DESC:SYSTEM TO DETECT ANOMALIES IN VEHICLE DATA
CROSS REFERENCE TO RELATED APPLICTIONS
The present application claims priority from Indian Provisional Patent Application No. 202421026808 filed on 31/03/2025, the entirety of which is incorporated herein by a reference.
TECHNICAL FIELD
The present disclosure generally relates to vehicle data analysis. Further, the present disclosure particularly relates to a system to detect anomalies in vehicle data.
BACKGROUND
Vehicle data analysis has become an essential aspect of modern transportation systems. Further, various methods have been employed to assess and monitor data associated with vehicle operations. Moreover, multiple state-of-the-art techniques have been developed to process vehicle data to enhance operational efficiency, detect irregularities, and optimize performance. Furthermore, the assessment of vehicle data involves analysing various parameters such as speed, acceleration, engine performance, braking patterns, and fuel consumption.
Further, conventional methods utilise rule-based techniques for detecting anomalies in vehicle data. Moreover, such techniques depend on predefined thresholds and manually crafted rules to identify deviations from expected behaviour. Furthermore, such techniques lack adaptability to dynamic driving conditions, road variations, and diverse driving styles, resulting in false positives and negatives. Additionally, manual rule-setting requires continuous updating to accommodate changing vehicle performance parameters, leading to increased complexity and inefficiency in data assessment.
Moreover, statistical methods have been widely used to analyse vehicle data for anomaly detection. Further, such methods rely on statistical distributions and historical data to establish patterns of normal vehicle behaviour. Furthermore, such methods face limitations when encountering outliers or rare events that deviate significantly from standard patterns. Additionally, statistical methods struggle with handling high-dimensional data, requiring extensive computational resources and leading to processing inefficiencies.
Additionally, model-based techniques have been employed for detecting irregularities in vehicle data. Further, such techniques construct mathematical models based on expected vehicle behaviour under different operational conditions. Moreover, such models require extensive calibration and validation using real-world data, increasing implementation complexity. Furthermore, variations in vehicle types, driving environments, and external factors affect the accuracy of such techniques, leading to potential misclassification of anomalies.
Moreover, recent advancements have introduced machine learning techniques for vehicle data analysis. Further, such techniques leverage data-driven models trained on historical data to identify anomalies. Furthermore, supervised learning approaches require labelled datasets for training, which may not always be available, limiting applicability. Additionally, unsupervised learning methods, while effective in identifying anomalies, often suffer from interpretability issues, making decision-making challenging.
Additionally, existing methods lack real-time adaptability, resulting in delays in detecting and responding to anomalies in vehicle data. Further, dependency on predefined rules, statistical assumptions, or rigid models affects the accuracy and reliability of anomaly detection mechanisms. Moreover, existing methods often fail to generalise across different vehicle categories, road conditions, and driving behaviours, reducing the effectiveness of vehicle data assessment.
In light of the above discussion, there exists an urgent need for solutions that overcome the problems associated with conventional systems and techniques for detecting anomalies in vehicle data.
.
SUMMARY
The aim of the present disclosure is to provide a system to detect the anomalies in vehicle data, wherein the system detects anomalies in vehicle data by processing time series data using an anomaly detection system, determining corrective actions, and transmitting commands to a connected vehicle for execution.
The present disclosure relates to a system and a method to detect anomalies in vehicle data by processing time series data using a machine learning model. Further, the system and the method aim to determine an action upon detection of an anomaly and enable execution of the determined action in a connected vehicle.
In an aspect, the present disclosure provides a system to detect anomalies in vehicle data. The system comprises a connected vehicle and an anomaly detection system communicatively linked to the connected vehicle. The anomaly detection system comprises at least one computing device comprising at least one processor and a non-transitory computer-readable medium storing instructions that, when executed by the processor, cause the anomaly detection system to receive a time series data record from the connected vehicle, process the received time series data record using a machine learning model to generate an anomaly score, and upon determination that the anomaly score exceeds an anomaly threshold value, determine an action to address the anomaly and transmit a command to the connected vehicle specifying the determined action to be executed.
Further, the connected vehicle receives the command from the anomaly detection system and executes the action specified in the command. Moreover, the time series data record comprises telemetry data selected from at least one of a motor speed, a control surface servo position, a linear velocity, an orientation, or an angular velocity. Furthermore, the anomaly detection system determines the anomaly threshold value based on historical data associated with the connected vehicle and updates the anomaly threshold value using real-time data received from the connected vehicle.
Additionally, the anomaly detection system trains the machine learning model using a dataset comprising telemetry data corresponding to at least two of a start portion of a trip, a driving portion of a trip, and an end portion of a trip. Further, the training comprises applying a weighting mechanism to identify anomalous time series data records to be excluded during processing, assigning a time series data record weight to each record in a training dataset and updating the time series data record weights based on anomaly frequency, alternating between optimizing a set of fitting weights and optimizing a set of time series data record weights, processing the time series data record using a multi-stage anomaly detection approach that comprises filtering, feature extraction, and classification, adjusting the anomaly threshold value using a feedback mechanism based on prior anomaly detection accuracy, and periodically updating the machine learning model using additional time series data collected from the connected vehicle.
Moreover, the action to address the anomaly comprises at least one of rescheduling a future trip or enabling remote control by an external operator. Further, the connected vehicle executes the action by at least one of navigating to an emergency repair location or performing an immediate halting procedure. Furthermore, the anomaly detection system generates an alert when the anomaly score exceeds the anomaly threshold value and receives a confirmation signal from the connected vehicle upon execution of the action. Additionally, the anomaly detection system compares the detected anomaly with a geographically indexed anomaly database to identify external factors influencing the anomaly.
In another aspect, the present disclosure provides a method for detecting anomalies in vehicle data. The method comprises receiving, by an anomaly detection system communicatively linked to a connected vehicle, a time series data record from the connected vehicle, processing the received time series data record using a machine learning model to generate an anomaly score, determining whether the anomaly score exceeds an anomaly threshold value, upon determination that the anomaly score exceeds the anomaly threshold value, determining an action to address the detected anomaly, and transmitting a command to the connected vehicle specifying the determined action to be executed.
Further, the anomaly detection system assigns a severity classification to the detected anomaly based on at least one of deviation magnitude from expected parameters, frequency of occurrence, and real-time operational conditions of the connected vehicle. Moreover, the anomaly detection system adjusts the anomaly threshold value based on historical anomaly detection data and real-time operational parameters of the connected vehicle. Furthermore, the anomaly detection system prioritizes execution of the determined action based on an operational safety risk assessment corresponding to the detected anomaly. Additionally, the anomaly detection system determines a recommended corrective measure based on a correlation between the detected anomaly and historical corrective actions stored in a remote database. Moreover, the anomaly detection system assigns a probability score to the detected anomaly based on historical occurrence patterns, environmental conditions, and operational parameters of the connected vehicle.
BRIEF DESCRIPTION OF DRAWINGS
The summary above, as well as the following detailed description of illustrative embodiments, is better understood when read in conjunction with the appended drawings. For the purpose of illustrating the present disclosure, exemplary constructions of the disclosure are shown in the drawings. However, the present disclosure is not limited to specific methods and instrumentalities disclosed herein. Moreover, those in the art will understand that the drawings are not to scale. Wherever possible, like elements have been indicated by identical numbers.
Embodiments of the present disclosure will now be described, by way of example only, with reference to the following diagrams wherein:
FIG. 1 illustrates a system 100 to detect the anomalies in vehicle data, in accordance with the embodiments of the present disclosure;
FIG. 2 illustrates a method 200 for detecting the anomalies in vehicle data, in accordance with the embodiments of the present disclosure;
FIG. 3 illustrates a class diagram of the system 100 for detecting anomalies in vehicle data, in accordance with the embodiments of the present disclosure; and
FIG. 4 illustrates a sequential process for detecting anomalies in vehicle data, in accordance with the embodiments of the present disclosure.
In the accompanying drawings, an underlined number is employed to represent an item over which the underlined number is positioned or an item to which the underlined number is adjacent. A non-underlined number relates to an item identified by a line linking the non-underlined number to the item. When a number is non-underlined and accompanied by an associated arrow, the non-underlined number is used to identify a general item at which the arrow is pointing.
DETAILED DESCRIPTION
The following detailed description illustrates embodiments of the present disclosure and ways in which they can be implemented. Although some modes of carrying out the present disclosure have been disclosed, those skilled in the art would recognise that other embodiments for carrying out or practising the present disclosure are also possible.
The description set forth below in connection with the appended drawings is intended as a description of certain embodiments of a system to detect the anomalies in vehicle data and is not intended to represent the only forms that may be developed or utilised. The description sets forth the various structures and/or functions in connection with the illustrated embodiments; however, it is to be understood that the disclosed embodiments are merely exemplary of the disclosure that may be embodied in various and alternative forms. The figures are not necessarily to scale; some features may be exaggerated or minimised to show details of particular components. Therefore, specific structural and functional details disclosed herein are not to be interpreted as limiting, but merely as a representative basis for teaching one skilled in the art to variously employ the present invention.
While the disclosure is susceptible to various modifications and alternative forms, specific embodiment thereof has been shown by way of example in the drawings and will be described in detail below. It should be understood, however, that it is not intended to limit the disclosure to the particular forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure.
In the following detailed description of the embodiments of the disclosure, reference is made to the accompanying drawings, and which are shown by way of illustration specific embodiments in which the disclosure may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the disclosure, and it is to be understood that other embodiments may be utilized and that changes may be made without departing from the scope of the present disclosure. The following description is, therefore, not to be taken in a limiting sense.
The present disclosure will be described herein below with reference to the accompanying drawings. In the following description, well known functions or constructions are not described in detail since they would obscure the description with unnecessary detail.
As used herein, the term "system" is used to refer to an arrangement of interconnected components that collectively perform a specific operation or set of operations. The components may comprise hardware, software, or a combination of both, working together to achieve a defined objective. The system may be implemented using various computational architectures, including centralized processing units, distributed computing networks, or cloud-based infrastructures. Additionally, the system may process data in real time or batch mode, depending on operational requirements. Further, the system may interact with external databases, sensors, and control units to exchange data and execute predefined instructions.
As used herein, the term "connected vehicle" is used to refer to any vehicle that is capable of wirelessly communicating with external systems, networks, or other vehicles. The connected vehicle may comprise vehicles integrated with telematics units, communication interfaces, and sensor networks that enable real-time data transmission and reception. The connected vehicle may utilize cellular networks, satellite communication, Wi-Fi, or dedicated short-range communication (DSRC) for exchanging data with remote servers, cloud platforms, or other vehicles in proximity. Additionally, the connected vehicle may transmit telemetry data, including speed, location, acceleration, braking patterns, and engine diagnostics, to external monitoring systems for analysis. Further, the connected vehicle may receive commands from remote control systems or automated decision-making units to adjust operational parameters, perform corrective actions, or respond to detected anomalies. The connected vehicle may comprise passenger cars, commercial trucks, buses, motorcycles, and autonomous vehicles integrated with connectivity features to enable intelligent transportation applications.
As used herein, the term "anomaly detection system" is used to refer to a computational framework that identifies deviations from expected patterns in data. The anomaly detection system may comprise real-time monitoring, data aggregation, and analysis to detect irregularities in operational parameters. The anomaly detection system may utilize predefined threshold values, statistical techniques, or machine learning-based approaches to distinguish between normal and abnormal data patterns. Additionally, the anomaly detection system may incorporate historical data and real-time inputs to dynamically adjust detection criteria based on changing conditions. Further, the anomaly detection system may generate alerts, recommendations, or corrective actions based on identified anomalies.
As used herein, the term "computing device" is used to refer to an electronic system that processes data and executes instructions. The computing device may comprise embedded systems, microcontrollers, single-board computers, general-purpose computing units, and specialized hardware platforms. The computing device may contain various hardware components, including a processor, memory units, input/output interfaces, and communication modules. Additionally, the computing device may support multiple operating systems, software frameworks, and programming environments to execute computational tasks. Further, the computing device may perform real-time processing, data storage, and transmission functions to facilitate decision-making and control operations. The computing device may interact with external sensors, databases, and cloud platforms to retrieve, process, and analyze information. The computing device may be implemented in various domains, including automotive systems, industrial automation, medical diagnostics, and smart infrastructure.
As used herein, the term "processor" is used to refer to an integrated circuit or microelectronic unit that executes computational instructions. Processor may comprise single-core, multi-core, or specialized processors designed for high-performance computing tasks. The processor may perform arithmetic and logical operations, data manipulation, and instruction execution based on predefined software programs. Additionally, the processor may operate in conjunction with memory units, input/output controllers, and peripheral devices to execute complex tasks efficiently. Further, the processor may support various instruction sets, parallel processing architectures, and optimization techniques to enhance computational performance. The processor may be utilized in various applications, including embedded systems, artificial intelligence, cryptographic computations, and real-time control systems. The processor may incorporate power-saving features, adaptive frequency scaling, and fault detection mechanisms to optimize performance and reliability under varying operational conditions.
As used herein, the term "non-transitory computer-readable medium" is used to refer to any tangible storage device that retains data and program instructions for execution by a computing device. The non-transitory computer-readable medium may comprise solid-state drives, magnetic disks, optical storage media, flash memory, and embedded memory units. The non-transitory computer-readable medium may store operating systems, software applications, configuration settings, and data logs required for computational tasks. Additionally, the non-transitory computer-readable medium may facilitate persistent data retention, preventing loss of information during power interruptions or system failures. Further, the non-transitory computer-readable medium may support read-only, write-once, or rewritable storage mechanisms, depending on application requirements. The non-transitory computer-readable medium may be integrated into computing devices, networked storage systems, or distributed computing infrastructures to enable efficient data access and retrieval. The non-transitory computer-readable medium may incorporate encryption, access control, and redundancy mechanisms to protect stored data from unauthorized access, corruption, or loss.
As used herein, the term "time series data record" is used to refer to a sequence of data points recorded at successive time intervals. The time series data record may comprise numerical measurements, sensor readings, or log entries that capture variations in parameters over time. The time series data record may be collected from sensors, monitoring devices, or automated logging systems in real-time or batch mode. Additionally, the time series data record may comprise data, supporting various data representation techniques such as timestamps, labels, and categorical attributes. Further, the time series data record may be utilized for trend analysis, anomaly detection, forecasting, and predictive modeling. The time series data record may be processed using filtering techniques, statistical transformations, and data compression methods to enhance efficiency and accuracy. The time series data record may be stored in relational databases, distributed storage systems, or cloud-based repositories to facilitate scalable data management and retrieval.
As used herein, the term "machine learning model" is used to refer to a computational framework that identifies patterns and relationships in data. The machine learning model may comprise supervised, unsupervised, or reinforcement learning techniques that enable adaptive decision-making. The machine learning model may be trained using historical datasets, feature selection methods, and optimization algorithms to improve predictive accuracy. Additionally, the machine learning model may incorporate real-time learning mechanisms to adapt to evolving data distributions and environmental conditions. Further, the machine learning model may process large-scale datasets, extracting relevant features and identifying correlations to enhance classification, clustering, and regression tasks. The machine learning model may be applied across various domains, including finance, healthcare, industrial automation, and autonomous systems. The machine learning model may utilize cloud computing, edge computing, or hybrid architectures to balance computational efficiency and scalability. The machine learning model may comprise mechanisms for interpretability, bias mitigation, and uncertainty quantification to enhance decision reliability.
FIG. 1 illustrates a system 100 to detect the anomalies in vehicle data, in accordance with the embodiments of the present disclosure. The system 100 comprises a connected vehicle 102 that is communicatively linked to external systems for data transmission and reception. The connected vehicle 102 comprises an onboard communication interface that facilitates data exchange with remote servers, control systems, and external networks. The connected vehicle 102 supports multiple communication technologies, including cellular networks, satellite communication, Wi-Fi, and dedicated short-range communication (DSRC). The connected vehicle 102 generates telemetry data representing various operational parameters, including speed, acceleration, braking patterns, engine diagnostics, and environmental conditions. The connected vehicle 102 transmits the telemetry data in the form of a time series data record to an external processing system for analysis. The connected vehicle 102 is further capable of receiving control commands from remote systems and executing predefined actions based on received instructions. The connected vehicle 102 comprises onboard computing hardware that processes received commands and initiates actions, such as adjusting driving parameters, activating emergency responses, or modifying operational settings. The connected vehicle 102 interacts with various external data sources, including traffic management systems, vehicle-to-vehicle communication networks, and cloud-based analytics platforms. The connected vehicle 102 continuously updates operational parameters based on received data and dynamically adjusts communication settings to maintain seamless connectivity. The connected vehicle 102 implements security mechanisms to protect transmitted and received data against unauthorized access. The connected vehicle 102 supports integration with multiple sensor systems, including cameras, radar units, and LiDAR systems, to enhance data collection and transmission capabilities. The connected vehicle 102 is applicable to various vehicle types, including passenger cars, commercial trucks, buses, motorcycles, and autonomous vehicles.
In an embodiment, the system 100 comprises an anomaly detection system 104 that is communicatively linked to the connected vehicle 102 for receiving and processing vehicle data. The anomaly detection system 104 comprises at least one computing device that comprises at least one processor and a non-transitory computer-readable medium storing instructions. The anomaly detection system 104 continuously monitors data transmitted from the connected vehicle 102 to identify deviations from expected operational patterns. The anomaly detection system 104 receives a time series data record representing real-time telemetry data from the connected vehicle 102. The anomaly detection system 104 processes the received time series data record using a machine learning model that evaluates data patterns and determines an anomaly score. The anomaly detection system 104 compares the anomaly score to an anomaly threshold value to determine whether an anomaly is present in the received data. Upon determining that the anomaly score exceeds the anomaly threshold value, the anomaly detection system 104 determines an action to address the detected anomaly. The anomaly detection system 104 transmits a command to the connected vehicle 102 specifying the determined action for execution. The anomaly detection system 104 dynamically updates the anomaly threshold value based on historical data and real-time operational parameters received from the connected vehicle 102. The anomaly detection system 104 generates alerts corresponding to detected anomalies and transmits notifications to external monitoring systems. The anomaly detection system 104 implements security mechanisms to prevent unauthorized access to data and assure the reliability of detected anomalies. The anomaly detection system 104 interacts with external databases and predictive analytics platforms to refine anomaly detection accuracy. The anomaly detection system 104 supports integration with remote diagnostic systems, enabling proactive maintenance and risk mitigation. The anomaly detection system 104 periodically updates stored instructions and computational models to improve anomaly detection capabilities over time.
In an embodiment, a connected vehicle 102 may receive a command from an anomaly detection system 104 and executes an action specified in the command. The connected vehicle 102 comprises an onboard communication system that enables wireless data transmission and reception. The communication system supports various network technologies, including cellular, satellite, and short-range communication. The connected vehicle 102 continuously monitors incoming messages from the anomaly detection system 104 and processes received commands in real-time. Upon receiving a command, the connected vehicle 102 verifies the authenticity and integrity of the transmitted data using encryption and validation techniques. The connected vehicle 102 accesses onboard control systems to determine appropriate mechanisms for executing the specified action. The action may involve modifying vehicle parameters, activating emergency protocols, or adjusting driving operations based on received instructions. The connected vehicle 102 interacts with onboard sensors, actuators, and computing units to implement command execution procedures. The connected vehicle 102 records executed actions in an event log for auditing and performance analysis. The connected vehicle 102 synchronizes data with external servers to maintain an updated record of operational responses. The connected vehicle 102 utilizes real-time telemetry inputs to assess environmental conditions before performing corrective actions. The connected vehicle 102 incorporates fail-safe mechanisms to enable proper execution of received commands while preventing unintended disruptions. The connected vehicle 102 generates status reports and confirmation signals upon executing the specified action and transmits the confirmation to the anomaly detection system 104. The connected vehicle 102 adjusts system parameters based on ongoing communication with external monitoring units. The connected vehicle 102 may comprise provisions for remote overrides by authorized control centers, allowing external operators to verify or modify executed actions.
In an embodiment, a time series data record may comprise telemetry data representing operational characteristics of a connected vehicle 102. The telemetry data comprises parameters selected from at least one of a motor speed, a control surface servo position, a linear velocity, an orientation, or an angular velocity. The time series data record is collected at predefined intervals, providing a continuous stream of data points that describe vehicle performance over time. The time series data record is generated by onboard sensors and electronic control units that measure various physical and dynamic properties of the connected vehicle 102. The motor speed represents the rotational speed of the propulsion system, providing insight into energy consumption and efficiency. The control surface servo position corresponds to the angular displacement of control surfaces, including steering mechanisms or aerodynamic control elements. The linear velocity defines the translational motion of the connected vehicle 102 in a specified direction, capturing acceleration and deceleration characteristics. The orientation parameter indicates the spatial alignment of the connected vehicle 102 with respect to a reference coordinate system. The angular velocity quantifies the rate of rotational movement around an axis, reflecting dynamic stability and manoeuvrability. The time series data record is transmitted from the connected vehicle 102 to the anomaly detection system 104 for further processing. The time series data record undergoes preprocessing operations, including filtering, noise reduction, and feature extraction. The time series data record serves as input for predictive analysis, anomaly detection, and operational diagnostics. The time series data record may be stored in a centralized database for historical trend analysis and comparative evaluation. The time series data record facilitates correlation of vehicle performance metrics with external conditions, such as road characteristics and environmental influences. The time series data record supports integration with machine learning techniques to refine predictive models and improve system responsiveness. The time series data record may be in a standardized format to enable interoperability across multiple data analysis platforms.
In an embodiment, an anomaly detection system 104 may determine an anomaly threshold value based on historical data associated with a connected vehicle 102 and updates the anomaly threshold value using real-time data received from the connected vehicle 102. The anomaly detection system 104 collects historical telemetry data from previous operational cycles of the connected vehicle 102, including motor speed, control surface positions, velocity, orientation, and angular velocity. The anomaly detection system 104 applies statistical analysis techniques to establish baseline operational patterns and identify acceptable ranges for vehicle parameters. The anomaly detection system 104 calculates an initial anomaly threshold value based on deviations observed in historical data, accounting for variations caused by environmental conditions and driving behavior. The anomaly detection system 104 continuously receives real-time telemetry data from the connected vehicle 102 and compares received data against predefined thresholds. The anomaly detection system 104 dynamically adjusts the anomaly threshold value based on observed trends in real-time data, making sure adaptability to evolving vehicle conditions. The anomaly detection system 104 incorporates a self-learning mechanism that refines anomaly threshold values through iterative analysis of historical and real-time inputs. The anomaly detection system 104 utilizes feedback mechanisms to recalibrate anomaly detection criteria when significant deviations are consistently observed. The anomaly detection system 104 prevents false positive detections by filtering out temporary fluctuations that do not indicate system failures or safety risks. The anomaly detection system 104 transmits updated anomaly threshold values to connected processing units to maintain synchronization across monitoring systems. The anomaly detection system 104 integrates external data sources, such as environmental sensors and road condition databases, to enhance the contextual accuracy of anomaly threshold determination.
In an embodiment, an anomaly detection system 104 may train the machine learning model using a dataset that comprises telemetry data corresponding to at least two of a start portion of a trip, a driving portion of a trip, and an end portion of a trip. The anomaly detection system 104 applies a weighting mechanism to identify anomalous time series data records to be excluded during processing. The anomaly detection system 104 assigns a time series data record weight to each record in a training dataset and updates the time series data record weights based on anomaly frequency. The anomaly detection system 104 alternates between optimizing a set of fitting weights and optimizing a set of time series data record weights, wherein the anomalous records are excluded during optimization. The anomaly detection system 104 processes the time series data record using a multi-stage anomaly detection approach that comprises filtering, feature extraction, and classification. The anomaly detection system 104 adjusts the anomaly threshold value using a feedback mechanism based on prior anomaly detection accuracy. The anomaly detection system 104 periodically updates the machine learning model using additional time series data collected from the connected vehicle 102. The anomaly detection system 104 enhances predictive capabilities by incorporating diverse driving scenarios and environmental conditions into the training dataset. The anomaly detection system 104 optimizes model parameters to minimize false positives and false negatives during anomaly detection. The anomaly detection system 104 dynamically refines detection criteria based on evolving operational trends observed in collected telemetry data. The anomaly detection system 104 supports scalable model updates, allowing the inclusion of new vehicle data to improve anomaly detection accuracy. The anomaly detection system 104 maintains a data repository to facilitate retrospective analysis and model validation. The anomaly detection system 104 integrates data augmentation techniques to enhance learning performance by simulating rare but significant anomaly scenarios.
In an embodiment, an action to address an anomaly may comprise at least one of rescheduling a future trip or enabling remote control by an external operator. The anomaly detection system 104 analyzes the anomaly detected in the time series data record received from the connected vehicle 102 and determines whether immediate corrective action is required. If the anomaly does not indicate an immediate safety risk or operational failure, the anomaly detection system 104 determines whether rescheduling a future trip would mitigate the potential impact of the detected anomaly. The anomaly detection system 104 evaluates factors such as route conditions, environmental influences, and vehicle availability before rescheduling a future trip. The anomaly detection system 104 updates the scheduling database and notifies fleet management systems or operators regarding the rescheduled trip. If the anomaly poses an immediate risk or requires external intervention, the anomaly detection system 104 determines whether enabling remote control by an external operator is necessary. The anomaly detection system 104 establishes a secure communication link between the connected vehicle 102 and an authorized external operator. The external operator receives real-time telemetry data from the connected vehicle 102 and monitors vehicle status before initiating control actions. The connected vehicle 102 transitions to a remote-control mode, allowing the external operator to modify vehicle operations, adjust navigation routes, or activate safety measures. The anomaly detection system 104 verifies authentication credentials before enabling remote access to prevent unauthorized intervention. The connected vehicle 102 logs all remote-control activities to maintain an audit trail of executed commands. The anomaly detection system 104 restores autonomous control to the connected vehicle 102 once the external operator determines that the anomaly has been addressed. The anomaly detection system 104 transmits a confirmation signal indicating the successful resolution of the anomaly and updates operational records accordingly.
In an embodiment, the connected vehicle 102 may execute an action by at least one of navigating to an emergency repair location or performing an immediate halting procedure. The connected vehicle 102 continuously monitors incoming commands from the anomaly detection system 104 and determines whether the detected anomaly requires vehicle navigation to an emergency repair location. The anomaly detection system 104 evaluates vehicle diagnostics and external conditions before selecting an appropriate repair location. The anomaly detection system 104 transmits location coordinates and navigation instructions to the connected vehicle 102, enabling optimal routing to the designated repair facility. The connected vehicle 102 communicates with traffic management systems and road infrastructure to facilitate seamless navigation to the repair location. If the anomaly detection system 104 determines that continued operation of the connected vehicle 102 poses a safety risk, an immediate halting procedure is initiated. The connected vehicle 102 receives a command from the anomaly detection system 104 instructing the onboard control system to safely stop vehicle movement. The connected vehicle 102 activates braking mechanisms and gradually reduces velocity while considering road conditions and surrounding traffic. The connected vehicle 102 transmits hazard alerts to nearby vehicles and road authorities to notify them of the emergency halting procedure. The connected vehicle 102 records all executed actions in an event log to enable post-event analysis and compliance verification. The anomaly detection system 104 monitors system status after execution of the halting procedure and determines whether further intervention is required. The connected vehicle 102 awaits further instructions from the anomaly detection system 104 before resuming operation
In an embodiment, the anomaly detection system 104 may generate an alert when an anomaly score exceeds an anomaly threshold value and receives a confirmation signal from a connected vehicle 102 upon execution of the action. The anomaly detection system 104 continuously processes telemetry data transmitted from the connected vehicle 102 and computes an anomaly score based on deviations observed in operational parameters. The anomaly detection system 104 compares the computed anomaly score against the anomaly threshold value to determine whether an alert should be generated. When the anomaly score exceeds the anomaly threshold value, the anomaly detection system 104 generates an alert indicating the detection of an abnormal operational condition. The alert contains information regarding the detected anomaly, including affected parameters, severity classification, and recommended corrective measures. The anomaly detection system 104 transmits the alert to external monitoring systems, fleet operators, or vehicle occupants to initiate an appropriate response. The anomaly detection system 104 supports multiple alert transmission methods, including direct communication with vehicle control systems, cloud-based notifications, and integration with external emergency response networks. The connected vehicle 102 executes the corrective action specified in the command received from the anomaly detection system 104 and transmits a confirmation signal upon completion. The confirmation signal comprises details of the executed action, the time of execution, and post-action system status. The anomaly detection system 104 records the confirmation signal in a centralized event log for future analysis and validation. The anomaly detection system 104 utilizes received confirmation signals to assess the effectiveness of corrective actions and refine anomaly detection criteria.
In an embodiment, the anomaly detection system 104 may compare a detected anomaly with a geographically indexed anomaly database to identify external factors influencing the anomaly. The anomaly detection system 104 maintains a geographically indexed anomaly database containing historical records of anomalies detected across various locations. The database comprises data related to road conditions, weather patterns, infrastructure characteristics, and common vehicle performance deviations associated with specific regions. When the anomaly detection system 104 detects an anomaly in telemetry data received from the connected vehicle 102, the system retrieves relevant entries from the geographically indexed anomaly database for comparison. The anomaly detection system 104 analyzes location-based trends and environmental factors that may have contributed to the detected anomaly. The anomaly detection system 104 assesses whether the anomaly is correlated with recurring patterns observed in similar geographic locations. The anomaly detection system 104 enhances anomaly classification accuracy by distinguishing between anomalies caused by internal vehicle faults and those influenced by external conditions. The anomaly detection system 104 updates anomaly detection models based on recurring geographic trends to improve predictive capabilities. The anomaly detection system 104 integrates external data sources, including weather monitoring stations and traffic reports, to provide an analysis of environmental influences on vehicle anomalies. The anomaly detection system 104 generates location-specific recommendations for mitigating anomaly occurrences in high-risk areas. The anomaly detection system 104 transmits geographic anomaly assessments to fleet management systems and transportation authorities for informed decision-making
FIG. 2 illustrates a method 200 for detecting the anomalies in vehicle data, in accordance with the embodiments of the present disclosure. The method 200 comprises sequential stages including receiving data (at step 202), processing data (at step 204), threshold evaluation (at step 206), and anomaly response (at step 208). The method 200 begins with receiving data (at step 202), where an anomaly detection system collects a time series data record from a connected vehicle. The collected data comprises telemetry parameters such as motor speed, linear velocity, and orientation. The method 200 proceeds to processing data (at step 204), where the received time series data is analyzed using a machine learning model to identify patterns and generate an anomaly score. The method 200 then advances to threshold evaluation (at step 206), where the generated anomaly score is compared against a predefined anomaly threshold value to determine whether an anomaly exists. If the anomaly score exceeds the threshold, the method 200 proceeds to anomaly response (at step 208), where an appropriate corrective action is determined. The corrective action is transmitted as a command to the connected vehicle, instructing execution of the response, such as adjusting operational parameters, initiating emergency measures, or transmitting alerts. The figure represents the workflow of the anomaly detection process, enable continuous monitoring and response to deviations in vehicle performance.
In an embodiment, the anomaly detection system 104 may assign a severity classification to a detected anomaly based on at least one of deviation magnitude from expected parameters, frequency of occurrence, and real-time operational conditions of a connected vehicle 102. The anomaly detection system 104 evaluates the extent to which received telemetry data deviates from normal operating ranges established through historical data analysis. The anomaly detection system 104 computes a severity score by comparing the magnitude of deviation against predefined thresholds corresponding to minor, moderate, and critical anomalies. The anomaly detection system 104 analyzes the frequency of occurrence of detected anomalies to assess whether an anomaly represents a recurring issue or an isolated event. The anomaly detection system 104 considers real-time operational conditions, including vehicle load, environmental factors, and road conditions, to refine severity classification. The anomaly detection system 104 assigns a severity level to the detected anomaly, categorizing the severity level as low-risk, medium-risk, or high-risk based on computed parameters. The anomaly detection system 104 prioritizes responses to high-risk anomalies by initiating immediate corrective actions and alerting external monitoring systems. The anomaly detection system 104 transmits severity classifications to fleet operators, maintenance personnel, or autonomous control units for informed decision-making. The anomaly detection system 104 refines severity classification models through continuous learning mechanisms that incorporate real-time data feedback. The anomaly detection system 104 enables risk assessment methodologies to improve predictive maintenance and operational reliability.
In an embodiment, the anomaly detection system 104 may adjust an anomaly threshold value based on historical anomaly detection data and real-time operational parameters of a connected vehicle 102. The anomaly detection system 104 continuously collects and stores historical data from multiple operational cycles of the connected vehicle 102, including telemetry records, previously detected anomalies, and corrective actions executed in response. The anomaly detection system 104 analyzes long-term trends in vehicle behavior to establish baseline performance metrics and determine typical operating ranges for various telemetry parameters. The anomaly detection system 104 dynamically updates the anomaly threshold value by comparing incoming real-time operational parameters against historical performance data. The anomaly detection system 104 prevents excessive false positive detections by refining anomaly threshold values based on seasonal variations, road conditions, and environmental influences. The anomaly detection system 104 incorporates feedback from executed corrective actions to improve anomaly threshold accuracy, making sure that detected anomalies represent significant deviations rather than minor fluctuations. The anomaly detection system 104 supports adaptive recalibration of threshold values using machine learning techniques to enhance detection precision. The anomaly detection system 104 integrates multiple data sources, including external diagnostic reports, manufacturer specifications, and predictive maintenance insights, to refine anomaly threshold values further. The anomaly detection system 104 transmits updated anomaly threshold values to the connected vehicle 102 and associated control systems, making sure that all monitoring components utilize the most recent detection criteria. The anomaly detection system 104 records all threshold adjustments in an operational log for auditing, regulatory compliance, and performance analysis.
In an embodiment, the anomaly detection system 104 may prioritize execution of a determined action based on an operational safety risk assessment corresponding to a detected anomaly. The anomaly detection system 104 evaluates the potential impact of the detected anomaly on vehicle performance, passenger safety, and surrounding traffic conditions. The anomaly detection system 104 calculates a risk score by assessing deviation severity, likelihood of system failure, and potential consequences if corrective action is delayed. The anomaly detection system 104 applies predefined prioritization rules to determine whether the detected anomaly requires immediate intervention or scheduled corrective action. The anomaly detection system 104 assigns higher priority to anomalies associated with critical vehicle components, including braking systems, propulsion units, and steering mechanisms. The anomaly detection system 104 prioritizes safety-critical anomalies over performance-related deviations to prevent hazardous operating conditions. The anomaly detection system 104 initiates emergency response protocols for anomalies that pose imminent safety threats, including transmission of high-priority alerts to remote operators and emergency service providers. The anomaly detection system 104 dynamically adjusts prioritization criteria based on real-time road conditions, vehicle load, and external environmental factors. The anomaly detection system 104 transmits prioritized execution commands to the connected vehicle 102, making sure that high-risk anomalies receive immediate corrective action. The anomaly detection system 104 logs all prioritization decisions and executed actions for post-event analysis and system optimization. The anomaly detection system 104 enhances system responsiveness by enabling proactive decision-making based on real-time risk evaluation.
In an embodiment, the anomaly detection system 104 may determine a recommended corrective measure based on a correlation between a detected anomaly and historical corrective actions stored in a remote database. The anomaly detection system 104 maintains a repository of previously detected anomalies, corresponding corrective actions, and post-correction performance data. The anomaly detection system 104 retrieves historical records from the remote database and analyzes recurring anomaly patterns to determine effective corrective measures for similar anomalies. The anomaly detection system 104 identifies similarities between the detected anomaly and historical cases by comparing anomaly characteristics, including affected vehicle parameters, deviation severity, and environmental conditions at the time of detection. The anomaly detection system 104 applies correlation techniques to assess the effectiveness of previously executed corrective actions in resolving similar anomalies. The anomaly detection system 104 selects a recommended corrective measure based on past resolutions that resulted in optimal vehicle recovery and operational stability. The anomaly detection system 104 transmits the recommended corrective measure to the connected vehicle 102 or external control systems for execution. The anomaly detection system 104 enables automated decision-making by dynamically selecting corrective measures that align with real-time operational conditions. The anomaly detection system 104 updates the remote database with new corrective action outcomes, enabling continuous improvement in anomaly resolution strategies. The anomaly detection system 104 integrates external diagnostic reports, manufacturer guidelines, and predictive maintenance insights to enhance the accuracy of recommended corrective measures. The anomaly detection system 104 transmits recommended corrective measures to fleet management systems and vehicle operators, enabling proactive maintenance planning and operational adjustments.
In an embodiment, the anomaly detection system 104 may assign a probability score to a detected anomaly based on historical occurrence patterns, environmental conditions, and operational parameters of a connected vehicle 102. The anomaly detection system 104 continuously monitors incoming telemetry data and identifies deviations that match predefined anomaly detection criteria. The anomaly detection system 104 calculates a probability score by analyzing historical records of similar anomalies and their frequency of occurrence across multiple operational cycles. The anomaly detection system 104 evaluates environmental factors, including road conditions, weather variations, and traffic density, to determine whether external influences contribute to the detected anomaly. The anomaly detection system 104 assesses real-time operational parameters, including vehicle load, engine performance, and sensor readings, to refine probability score calculations. The anomaly detection system 104 applies statistical techniques to differentiate between transient fluctuations and anomalies indicative of potential system failures. The anomaly detection system 104 transmits probability scores to external monitoring systems, fleet operators, and maintenance personnel for informed decision-making. The anomaly detection system 104 utilizes probability scoring mechanisms to prioritize anomalies requiring immediate intervention and filter out low-risk deviations. The anomaly detection system 104 continuously updates probability scoring models based on evolving vehicle performance trends and operational insights. The anomaly detection system 104 supports integration with predictive analytics platforms, enabling automated risk assessment and anomaly classification based on historical and real-time data correlations.
In an embodiment, the system 100 comprises the connected vehicle 102 and the anomaly detection system 104 communicatively linked to the connected vehicle 102. The anomaly detection system 104 comprises at least one computing device including at least one processor and a non-transitory computer-readable medium storing instructions. The anomaly detection system 104 receives a new time series data record from the connected vehicle 102 and processes the received data using a machine learning model to generate an anomaly score. The anomaly detection system 104 determines whether the anomaly score exceeds an anomaly threshold value, indicating a deviation from expected operational parameters. If the anomaly threshold value is exceeded, the anomaly detection system 104 determines an action to address the detected anomaly and transmits a command to the connected vehicle 102 specifying the determined action for execution. The system 100 provides real-time anomaly detection and response, reducing the likelihood of operational failures and enhancing safety by affirming corrective actions are taken promptly. The integration of data-driven anomaly identification and automated response mechanisms facilitates proactive vehicle monitoring and reduces reliance on manual diagnostics. The anomaly detection system 104 continuously refines anomaly detection accuracy by incorporating historical data and real-time telemetry inputs, supporting adaptive decision-making processes.
In an embodiment, the connected vehicle 102 is configured to receive a command from the anomaly detection system 104 and execute the action specified in the command. The connected vehicle 102 maintains a continuous communication link with the anomaly detection system 104, enabling real-time data transmission and reception. Upon receiving a command, the connected vehicle 102 validates the command authenticity and retrieves execution parameters from onboard control systems. The connected vehicle 102 initiates corrective measures based on the received command, modifying operational settings, adjusting driving parameters, or activating safety mechanisms as necessary. The connected vehicle 102 interacts with internal sensors and actuators to implement the specified action while maintaining stability and performance. The connected vehicle 102 logs executed actions and transmits confirmation signals to the anomaly detection system 104, allowing remote monitoring of executed responses. The ability of the connected vehicle 102 to autonomously execute corrective actions enables timely responses to detected anomalies, reducing system downtime and mitigating potential hazards. The continuous interaction between the connected vehicle 102 and the anomaly detection system 104 enhances real-time decision-making and operational control.
In an embodiment, a time series data record comprises telemetry data representing various operational parameters of the connected vehicle 102. The telemetry data comprises at least one of motor speed, control surface servo position, linear velocity, orientation, or angular velocity. The time series data record is generated in real-time by onboard sensors, electronic control units, and monitoring systems embedded within the connected vehicle 102. The time series data record provides a continuous stream of data points that reflect changes in vehicle dynamics over time. The motor speed parameter captures the rotational speed of propulsion components, providing insight into engine efficiency and power output. The control surface servo position measures adjustments in steering, braking, or aerodynamic control surfaces, enabling detection of mechanical deviations. The linear velocity parameter represents the translational motion of the connected vehicle 102 along a given trajectory, aiding in speed regulation and navigation analysis. The orientation parameter determines the spatial alignment of the connected vehicle 102 in relation to a reference axis, assuring accurate positioning for navigation and stability control. The angular velocity parameter captures the rate of rotation around a given axis, allowing real-time assessment of manoeuvrability and dynamic behavior. The continuous collection of telemetry data within the time series data record enables predictive maintenance, real-time diagnostics, and early detection of abnormal performance trends. The time series data record serves as the foundation for anomaly detection, providing essential inputs for comparative analysis against baseline operational models. The integration of multiple telemetry parameters into the time series data record improves the accuracy and reliability of anomaly detection processes.
In an embodiment, the anomaly detection system 104 determines an anomaly threshold value based on historical data associated with a connected vehicle 102 and updates the anomaly threshold value using real-time data received from the connected vehicle 102. The anomaly detection system 104 maintains a repository of historical telemetry data, allowing pattern analysis and identification of expected operational ranges for various vehicle parameters. The anomaly detection system 104 applies statistical models and historical trend analysis to establish an initial anomaly threshold value, defining acceptable deviation limits for received telemetry data. The anomaly detection system 104 continuously monitors incoming time series data records from the connected vehicle 102, comparing real-time operational parameters against previously established thresholds. When persistent deviations are observed, the anomaly detection system 104 recalibrates the anomaly threshold value, refining anomaly detection criteria to account for evolving vehicle conditions. The anomaly detection system 104 dynamically adjusts anomaly detection sensitivity to prevent excessive false positive detections while maintaining high accuracy in identifying critical faults. The anomaly detection system 104 integrates environmental data, road conditions, and vehicle load variations into anomaly threshold updates. The anomaly detection system 104 transmits updated anomaly threshold values to the connected vehicle 102 and associated monitoring units to synchronize detection criteria across multiple operational components. The continuous refinement of anomaly threshold values enhances system adaptability, allowing proactive identification of anomalies while minimizing unnecessary corrective interventions.
In an embodiment, the anomaly detection system 104 trains a machine learning model using a dataset that comprises telemetry data corresponding to at least two of a start portion of a trip, a driving portion of a trip, and an end portion of a trip. The anomaly detection system 104 applies a weighting mechanism to identify anomalous time series data records to be excluded during processing. The anomaly detection system 104 assigns a time series data record weight to each record in a training dataset and updates the time series data record weights based on anomaly frequency. The anomaly detection system 104 alternates between optimizing a set of fitting weights and optimizing a set of time series data record weights, wherein anomalous records are excluded during optimization. The anomaly detection system 104 processes the time series data record using a multi-stage anomaly detection approach that comprises filtering, feature extraction, and classification. The anomaly detection system 104 adjusts the anomaly threshold value using a feedback mechanism based on prior anomaly detection accuracy. The anomaly detection system 104 periodically updates the machine learning model using additional time series data collected from the connected vehicle 102. The training dataset comprises vehicle operation data captured across varying trip conditions, making sure that the machine learning model accounts for dynamic driving scenarios. The anomaly detection system 104 prevents bias in training by excluding outliers and assigning adaptive weights to telemetry records. The continuous training and optimization process enhances anomaly detection accuracy and adaptability to evolving vehicle conditions.
In an embodiment, the action to address an anomaly comprises at least one of rescheduling a future trip or enabling remote control by an external operator. The anomaly detection system 104 determines the severity of a detected anomaly and assesses whether immediate intervention is required. If an anomaly does not pose an immediate safety risk, the anomaly detection system 104 evaluates operational schedules and determines whether rescheduling a future trip can mitigate potential disruptions. The anomaly detection system 104 modifies trip schedules, updates route assignments, and notifies relevant personnel or automated systems about rescheduled travel plans. If an anomaly requires immediate intervention, the anomaly detection system 104 initiates remote control capabilities, allowing an external operator to assume command of the connected vehicle 102. The anomaly detection system 104 establishes a secure communication link between the connected vehicle 102 and the external operator, enabling real-time control transfer without latency disruptions. The external operator receives telemetry data updates and executes corrective actions, including adjusting navigation routes or initiating emergency stop procedures. The anomaly detection system 104 continuously monitors vehicle status during remote operation, enabling safe execution of external commands. The anomaly detection system 104 records all executed actions, generating event logs for post-incident analysis and system optimization. The ability to reschedule trips or enable remote control minimizes operational risks, preventing unnecessary downtime while maintaining vehicle safety and performance.
In an embodiment, the connected vehicle 102 executes an action by at least one of navigating to an emergency repair location or performing an immediate halting procedure. The anomaly detection system 104 continuously monitors real-time telemetry data and assesses whether a detected anomaly affects vehicle components. If the anomaly indicates a mechanical or system failure that requires immediate attention, the anomaly detection system 104 determines a suitable emergency repair location and transmits navigation instructions to the connected vehicle 102. The connected vehicle 102 integrates the received instructions into the onboard navigation system and initiates autonomous or assisted routing to the designated repair facility. The connected vehicle 102 communicates with external traffic management systems to optimize travel paths and avoid congestion while navigating to the repair location. If the anomaly poses an immediate safety risk, the anomaly detection system 104 transmits a command to perform an immediate halting procedure. The connected vehicle 102 processes the received command and engages braking mechanisms while considering road conditions, surrounding vehicles, and passenger safety. The connected vehicle 102 activates warning signals, including hazard lights and alert notifications, to notify nearby traffic and external monitoring systems about the emergency stop. The connected vehicle 102 records executed actions and transmits confirmation signals to the anomaly detection system 104. The anomaly detection system 104 evaluates post-execution telemetry data to assess whether additional interventions are necessary. The ability of the connected vehicle 102 to autonomously navigate to a repair location or execute emergency stopping procedures reduces risks associated with system failures while maintaining operational safety.
In an embodiment, the anomaly detection system 104 generates an alert when an anomaly score exceeds an anomaly threshold value and receives a confirmation signal from a connected vehicle 102 upon execution of an action. The anomaly detection system 104 continuously processes telemetry data transmitted from the connected vehicle 102, calculating an anomaly score based on observed deviations from baseline operational patterns. When the anomaly score exceeds the anomaly threshold value, the anomaly detection system 104 triggers an alert, notifying monitoring systems, fleet operators, or onboard control units of the detected anomaly. The anomaly detection system 104 classifies the alert based on severity, assigning priority levels to assure timely corrective action. The anomaly detection system 104 transmits the alert through multiple communication channels, including direct vehicle messaging, cloud-based notifications, or integration with external emergency response systems. The connected vehicle 102 processes the received alert and executes the specified corrective action, adjusting operational parameters or initiating emergency response mechanisms as required. Upon completion of the corrective action, the connected vehicle 102 generates a confirmation signal, transmitting execution details back to the anomaly detection system 104. The anomaly detection system 104 records confirmation signals, validating that the executed action aligns with the recommended response. The anomaly detection system 104 utilizes confirmation data to refine future anomaly response strategies, assuring continuous improvement in automated detection and intervention processes. The automated alerting and confirmation mechanism supports efficient anomaly management, reducing system failures and optimizing response time.
In an embodiment, the anomaly detection system 104 compares a detected anomaly with a geographically indexed anomaly database to identify external factors influencing the anomaly. The anomaly detection system 104 maintains a database containing historical anomaly occurrences indexed by geographic location, allowing correlation between detected anomalies and location-specific conditions. The database comprises records of environmental factors such as road quality, weather conditions, terrain variations, and traffic congestion, as well as recurring vehicle performance deviations reported in specific regions. When the anomaly detection system 104 identifies an anomaly, the system retrieves relevant data from the geographically indexed anomaly database and assesses whether external conditions have contributed to the detected deviation. The anomaly detection system 104 analyzes location-based trends to determine whether the detected anomaly aligns with patterns observed in similar geographic locations. The anomaly detection system 104 enhances anomaly classification accuracy by distinguishing between anomalies caused by internal vehicle faults and those influenced by external environmental conditions. The anomaly detection system 104 dynamically updates anomaly detection criteria based on geographic trends, affirming contextual accuracy in anomaly classification. The anomaly detection system 104 integrates external data sources, such as weather monitoring systems and road infrastructure databases, to refine geographic anomaly detection models. The anomaly detection system 104 generates location-specific recommendations to mitigate the impact of anomalies linked to environmental factors. The anomaly detection system 104 transmits geographic anomaly assessments to fleet management systems, traffic authorities, and vehicle control systems, supporting real-time adjustments to vehicle operations based on geographic anomaly trends.
In an embodiment, the method 200 for detecting anomalies in vehicle data comprises receiving, by an anomaly detection system 104 communicatively linked to a connected vehicle 102, a time series data record from the connected vehicle 102. The anomaly detection system 104 processes the received time series data record using a machine learning model to generate an anomaly score. The anomaly detection system 104 determines whether the anomaly score exceeds an anomaly threshold value by comparing detected deviations with pre-established operational baselines. Upon determining that the anomaly score exceeds the anomaly threshold value, the anomaly detection system 104 classifies the detected anomaly and assesses the impact on vehicle performance. The anomaly detection system 104 determines an action to address the detected anomaly by evaluating predefined corrective measures based on historical response data. The anomaly detection system 104 transmits a command to the connected vehicle 102 specifying the determined action for execution. The connected vehicle 102 receives the transmitted command and executes the specified action, such as modifying operational parameters, activating safety mechanisms, or transmitting alerts to external monitoring systems. The anomaly detection system 104 records all detected anomalies, computed anomaly scores, executed actions, and post-action system responses in an operational log. The anomaly detection system 104 continuously refines anomaly detection models using historical data and real-time feedback to enhance predictive accuracy. The anomaly detection system 104 supports automated anomaly identification and corrective action execution, reducing system failures and optimizing vehicle performance under varying operational conditions.
In an embodiment, the anomaly detection system 104 assigns a severity classification to a detected anomaly based on at least one of deviation magnitude from expected parameters, frequency of occurrence, and real-time operational conditions of a connected vehicle 102. The anomaly detection system 104 evaluates the extent of deviation from predefined operational baselines, assigning severity levels based on the magnitude of observed deviations. The anomaly detection system 104 analyzes historical occurrence patterns of similar anomalies, determining whether the detected anomaly represents a recurring issue or an isolated event. The anomaly detection system 104 considers real-time operational parameters, such as vehicle load, environmental conditions, and system diagnostics, to refine severity classification. The anomaly detection system 104 categorizes anomalies as low-risk, moderate-risk, or high-risk, prioritizing response measures accordingly. The anomaly detection system 104 transmits severity classifications to fleet operators, remote monitoring systems, and vehicle control units, enabling informed decision-making. The anomaly detection system 104 prioritizes high-risk anomalies by initiating immediate corrective actions or alerting emergency response teams. The anomaly detection system 104 dynamically updates severity classification models using real-time feedback and historical anomaly resolution outcomes.
In an embodiment, the anomaly detection system 104 adjusts an anomaly threshold value based on historical anomaly detection data and real-time operational parameters of a connected vehicle 102. The anomaly detection system 104 maintains a repository of past anomaly occurrences, allowing pattern recognition and refinement of detection criteria. The anomaly detection system 104 evaluates long-term trends in telemetry data to recalibrate anomaly threshold values, affirming adaptability to evolving vehicle conditions. The anomaly detection system 104 prevents excessive false positive detections by refining anomaly detection sensitivity based on observed operational variations. The anomaly detection system 104 integrates external data sources, including manufacturer specifications, environmental monitoring systems, and predictive maintenance reports, to enhance anomaly threshold adjustments. The anomaly detection system 104 transmits updated anomaly threshold values to the connected vehicle 102 and associated monitoring systems, assuring consistency in detection criteria across operational components. The anomaly detection system 104 logs all threshold recalibrations for compliance tracking and future anomaly detection model optimization.
In an embodiment, the anomaly detection system 104 prioritizes execution of a determined action based on an operational safety risk assessment corresponding to a detected anomaly. The anomaly detection system 104 evaluates the potential impact of a detected anomaly on vehicle performance, passenger safety, and surrounding traffic conditions. The anomaly detection system 104 calculates a risk score by analyzing deviation severity, likelihood of system failure, and potential consequences of delayed corrective action. The anomaly detection system 104 applies predefined prioritization rules to determine whether immediate intervention is required or if corrective action can be scheduled. The anomaly detection system 104 assigns high priority to anomalies affecting critical vehicle components, such as braking, propulsion, and steering systems. The anomaly detection system 104 initiates emergency response protocols for high-risk anomalies, transmitting immediate intervention commands to the connected vehicle 102 and notifying external monitoring systems. The anomaly detection system 104 dynamically adjusts prioritization criteria based on real-time traffic conditions, road hazards, and environmental influences. The anomaly detection system 104 logs all prioritization decisions and executed actions for post-event analysis and system optimization.
In an embodiment, the anomaly detection system 104 determines a recommended corrective measure based on a correlation between a detected anomaly and historical corrective actions stored in a remote database. The anomaly detection system 104 maintains a repository of previously detected anomalies, associated corrective actions, and post-correction vehicle performance data. The anomaly detection system 104 retrieves historical records and identifies recurring anomaly patterns, selecting effective corrective measures based on past resolutions. The anomaly detection system 104 correlates anomaly characteristics, including affected telemetry parameters and severity classification, with successful historical interventions. The anomaly detection system 104 transmits recommended corrective measures to the connected vehicle 102 or external operators for execution. The anomaly detection system 104 updates the remote database with new corrective action outcomes, enabling continuous refinement of response strategies. The anomaly detection system 104 integrates predictive maintenance insights and external diagnostic reports to enhance corrective measure selection.
In an embodiment, the anomaly detection system 104 assigns a probability score to a detected anomaly based on historical occurrence patterns, environmental conditions, and operational parameters of a connected vehicle 102. The anomaly detection system 104 continuously monitors incoming telemetry data and applies probability analysis to assess whether a detected anomaly represents a system fault or a transient deviation. The anomaly detection system 104 calculates a probability score by analyzing the frequency of past occurrences of similar anomalies, identifying recurring patterns in operational data. The anomaly detection system 104 evaluates external influences, such as road conditions, weather variations, and traffic density, to determine whether environmental factors contribute to the detected anomaly. The anomaly detection system 104 refines probability scores dynamically by integrating real-time operational telemetry data and historical failure analysis. The anomaly detection system 104 transmits probability scores to fleet management systems and external monitoring platforms, enabling predictive diagnostics and proactive decision-making. The anomaly detection system 104 continuously updates probability models based on evolving vehicle performance data.
FIG. 3 illustrates a class diagram of the system 100 for detecting anomalies in vehicle data, in accordance with the embodiments of the present disclosure. The system 100 comprises a connected vehicle 102 and an anomaly detection system 104. The connected vehicle 102 interacts with the anomaly detection system 104 by transmitting time series data, which comprises telemetry parameters such as motor speed and orientation. The anomaly detection system 104 consists of a computing device, which comprises at least one processor and a non-transitory computer-readable medium storing executable instructions. The anomaly detection system 104 processes the received time series data using a machine learning model to generate an anomaly score. If the anomaly score exceeds a predefined threshold, the anomaly detection system 104 determines an appropriate action and transmits a command to the connected vehicle 102 to execute the corrective action. The machine learning model assists in analyzing data and predicting anomalies, improving real-time decision-making. The computing device executes instructions through the processor and stores data in the non-transitory medium for future analysis
FIG. 4 illustrates a sequential process for detecting anomalies in vehicle data, in accordance with the embodiments of the present disclosure. The system 100 for detecting anomalies in vehicle data uses a connected vehicle 102 and an anomaly detection system 104. The connected vehicle 102 transmits a time series data record to the anomaly detection system 104, which processes the received data using a machine learning model to generate an anomaly score. The anomaly detection system 104 then evaluates whether the anomaly score exceeds a predefined anomaly threshold value. If the anomaly score remains within acceptable limits, no further action is taken. However, if the anomaly score exceeds the anomaly threshold value, the anomaly detection system 104 determines an appropriate corrective action based on the detected anomaly. Once the corrective action is determined, the anomaly detection system 104 transmits a command to the connected vehicle 102, specifying the corrective action for execution. The connected vehicle 102 processes the received command and executes the prescribed action to mitigate the detected anomaly. The automated sequence enables real-time anomaly detection, decision-making, and corrective action execution, enabling continuous monitoring and response to deviations in vehicle performance.

In the description of the present invention, it is also to be noted that, unless otherwise explicitly specified or limited, the terms “disposed,” “mounted,” and “connected” are to be construed broadly, and may for example be fixedly connected, detachably connected, or integrally connected, either mechanically or electrically. They may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meaning of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Modifications to embodiments and combination of different embodiments of the present disclosure described in the foregoing are possible without departing from the scope of the present disclosure as defined by the accompanying claims. Expressions such as “comprising”, “comprising”, “incorporating”, “have”, “is” used to describe and claim the present disclosure are intended to be construed in a non- exclusive manner, namely allowing for items, components or elements not explicitly described also to be present. Reference to the singular is also to be construed to relate to the plural where appropriate.
Although embodiments have been described with reference to a number of illustrative embodiments thereof, it should be understood that numerous other modifications and embodiments can be devised by those skilled in the art that will fall within the spirit and scope of the principles of this disclosure. More particularly, various variations and modifications are possible in the component parts and/or arrangements of the subject combination arrangement within the scope of the present disclosure, the drawings and the appended claims. In addition to variations and modifications in the component parts and/or arrangements, alternative uses will also be apparent to those skilled in the art.
,CLAIMS:WE CLAIM:
1. A system 100 to detect the anomalies in vehicle data, comprising:
a connected vehicle 102; and
an anomaly detection system 104 communicatively linked to the connected vehicle 102, wherein the anomaly detection system 104 comprises at least one computing device comprising at least one processor and a non-transitory computer-readable medium storing instructions that, when executed by the at least one processor, cause the anomaly detection system 104 to:
receive, new time series data record from the connected vehicle 102;
process, the received new time series data record using a machine learning model to generate an anomaly score; and
upon determination, the anomaly score exceeds an anomaly threshold value:
determine an action to address the anomaly; and
transmit a command to the connected vehicle 102 specifying the determined action to be executed.
2. The system 100 of claim 1, wherein the connected vehicle 102 is configured to:
receive, the command from the anomaly detection system 104; and
execute, the action specified in the command.
3. The system 100 of claim 1, wherein the time series data record comprises telemetry data selected from at least one of a motor speed, a control surface servo position, a linear velocity, an orientation, or an angular velocity.
4. The system 100 of claim 1, wherein the anomaly detection system 104 determines the anomaly threshold value based on historical data associated with the connected vehicle 102, and updates the anomaly threshold value using real-time data received from the connected vehicle 102.
5. The system 100 of claim 1, wherein the anomaly detection system 104 trains the machine learning model using a dataset that comprises telemetry data corresponding to at least two of a start portion of a trip, a driving portion of a trip, and an end portion of a trip, wherein the training comprises:
applying a weighting mechanism to identify anomalous time series data records to be excluded during processing;
assigning a time series data record weight to each record in a training dataset and updating the time series data record weights based on anomaly frequency;
alternating between optimizing a set of fitting weights and optimizing a set of time series data record weights, wherein the anomalous records are excluded during optimization;
processing the time series data record using a multi-stage anomaly detection approach that comprises filtering, feature extraction, and classification;
adjusting the anomaly threshold value using a feedback mechanism based on a prior anomaly detection accuracy; and
periodically updating the machine learning model using additional time series data collected from the connected vehicle 102.
6. The system 100 of claim 1, wherein the action to address the anomaly comprises at least one of rescheduling a future trip or enabling remote control by an external operator.
7. The system 100 of claim 1, wherein the connected vehicle 102 executes the action by at least one of navigating to an emergency repair location or performing an immediate halting procedure.
8. The system 100 of claim 1, wherein the anomaly detection system 104 generates an alert when the anomaly score exceeds the anomaly threshold value, and receives a confirmation signal from the connected vehicle 102 upon execution of the action.
9. The system 100 of claim 1, wherein the anomaly detection system 104 compares the detected anomaly with a geographically indexed anomaly database to identify the external factors influencing the anomaly.
10. A method 200 for detecting the anomalies in vehicle data, comprising:
receiving, by an anomaly detection system 104 communicatively linked to a connected vehicle 102, a new time series data record from the connected vehicle 102;
processing, by the anomaly detection system 104, the received new time series data record using a machine learning model to generate an anomaly score;
determining, by the anomaly detection system 104, whether the anomaly score exceeds an anomaly threshold value;
upon determining, the anomaly score exceeds the anomaly threshold value:
determining, by the anomaly detection system 104, an action to address the detected anomaly; and
transmitting, by the anomaly detection system 104, a command to the connected vehicle 102 specifying the determined action to be executed.
11. The method 200 of claim 10, wherein the anomaly detection system 104 assigns a severity classification to the detected anomaly based on at least one of deviation magnitude from the expected parameters, a frequency of occurrence, and the real-time operational conditions of the connected vehicle 102.
12. The method 200 of claim 10, wherein the anomaly detection system 104 adjusts the anomaly threshold value based on historical anomaly detection data and the real-time operational parameters of the connected vehicle 102.
13. The method 200 of claim 10, wherein the anomaly detection system 104 prioritizes execution of the determined action based on an operational safety risk assessment corresponding to the detected anomaly.
14. The method 200 of claim 10, wherein the anomaly detection system 104 determines a recommended corrective measure based on a correlation between the detected anomaly and the historical corrective actions stored in a remote database.
15. The method 200 of claim 10, wherein the anomaly detection system 104 assigns a probability score to the detected anomaly based on the historical occurrence patterns, the environmental conditions, and the operational parameters of the connected vehicle 102.