data, images, contextual data associated with the one or more trips. The telematics data is collected during one or more trips of a driver and indicates driving behaviors of the driver during the one or more trips. As an example, the driving behavior represents a manner in which the driver has operated a vehicle, such as the driver’s driving habits and / or driving patterns. The telematics data may be collected from one or more sensors associated with a vehicle and / or a driver’s computing device such as a mobile device. For example, the one or more sensors include any type and number of accelerometers, gyroscopes, magnetometers, location sensors (e.g., GPS sensors), and / or any other suitable sensors that measure the state and / or movement of the vehicle and / or the mobile device. The computing device as used herein may be a wearable computing device such as a ring, watch, glasses, bracelet, and the like. In certain embodiments, the telematics data may be collected continuously or at predetermined time intervals.

[0021] In the illustrative embodiment, the trip data may further include context data. For example, the context data includes road data, driver data, and / or world data. The road data associated with the one or more trips includes information about one or more roads taken during the one or more trips. For example, the road data includes a type of the road (e.g., highway, freeway, toll, local, or parking lot), a road map (e.g., curvature, incline, gradient, elevation, direction, and / or a number of lanes), and / or road conditions (e.g., road moisture, traffic). The driver data associated with the one or more trips of a driver includes any sociodemographic information of the driver. For example, the driver data includes age, race,ethnicity, gender, marital status, income, education, employment, and / or credit score. The world data associated with the one or more trips includes an indication whether the one or more trips was taken on a holiday, a weather condition during the one or more trips, and / or an indication of when the one or more trips was taken (e.g., time of day, day of week, day of month, and / or month of year).

[0022] At the process 104, each trip data segment corresponds to a portion of the one or more trips for the predetermined time period. For example, each trip may be divided into a plurality of trip data segments every 5 minutes. According to some embodiments, the trip data collected during a first predetermined time period of each trip may be selected and further divided into a second predetermined time period. For example, first 15 minute of trip data is selected for each trip and is then divided into three 5-minute segments.

[0023] According to some embodiments, each trip data segment may correspond to a portion of the one or more trips for a predetermined distance. For example, each trip may be divided into a plurality of trip data segments every 5 miles. According to some embodiments, the trip data collected during a first predetermined distance of each trip may be selected and further divided into a second predetermined distance. For example, first 20 miles of trip data is selected for each trip and is then divided into four-5 mile segments.

[0024] At the process 106, each trip data segment is transformed into an image representation. For example, the image representation is a high-depth image-like tensor. To do so, at process 108, for each trip data segment, a graphical representation representing relative positions (e.g., latitude and longitude) of the driver during the predetermined time period is generated by extracting location information from the corresponding trip data segment. In other words, the relative longitude and latitude coordinates are used as indexes to generate the graphical representation as illustrated in Figure 3.

[0025] Once the graphical representation is generated, at process 110, depth to each point of the graphical representation is added. For example, the depth includes one or more channels that represent sensor data. As described above, the sensor data may be collected from one or more sensors associated with a vehicle and / or a driver’s mobile device. For example, the one or more sensors include any type and number of accelerometers, gyroscopes, magnetometers, location sensors (e.g., GPS sensors), and / or any other suitable sensors that measure the state and / or movement of the vehicle and / or the mobile device. For example, the sensor data may include speed, acceleration, braking, orientation and angular velocity, direction, heading, location, idling time, and / or fuel consumption. It should be appreciated that, according to certain embodiments, the one or more channels may include atime channel to, for example, indicate a chronological sequence of the graphical representations.

[0026] At process 1 12, an image representation of each trip data segment is generated. For example, each image representation is a n-dimensional graphical representation. As described above, according to some embodiments, each image representation maps relative positions of the driver during the predetermined time period. For example, location data of the trip data is used to plot relative longitude and latitude coordinates of the driver for each predetermined time period. In other words, the relative longitude and latitude coordinates are used as indexes to generate the graphical representation. For each point of the graphical representation, n-number of sensor data is added to generate a n-dimensional graphical representation of the corresponding trip data segment.

[0027] At the process 114, a prediction model is used to predict driving behaviors of the driver based on the image representation of the one or more trips. Specifically, the prediction model is a predictive model that is trained to extract features from the image representation of data segments associated with the one or more trips of the driver. Extracted features are indicative of one or more driving behaviors of the driver. For example, the extracted features may include sudden acceleration or braking, frequent braking, sharp cornering, and / or slow cornering. According to some embodiments, the driving behaviors of the driver are predicted based on the extracted features. For example, driving behaviors of the driver in a next trip are predicted based on one or more features extracted from the trip data of one or more trips taken by the driver using the prediction model. By being able to have or use the extracted features, a more accurate prediction can be done using a predictive model then with simple telematics data.

[0028] According to some embodiments, the image representation of the one or more trips may be further divided into smaller patches using a patchify algorithm. For example, the image representation may be 20x20 pixel image, which can be divided into 1000 square patches of 2x2 pixel each. These smaller patches may be inputted into the prediction model. The prediction model may be any predictive model that is trained using a deep learning algorithm, such as a convolution neural network (CNN).II. ONE OR MORE SYSTEMS FOR TRAINING A PREDICTION MODEL ACCORDING TO CERTAIN EMBODIMENTS

[0029] Figure 2 is a simplified diagram showing a method 200 for training a prediction model for predicting driving behaviors of a driver according to certain embodiments of the present disclosure. This diagram is merely an example, which should not unduly limit the scope of the claims. One of ordinary skill in the art would recognize many variations, alternatives, and modifications. In the illustrative embodiment, the method 200 is performed by a computing device (e.g., a server 506).

[0030] The method 200 includes process 202 for receiving trip data of reference trips of a plurality of drivers, process 204 for transforming the trip data of the reference trips into a training image representation, and process 214 for training a prediction model using the training image representation of the reference trips. The plurality of drivers may be similar situated group of drivers such as having one or more similar occupation, locality, vehicles, age, education, driving record, and the like.

[0031] Specifically, at the process 202, the trip data includes telematics data, vehicle data, ADAS data, images, contextual data associated with the one or more reference trips. The telematics data is collected during one or more reference trips of the plurality of drivers and indicates driving behaviors of each driver during one or more reference trips taken by the corresponding driver. As an example, the driving behavior represents a manner in which the corresponding driver has operated a vehicle such as the driver’s driving habits and / or driving patterns. The telematics data may be collected from one or more sensors associated with a vehicle and / or a driver’s mobile device. For example, the one or more sensors include any type and number of accelerometers, gyroscopes, magnetometers, location sensors (e.g., GPS sensors), and / or any other suitable sensors that measure the state and / or movement of the vehicle and / or the mobile device. In certain embodiments, the telematics data may be collected continuously or at predetermined time intervals.

[0032] In the illustrative embodiment, the trip data may further include context data. For example, the context data includes road data, driver data, and / or world data. The road data associated with the one or more reference trips includes information about one or more roads taken during the one or more reference trips. For example, the road data includes a type of the road (e.g., highway, freeway, toll, local, or parking lot), a road map (e.g., curvature, incline, gradient, elevation, direction, and / or a number of lanes), and / or road conditions (e.g., road moisture, traffic). The driver data associated with the one or more reference trips of adriver includes any socio-demographic information of the driver. For example, the driver data includes age, race, ethnicity, gender, marital status, income, education, employment, and / or credit score. The world data associated with the one or more reference trips includes an indication whether the one or more reference trips was taken on a holiday, a weather condition during the one or more reference trips, and / or an indication of when the one or more reference trips was taken (e.g., time of day, day of week, day of month, and / or month of year).

[0033] At the process 204, training image representation are generated using the trip data of the reference trips. For example, the image representation is a high-depth image-like tensor. To do so, at process 206, for each reference trip, the trip data is divided into a plurality of trip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the corresponding reference trip for the predetermined time period. For example, each reference trip may be divided into a plurality of trip data segments every 5 minutes. According to some embodiments, the trip data collected during a first predetermined time period of each reference trip may be selected and further divided into a second predetermined time period. For example, first 15 minute of trip data is selected for each reference trip and is then divided into three 5-minute segments.

[0034] According to some embodiments, each trip data segment may correspond to a portion of the one or more trips for a predetermined distance. For example, each trip may be divided into a plurality of trip data segments every 5 miles. According to some embodiments, the trip data collected during a first predetermined distance of each trip may be selected and further divided into a second predetermined distance. For example, first 20 miles of trip data is selected for each trip and is then divided into four-5 mile segments.

[0035] At process 208, for each trip data segment, a graphical representation representing relative positions (e.g., latitude and longitude) of the corresponding driver during the predetermined time period is generated by extracting location information from the corresponding trip data segment. In other words, the relative longitude and latitude coordinates are used as indexes to generate the graphical representation as illustrated in Figure 4.

[0036] Once the graphical representation is generated, at process 210, depth to each point of the graphical representation is added. For example, the depth includes one or more channels that represent sensor data. As described above, the sensor data may be collected from one or more sensors associated with a vehicle and / or a driver’s mobile device. For example, the one or more sensors include any type and number of accelerometers,gyroscopes, magnetometers, location sensors (e.g., GPS sensors), and / or any other suitable sensors that measure the state and / or movement of the vehicle and / or the mobile device. For example, the sensor data may include speed, acceleration, braking, orientation and angular velocity, direction, heading, location, idling time, and / or fuel consumption. It should be appreciated that, according to certain embodiments, the one or more channels may include a time channel to, for example, indicate a chronological sequence of the graphical representations.

[0037] At process 212, an image representation of each trip data segment is generated. For example, each image representation is a n-dimensional graphical representation. As described above, according to some embodiments, each image representation maps relative positions of the corresponding driver during the predetermined time period. For example, location data of the trip data is used to plot relative longitude and latitude coordinates of the corresponding driver for each predetermined time period. In other words, the relative longitude and latitude coordinates are used as indexes to generate the graphical representation. For each point of the graphical representation, n-number of sensor data is added to generate a n-dimensional graphical representation of the corresponding trip data segment. According to some embodiments, the processes 206-212 repeat until all the trip data segments of the reference trips are transformed into the image representation, which are used to train a prediction model. In another embodiment, the prediction model may be a generative artificial intelligence (Al) model.

[0038] At the process 214, the image representation transformed from the trip data of reference trips are used as training data to train the prediction model for predicting driving behaviors of a driver. For example, the prediction model may be trained using a deep learning algorithm, such as a convolution neural network (CNN). For example, the prediction model may be an auto-regressive model or a causal auto-regressive model that predicts future behavior based on past behavior data. According to some embodiments, each training image representation may be further divided into smaller patches (e.g., 420) prior to training the prediction model, as illustrated in Figure 4. For example, the training image representation may be 20x20 pixel image, which can be divided into 1000 square patches of 2x2 pixel each.

[0039] Specifically, the prediction model is trained to extract features from image representation of data segments of a driver. Extracted features are indicative of one or more driving behaviors of the corresponding driver. For example, the extracted features may include sudden acceleration or braking, frequent braking, sharp cornering, and / or slowcomering. According to some embodiments, the driving behaviors of the corresponding driver are predicted based on the extracted features. For example, driving behaviors of the corresponding driver in a next trip (e.g., a next patch or a next trip data segment) are predicted based on one or more features extracted from the trip data of one or more trips taken by the corresponding driver using the prediction model.

[0040] Although the above has been shown using a selected group of processes for the method, there can be many alternatives, modifications, and variations. For example, some of the processes may be expanded and / or combined. Other processes may be inserted to those noted above. Depending upon the embodiment, the sequence of processes may be interchanged with others replaced. For example, although the methods 100 and 200 are described as performed by the computing device above, some or all processes of the method are performed by any computing device or a processor directed by instructions stored in memory. As an example, some or all processes of the method are performed according to instructions stored in a non-transitory computer-readable medium.

[0041] Figure 3 is an exemplary diagram 300 illustrating transformation of trip data into graphical representations of image representation of trip data according to certain embodiments of the present disclosure. As described above in Figures 1 and 2, trip data may be transformed into an image representation that can be interpreted and understood by a prediction model. To do so, the trip data is divided into a plurality of trip data segments based on a predetermined time period or predetermined distance. For example, each reference trip may be divided into a plurality of trip data segments every 5 minutes. According to some embodiments, the trip data collected during a first predetermined time period of each reference trip may be selected and further divided into a second predetermined time period. For example, first 15 minute of trip data is selected for each reference trip and is then divided into three 5-minute segments. Additionally, each image representation is a n- dimensional graphical representation.

[0042] As described above, according to some embodiments, each image representation maps relative positions of the driver during the predetermined time period. For example, location data of the trip data is used to plot relative longitude and latitude coordinates of the driver for each predetermined time period, as illustrated in the graphical representations 310. In other words, each graphical representation 310 corresponds to relative positions of the driver during the predetermine time period. Since the relative positions are not absolute positional data, the relative longitude and latitude coordinates of the graphical representations 310 can be used as indexes (e.g., 0 to N) to generate the graphical representations 320. Foreach point of the graphical representation 320, n-number of sensor data is added to generate a n-dimensional graphical representation of the corresponding trip data segment. It should be noted that the graphical representations 310 and 320 are not drawn to scale.

[0043] Figure 4 is an exemplary diagram 400 illustrating using image representation of the trip data to predict driving behavior of a driver using a trained prediction model according to certain embodiments of the present disclosure. Similar to the graphical representation 310, each graphical representation 410 represents relative longitude and latitude coordinates of the driver during a predetermined time period by extracting location data of the trip data. Since the relative positions are not absolute positional data, the relative longitude and latitude coordinates of the graphical representations 410 can be used as indexes (e.g., 0 to N) to generate the graphical representations 430, which may be fed into a prediction model for predicting driving behaviors of a driver. As described above, the prediction model is trained using a deep learning algorithm, such as a convolution neural network (CNN). In some embodiments, the graphical representations 410 may be further divided into smaller patches 420 using a patchify algorithm. It should be appreciated that for each point of the graphical representation 410, n-number of sensor data may be added to generate a n-dimensional graphical representation of the corresponding trip data segment, prior to dividing into smaller patches 420. In other words, the smaller patches 420 are also n-dimensional graphical representation of a corresponding portion of the trip data segment. The smaller patches 420 may then be fed into the prediction model for predicting driving behaviors of a driver. It should be noted that the graphical representations 410, 420, 430 are not drawn to scale.

[0044] Figure 5 is a simplified diagram showing a system for determining predicted driving behaviors of a driver according to certain embodiments of the present disclosure. This diagram is merely an example, which should not unduly limit the scope of the claims. One of ordinary skill in the art would recognize many variations, alternatives, and modifications. In the illustrative embodiment, the system 500 includes a computing device 502, a network 504, and a server 506. Although the above has been shown using a selected group of components for the system, there can be many alternatives, modifications, and variations. For example, some of the components may be expanded and / or combined. Other components may be inserted to those noted above. Depending upon the embodiment, the arrangement of components may be interchanged with others replaced.

[0045] In various embodiments, the system 500 is used to implement the method 100 the method 200, the method 300, and / or the method 400. According to certain embodiments, the mobile device 502 is communicatively coupled to the server 506 via the network 504. Thecomputing device 502 may be a mobile device or a vehicle system. As an example, the mobile device 502 includes one or more processors 516 (e.g., a central processing unit (CPU), a graphics processing unit (GPU)), a memory 518 (e.g., random-access memory (RAM), read-only memory (ROM), flash memory), a communications unit 520 (e.g., a network transceiver), a display unit 522 (e.g., a touchscreen), and one or more sensors 524 (e.g., an accelerometer, a gyroscope, a magnetometer, a location sensor). For example, the one or more sensors 524 are configured to generate sensor data. According to some embodiments, the data are collected continuously, at predetermined time intervals, and / or based on a triggering event (e.g., when each sensor has acquired a threshold amount of sensor measurements).

[0046] In some embodiments, the mobile device 502 is operated by the user. For example, the user installs an application associated with an insurer on the mobile device 502 and allows the application to communicate with the one or more sensors 524 to collect sensor data. According to some embodiments, the application collects the sensor data continuously, at predetermined time intervals, at predetermined distance, and / or based on a triggering event (e.g., when each sensor has acquired a threshold amount of sensor measurements). In certain embodiments, the sensor data represents the user’s activity / behavior, such as the user driving behavior, in the method 100, the method 200, the method 300, and / or the method 400.

[0047] According to certain embodiments, the collected data are stored in the memory 518 before being transmitted to the server 506 using the communications unit 522 via the network 504 (e.g., via a local area network (LAN), a wide area network (WAN), the Internet). In some embodiments, the collected data are transmitted directly to the server 506 via the network 5 04. In certain embodiments, the collected data are transmitted to the server 506 via a third party. For example, a data monitoring system stores any and all data collected by the one or more sensors 524 and transmits those data to the server 506 via the network 504 or a different network.

[0048] According to certain embodiments, the server 506 includes a processor 530 (e.g., a microprocessor, a microcontroller), a memory 532, a communications unit 534 (e.g., a network transceiver), and a data storage 536 (e.g., one or more databases). In some embodiments, the server 506 is a single server, while in certain embodiments, the server 506 includes a plurality of servers with distributed processing. As an example, in Figure 5, the data storage 536 is shown to be part of the server 506. In some embodiments, the data storage 536 is a separate entity coupled to the server 506 via a network such as the network 504. In certain embodiments, the server 506 includes various software applications stored inthe memory 532 and executable by the processor 530. For example, these software applications include specific programs, routines, or scripts for performing functions associated with the method 100, the method 200, the method 300, and / or the method 400. As an example, the software applications include general-purpose software applications for data processing, network communication, database management, web server operation, and / or other functions typically performed by a server.

[0049] According to various embodiments, the server 506 receives, via the network 504, the sensor data collected by the one or more sensors 524 from the application using the communications unit 534 and stores the data in the data storage 536. For example, the server 506 then processes the data to perform one or more processes of the method 100, one or more processes of the method 200, one or more processes of the method 300, and / or one or more processes of the method 400.

[0050] According to certain embodiments, the notification in response to the application being determined not working properly in the method 100, the method 200, the method 300, and / or the method 400 is transmitted back to the mobile device 502, via the network 504, to be provided (e.g., displayed) to the user via the display unit 522.

[0051] In some embodiments, one or more processes of the method 100, one or more processes of the method 200, one or more processes of the method 300, and / or one or more processes of the method 400 are performed by the mobile device 502. For example, the processor 516 of the mobile device 502 processes the data collected by the one or more sensors 524 to perform one or more processes of the method 100, one or more processes of the method 200, one or more processes of the method 300, and / or one or more processes of the method 400.III. EXAMPLES OF CERTAIN EMBODIMENTS OF THE PRESENT DISCLOSURE

[0052] According to some embodiments, a method for predicting driving behaviors of a driver by transforming trip data into an image representation includes receiving trip data of one or more trips of a driver and dividing the trip data into a plurality of trip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the one or more trips. The method further includes transforming the plurality of trip data segments into an image representation, and determining predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model.For example, the method is implemented according to at least Figure 1, Figure 2, Figure 3, and / or Figure 4.

[0053] According to certain embodiments, a computing device for predicting driving behaviors of a driver by transforming trip data into an image representation includes a processor and a memory having a plurality of instructions stored thereon that, when executed by the processor. The instructions, when executed, cause the one or more processors to receive trip data of one or more trips of a driver and divide the trip data into a plurality of trip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the one or more trips. Also, the instructions, when executed, cause the one or more processors to transform the plurality of trip data segments into an image representation, and determine predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model. For example, the computing device is implemented according to at least Figure 5.

[0054] According to some embodiments, a non-transitory computer-readable medium stores instructions for predicting driving behaviors of a driver by transforming trip data into an image representation. The instructions are executed by one or more processors of a computing device. The non-transitory computer-readable medium includes instructions rreceive trip data of one or more trips of a driver, and divide the trip data into a plurality of trip data segments based on a predetermined time period. Each trip data segment corresponds to a portion of the one or more trips. Also, the non-transitory computer-readable medium includes instructions to transform the plurality of trip data segments into an image representation, and determine predicted driving behaviors of the driver based on the image representation of the one or more trips using a prediction model. For example, the non- transitory computer-readable medium is implemented according to at least Figure 1 , Figure 2, Figure 3, and / or Figure 4.IV. EXAMPLES OF MACHINE LEARNING ACCORDING TO CERTAIN EMBODIMENTS

[0055] According to some embodiments, a processor or a processing element may be trained using supervised machine learning, unsupervised machine learning, and the machine learning may employ an artificial neural network, which, for example, may be a convolutional neural network (CNN), a recurrent neural network, a deep learning neural network, a reinforcement learning module or program, or a combined learning module orprogram that learns in two or more fields or areas of interest. Machine learning may involve identifying and recognizing patterns in existing data in order to facilitate making predictions for subsequent data. Models may be created based upon example inputs in order to make valid and reliable predictions for novel inputs.

[0056] According to certain embodiments, machine learning programs may be trained by inputting sample data sets or certain data into the programs, such as images, object statistics and information, historical estimates, and / or actual repair costs. The machine learning programs may utilize deep learning algorithms that may be primarily focused on pattern recognition and may be trained after processing multiple examples. The machine learning programs may include Bayesian Program Learning (BPL), voice recognition and synthesis, image or object recognition, optical character recognition, and / or natural language processing. The machine learning programs may also include natural language processing, semantic analysis, automatic reasoning, and / or other types of machine learning.

[0057] According to some embodiments, supervised machine learning techniques, unsupervised machine learning techniques, and / or self-supervised machine learning techniques may be used. In supervised machine learning, a processing element may be provided with example inputs and their associated outputs and may seek to discover a general rule that maps inputs to outputs, so that when subsequent novel inputs are provided the processing element may, based upon the discovered rule, accurately predict the correct output. In unsupervised machine learning, the processing element may need to find its own structure in unlabeled example inputs. Similar to the unsupervised machine learning, in selfsupervised machine learning, the processing element may need to find its own structure in unlabeled example inputs. However, the self-supervised machine learning has a lot of supervisory signals that may act as feedback in the training process.V. ADDITIONAL CONSIDERATIONS ACCORDING TO CERTAIN EMBODIMENTS

[0058] For example, some or all components of various embodiments of the present disclosure each are, individually and / or in combination with at least another component, implemented using one or more software components, one or more hardware components, and / or one or more combinations of software and hardware components. As an example, some or all components of various embodiments of the present disclosure each are, individually and / or in combination with at least another component, implemented in one ormore circuits, such as one or more analog circuits and / or one or more digital circuits. For example, while the embodiments described above refer to particular features, the scope of the present disclosure also includes embodiments having different combinations of features and embodiments that do not include all of the described features. As an example, various embodiments and / or examples of the present disclosure can be combined.

[0059] Additionally, the methods and systems described herein may be implemented on many different types of processing devices by program code comprising program instructions that are executable by the device processing subsystem. The software program instructions may include source code, object code, machine code, or any other stored data that is operable to cause a processing system to perform the methods and operations described herein. Certain implementations may also be used, however, such as firmware or even appropriately designed hardware configured to perform the methods and systems described herein.

[0060] The systems’ and methods’ data (e.g., associations, mappings, data input, data output, intermediate data results, final data results) may be stored and implemented in one or more different types of computer-implemented data stores, such as different types of storage devices and programming constructs (e.g., RAM, ROM, EEPROM, Flash memory, flat files, databases, programming data structures, programming variables, IF-THEN (or similar type) statement constructs, application programming interface). It is noted that data structures describe formats for use in organizing and storing data in databases, programs, memory, or other computer-readable media for use by a computer program.

[0061] The systems and methods may be provided on many different types of computer- readable media including computer storage mechanisms (e.g., CD-ROM, diskette, RAM, flash memory, computer’s hard drive, DVD) that contain instructions (e.g., software) for use in execution by a processor to perform the methods’ operations and implement the systems described herein. The computer components, software modules, functions, data stores and data structures described herein may be connected directly or indirectly to each other in order to allow the flow of data needed for their operations. It is also noted that a module or processor includes a unit of code that performs a software operation, and can be implemented for example as a subroutine unit of code, or as a software function unit of code, or as an object (as in an object-oriented paradigm), or as an applet, or in a computer script language, or as another type of computer code. The software components and / or functionality may be located on a single computer or distributed across multiple computers depending upon the situation at hand.

[0062] The computing system can include mobile devices and servers. A mobile device and server are generally remote from each other and typically interact through a communication network. The relationship of mobile device and server arises by virtue of computer programs running on the respective computers and having a mobile device-server relationship to each other.

[0063] This specification contains many specifics for particular embodiments. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations, one or more features from a combination can in some cases be removed from the combination, and a combination may, for example, be directed to a subcombination or variation of a subcombination.

[0064] Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0065] Although specific embodiments of the present disclosure have been described, it will be understood by those of skill in the art that there are other embodiments that are equivalent to the described embodiments. Accordingly, it is to be understood that the present disclosure is not to be limited by the specific illustrated embodiments.

Claims

CLAIMSWhat is claimed is:

1. A computer-implemented method, performed by one or more processors, for predicting driving behaviors of a driver, comprising: receiving trip data of one or more trips of the driver from one or more sensors; dividing the trip data into one or more trip data segments based on a predetermined time period, each trip data segment corresponding to at least a portion of the one or more trips; transforming the one or more trip data segments into a corresponding image representation; and predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using a prediction model.

2. The method of claim 1, wherein transforming the one or more trip data segments into the corresponding image representation comprises: generating, for each trip data segment of the one or more trip data segments, a graphical representation representing relative positions of a corresponding driver during the predetermined time period by extracting location information from a corresponding trip data segment; adding depth to each point of the graphical representation; and generating the corresponding image representation for each trip data segment of the one or more trip data segments.

3. The method of any one of claims 1 or 2, wherein the depth includes one or more channels that represent sensor data, and the corresponding image representation comprises an n-dimensional graphical representation with an n number of the sensor data associated with each trip data segment of the one or more trip data segments, and wherein the one or more channels include at least a time channel to indicate a chronological sequence of the corresponding image representation.

4. The method of any one of claims 1-3, wherein predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using the prediction model comprises:dividing the corresponding image representation into patches; and predicting the driving behaviors of the driver by putting the patches of the corresponding image representation into the prediction model.

5. The method of any one of claims 1-4, wherein predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using the prediction model comprises: extracting one or more features from the corresponding image representation of the one or more trips; and predicting the driving behaviors of the driver based at least in part upon the one or more features that have been extracted.

6. The method of claim 5, wherein the one or more features that have been extracted, includes at least one of average speed, average braking, average cornering, or average acceleration.

7. The method of claim 1, further comprising training the prediction model, wherein training the prediction model comprises: receiving reference trip data corresponding to one or more reference trips of one or more reference drivers; transforming the reference trip data into a training image representation; and training the prediction model using the training image representation of the reference trips, wherein the reference trip data includes telematics data associated with a one or more reference trips taken by at least one driver of the one or more reference drivers.

8. The method of claim 7, wherein transforming the reference trip data into the training image representation comprises: dividing, for each reference trip of the one or more reference trips, the reference trip data into a one or more reference trip data segments based on the predetermined time period, wherein each reference trip data segment corresponding to at least a portion of the reference trip; generating, for each reference trip data segment of the one or more reference trip data segments, a graphical representation representing relative positions of a corresponding driverduring the predetermined time period by extracting location information from a corresponding reference trip data segment; adding depth to each point of the graphical representation, wherein the depth includes one or more channels that represent sensor data associated with the corresponding reference trip data segment; and generating the corresponding image representation for each reference trip data segment of the one or more reference trip data segments.

9. The method of claim 8, wherein the training image representation comprises an n-dimensional graphical representation with an n number of the sensor data associated with each reference trip data segment of the one or more reference trip data segments.

10. The method of any one of claims 4, further comprising: dividing the corresponding image representation for each reference trip data segment of one or more reference trip data segments into the patches; and training the prediction model using the patches of a training image representation.

11. The method of any one of claims 1-10, wherein the prediction model comprises a convolution neural network (CNN).

12. A computing device for predicting driving behaviors of a driver comprising: a processor; and a memory having instructions stored thereon that, when executed by the processor, causes the computing device to: receive trip data of one or more trips of the driver from one or more sensors; divide the trip data into a one or more trip data segments based on a predetermined time period, each trip data segment corresponding to at least a portion of the one or more trips; transform the one or more trip data segments into a corresponding image representation; and predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using a prediction model.

13. The computing device of claim 12, wherein transforming the one or more trip data segments into the corresponding image representation comprises to:generate, for each trip data segment of the one or more trip data segments, a graphical representation representing relative positions of a corresponding driver during the predetermined time period by extracting location information from a corresponding trip data segment; add depth to each point of a graphical representation; and generate the corresponding image representation for each trip data segment of the one or more trip data segments, wherein the depth includes one or more channels that represent sensor data, and the image representation comprising an n-dimensional graphical representation with an n number of the sensor data associated with each trip data segment of the one or more trip data segments.

14. The computing device of any one of claims 12-13, wherein predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using the prediction model comprises to: extract one or more features from the corresponding image representation of the one or more trips; and predicting the driving behaviors of the driver based at least in part upon the one or more features that have been extracted.

15. The computing device of claim 12, wherein the instructions, when executed, further cause the computing device to: receive reference trip data corresponding to one or more reference trips of one or more reference drivers; transform the reference trip data into a training image representation; and train the prediction model using the training image representation of the reference trips, wherein the reference trip data includes telematics data associated with a one or more reference trips taken by at least one driver of the one or more reference drivers.

16. The computing device of claim 15, wherein transforming the reference trip data into the training image representation comprises to: divide, for each reference trip of the one or more reference trips, the reference trip data into a one or more reference trip data segments based on the predetermined time period,wherein each reference trip data segment corresponding to at least a portion of the reference trips; generate, for each reference trip data segment of the one or more reference trip data segments, a graphical representation representing relative positions of a corresponding driver during the predetermined time period by extracting location information from a corresponding reference trip data segment; add depth to each point of a graphical representation, wherein the depth includes one or more channels that represent sensor data associated with the corresponding reference trip data segment; and generate the corresponding image representation for each reference trip data segment of the one or more reference trip data segments, wherein the training image representation comprises an n-dimensional graphical representation with an n number of the sensor data associated with each reference trip data segment of the one or more reference trip data segments.

17. A non- transitory computer-readable medium storing instructions for predicting driving behaviors of a driver, wherein instructions, when executed by one or more processors of a computing device, cause the computing device to: receive trip data of one or more trips of the driver from one or more sensors; divide the trip data into one or more trip data segments based on a predetermined time period, each trip data segment corresponding to at least a portion of the one or more trips; transform the one or more trip data segments into a corresponding image representation; and predicting the driving behaviors of the driver based on the corresponding image representation of the one or more trips using a prediction model.

18. The non-transitory computer-readable medium storing instructions for predicting the driving behaviors of claim 17, wherein transforming the one or more trip data segments into the corresponding image representation comprises to: generate, for each trip data segment, a graphical representation representing relative positions of the driver during the predetermined time period by extracting location information from a corresponding trip data segment; add depth to each point of a graphical representation; andgenerate the corresponding image representation for each trip data segment of the one or more trip data segments, wherein the depth includes one or more channels that represent sensor data, and the corresponding image representation comprising an n-dimensional graphical representation with an n number of the sensor data associated with each trip data segment of the one or more trip data segments.

19. The non-transitory computer-readable medium storing instructions for predicting driving the behaviors of any one of claims 17-18, further comprising: receive a reference trip data of reference trips of one or more reference drivers; transform the reference trip data into a training image representation; and train the prediction model using the training image representation of the reference trips; and wherein the reference trip data includes telematics data associated with the reference trips taken by at least one driver of the one or more reference drivers.

20. The non-transitory computer-readable medium storing instructions for predicting the driving behaviors of any one of claims 17-19, wherein transforming the reference trip data into the training image representation comprises to: divide, for each reference trip of the reference trips, the reference trip data into a one or more reference trip data segments based on the predetermined time period, each reference trip data segment corresponding to at least the portion of the reference trips; generate, for each reference trip data segment of the one or more reference trip data segments, the graphical representation representing relative positions of a corresponding driver during the predetermined time period by extracting the location information from a corresponding reference trip data segment; add the depth to each point of the graphical representation, wherein the depth includes the one or more channels that represent the sensor data associated with the corresponding reference trip data segment; and generate the corresponding image representation for each reference trip data segment of the one or more reference trip data segments, wherein the training image representation comprising the n-dimensional graphical representation with the n number of the sensor data associated with each reference trip data segment of the one or more reference trip data segments.

Images