Device, system, and method to generate recommendations for fixing malfunctions in a vehicle

Abstract

The present invention discloses a device (100), a system (300), and a method (200) for generating recommendations for predicting malfunctions in a vehicle (500). The invention includes a device (100) for generating recommendations after detecting vehicle (500) malfunctions (500). The device (100) comprises a control unit (101) with a transceiver (103) receiving sensed parameters from vehicle sensors (102) and user inputs. The processors (101-1) within the control unit (101) analyze the combined data to create the datasets to identify faults and predict potential malfunctions. The control unit (101) can further employ a fault identification model (104) like machine learning or artificial intelligence models to aid the analysis. Based on the analysis, the device (100) generates recommendations for users to address potential vehicle (500) malfunctions.

Description

DEVICE, SYSTEM, AND METHOD TO GENERATE RECOMMENDATIONS FOR FIXING MALFUNCTIONS IN A VEHICLETECHNICAL FIELD

[0001] The embodiments of the present disclosure generally relate to the field of vehicle diagnostics and maintenance, specifically to systems and methods implemented for providing maintenance recommendations to fix the issues in a vehicle. More particularly, the present disclosure relates to the device, system, and method to generate one or more recommendations (“hereinafter recommendations”) to fix the malfunctions in the vehicle.BACKGROUND

[0002] Background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.

[0003] The existing technologies include diagnostic scan tools, telematics systems, predictive maintenance software, and remote diagnostics platforms. The diagnostic scan tools utilize OBD-II ports to gather data and diagnose issues related to engine performance and emissions. Telematics systems combine GPS and onboard diagnostics to monitor vehicle performance and provide remote diagnostics capabilities. Predictive maintenance software analyzes data from various sensors to predict maintenance needs, reducing downtime and preventing costly repairs. Remote diagnostics platforms enable technicians to access vehicle data remotely for diagnosis and software updates. However, the existing technologies have limitations such as limited coverage, dependency on data quality, high costs, complexity requiring specialized training, privacy concerns related to data collection, and reliance on historical data for predictive maintenance, which may not always be accurate.

[0004] Further, the present invention addresses the challenges posed by existing technologies as below.

[0005] Unlike traditional diagnostic tools that rely solely on specific error codes or sensor data, the symptomatic detection framework considers a wide range of symptoms exhibited by the vehicle including mechanical components. This approach allows for a more holistic and comprehensive diagnosis, reducing the risk of overlooking potential issues.

[0006] Unlike telematics systems that raise concerns about privacy and data security due to continuous monitoring and data collection, the symptomatic detection framework operates on a symptom-based approach, minimizing the collection and transmission of sensitive information about vehicle usage and performance. Overall, the present invention offers a solution to address the challenges faced by existing technologies in vehicle diagnostics and maintenance, providing a more comprehensive, adaptable, cost-effective, and privacypreserving approach to ensuring vehicle safety and reliability.

[0007] A prior-art reference “US 20,220,108,569 Al” titled “Automated detection of vehicle data manipulation and mechanical failure" provides systems and methods to detect and identify vehicular anomalies. The system disclosed in the prior art includes a plurality of sensors, one or more processors, a system memory to store the instructions to cause the system to: receive a plurality of signals from the plurality of sensors that contains vehicular data, convert the group the plurality of signals into a plurality of detection sets, detect an anomaly within at least one detection set of the plurality of detection sets based on a comparison of signals in the at least one detection set to a normal behavior model, and crossreference the at least one detection set with at least one other detection set of the plurality of detection sets to identify a source of the anomaly.

[0008] Whereas, the present disclosure discloses a device, a system, and a method for generating the recommendations after detecting malfunctions in a vehicle. The invention includes a device generating repair recommendations based on the detected vehicle malfunctions and. The device comprises a control unit with a transceiver receiving sensed parameters from vehicle sensors and user inputs. The processors within the control unit analyze the combined data to create the datasets to identify faults and predict potential malfunctions. The control unit can further employ a fault identification model like machine learning or artificial intelligence models to aid the analysis. Based on the analysis, the device generates recommendations for users to address potential vehicle malfunctions.

[0009] Another prior-art reference KR 101,974,347 Bl titled “Fault diagnosis system for vehicle and data security method thereof’ discloses a vehicle failure diagnosis system and a diagnostic method to generate a failure prediction reference data by machine learning the vehicle status and the data collected from a plurality of vehicles, and predicting failure in each vehicle. The system disclosed in the prior art is a diagnostic server. A diagnostic server collects data on the state of a part of the vehicle and whether there is a failure from a plurality of vehicles to generate failure prediction reference data for predicting a failure of the vehicle.It is installed in the vehicle and collects data on the status of each part of the vehicle and whether or not there is a failure to provide to the diagnostic server, and compares the failure prediction reference data provided from the diagnostic server with the data on the vehicle, the failure of the vehicle Among the expected devices or parts, a vehicle diagnostic device for displaying failure prediction information on a device or part having a predicted failure time below a predetermined schedule on a display of the own vehicle.

[0010] On the other hand, the present disclosure discloses a device, a system, and a method for generating the recommendations for fixing malfunctions in a vehicle. The invention includes a device for generating recommendations after detecting vehicle malfunctions. The device comprises a control unit with a transceiver receiving sensed parameters from vehicle sensors and user inputs. The processors within the control unit analyze the combined data to create the datasets to identify faults and predict potential malfunctions. The control unit can further employ a fault identification model like machine learning or artificial intelligence models to aid the analysis. Based on the analysis, the device generates recommendations for users to address potential vehicle malfunctions.

[0011] Thus, there exists a dire need in the art, to provide a device, a system, and a method for generating repair recommendations for fixing malfunctions in the vehicle.OBJECTS OF INVENTION

[0012] Some of the objects of the present disclosure, that at least one embodiment herein satisfies are as listed herein below.

[0013] It is a general object of the present disclosure to overcome the drawbacks and limitations of the existing vehicle diagnostics systems.

[0014] It is an object of the present disclosure to provide a device, a system, and a method to generate the recommendations for fixing the malfunctions in the vehicle.

[0015] It is another object of the present invention to simplify the diagnostic process by focusing on observable symptoms rather than complex data analysis or specialized training.

[0016] It is another object of the present disclosure to provide a system that utilizes a fault identification model to predict the likelihood of a malfunction occurring within the vehicle.

[0017] It is another object of the present disclosure to generate maintenance recommendations to users to identify and thereby fix the issues that occurs in the vehicle.

[0018] It is another object of the present disclosure to provide a more comprehensive, adaptable, cost-effective, and privacy-preserving solution for ensuring vehicle safety and reliability.SUMMARY

[0019] Within the scope of this application, it is expressly envisaged that the various aspects, embodiments, examples, and alternatives set out in the preceding paragraphs, in the claims and / or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. Features described in connection with one embodiment are applicable to all embodiments, unless such features are incompatible.

[0020] Aspects of the invention relate to a device, a system, and a method for generating recommendations to fix the malfunctions present in the vehicle. The present disclosure includes the transceiver to collect sensory symptoms perceivable through human senses such as auditory, tactile, visual, olfactory, and additional data inputs from one or more sensors (hereinafter "sensors") embedded in the vehicle. The sensors capture one or more parameters (hereinafter "parameters") of the vehicle including engine noises, brake squeals, vibrations, exhaust emissions, battery voltage, engine load percentage, and other relevant parameters. The collected sensory inputs and sensed parameters of the vehicles are processed to identify and classify symptoms indicative of potential vehicle issues. A fault identification model analyses patterns and anomalies in the inputs to classify symptoms based on their severity and relevance to a particular vehicle part’s malfunction. The fault identification model uses machine learning (ML), artificial intelligence (Al) techniques or other advanced techniques to predict the likelihood of a malfunction occurring within the vehicle.

[0021] Various objects, features, aspects, and advantages of the inventive subject matter will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which like numerals represent like components.BRIEF DESCRIPTION OF DRAWINGS

[0022] The accompanying drawings, which are incorporated herein, and constitute a part of this invention, illustrate exemplary embodiments of the disclosed methods and systems in which like reference numerals refer to the same parts throughout the different drawings. Components in the drawings are not necessarily to scale, emphasis instead being placed uponclearly illustrating the principles of the present invention. Some drawings may indicate the components using block diagrams and may not represent the internal circuitry of each component. It will be appreciated by those skilled in the art that invention of such drawings includes the invention of electrical components, electronic components or circuitry commonly used to implement such components.

[0023] FIG. 1 illustrates an exemplary block diagram of the device to generate recommendations for fixing malfunctions in the vehicle, in accordance with an exemplary embodiment of the present disclosure.

[0024] FIG. 2 illustrates an exemplary flow diagram that describes the step-wise illustration of the method for generating repair recommendations to fix the malfunctions of the vehicle, in accordance with an embodiment of the present disclosure.

[0025] FIG. 3 illustrates an exemplary block diagram of the system to generate recommendations for fixing malfunctions in the vehicle, in accordance with an exemplary embodiment of the present disclosure.

[0026] FIG. 4 illustrates the exemplary representation of the components of the system for vehicle inspections, in accordance with an exemplary embodiment of the present disclosure.

[0027] Other objects, advantages, and novel features of the invention will become apparent from the following more detailed description of the present embodiment when taken in conjunction with the accompanying drawings.LIST OF REFERENCE NUMERALS100- Device101- Control unit101-1- One or more processors102- One or more sensors103- Transceiver104- Fault identification model300- System500- VehicleDETAILED DESCRIPTION

[0028] Various example embodiments will now be described more fully with reference to the accompanying drawings in which only some example embodiments are shown. Specific structural and functional details disclosed herein are merely representative for purposes of describing example embodiments. The present invention, however, may be embodied in many alternate forms and should not be construed as limited to only the example embodiments set forth herein.

[0029] Accordingly, while example embodiments of the invention are capable of various modifications and alternative forms, embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that there is no intent to limit example embodiments of the present invention to the particular forms disclosed. On the contrary, example embodiments are to cover all modifications, equivalents, and alternatives falling within the scope of the invention. Like numbers refer to like elements throughout the description of the figures.

[0030] It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments of the present invention. As used herein, the term “and / or,” includes any and all combinations of one or more of the associated listed items.

[0031] Various methods described herein may be practiced by combining one or more machine-readable storage media containing the code / instruction according to the present invention with appropriate standard device hardware to execute the instruction contained therein. An apparatus for practicing various embodiments of the present invention may involve one or more computers (say server) (or one or more processors within a single computer) and storage systems containing or having network access to computer program(s) coded in accordance with various methods described herein, and the method steps of the invention could be accomplished by modules, devises, routines, subroutines, or subparts of a computer program product.

[0032] Thus, for example, it will be appreciated by those of ordinary skill in the art that the diagrams, schematics, illustrations, and the like represent conceptual views or processesillustrating systems and methods embodying this invention. The functions of the various elements shown in the figures may be provided through the use of dedicated hardware as well as hardware capable of executing associated software. Similarly, any switches shown in the figures are conceptual only. Their function may be carried out through the operation of program logic, through dedicated logic, through the interaction of program control and dedicated logic, or even manually, the particular technique being selectable by the entity implementing this invention. Those of ordinary skill in the art further understand that the exemplary hardware, software, processes, methods, and / or operating systems described herein are for illustrative purposes and, thus, are not intended to be limit the present disclosure.

[0033] The following is a detailed description of embodiments of the disclosure depicted in the accompanying drawings. The embodiments are in such detail as to clearly communicate the disclosure. However, the amount of detail offered is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure as defined by the appended claims.

[0034] Accordingly, an embodiment of the present disclosure, a system, and a method for is disclosed.

[0035] Hereinafter, exemplary embodiments of the present disclosure will be described with reference to the accompanying drawings FIG. 1-4.

[0036] FIG. l illustrates an exemplary block diagram of the device (100) to generate recommendations for fixing malfunctions in the vehicle (500), in accordance with an exemplary embodiment of the present disclosure.

[0037] The present disclosure relates to a device for generating repair recommendations after detecting vehicle malfunctions. In these embodiments, the device (100) comprises below components:

[0038] The device (100) comprises a control unit (101). The control unit (101) comprises a transceiver (103) and one or more processors (101-1). The transceiver (103) is configured to receive two types of data - sensed parameters of the vehicle (500) sensed by the sensors (102) and sensory inputs obtained from the users.

[0039] Sensed parameters of the vehicle (500) data originate from various sensors (102) installed within the vehicle (500) at predetermined locations. These sensors (102)continuously monitor and capture various aspects of the vehicle's health, such as engine revolutions per minute (RPM), coolant temperature, oil pressure, air intake temperature, exhaust gas temperature, and battery voltage (all reference numeral 4). The specific parameters monitored can vary depending on the make and model of the vehicle (500).

[0040] Sensory input data comes directly from users. The users can provide information about their experience with the vehicle (500) through the device (100). Examples of such sensory inputs include observations of unusual noises like engine misfire or brake squeal, vibrations, exhaust smoke emission, or dashboard warning light illumination. Additionally, users can report physical sensations like unbalanced tires or abnormal wear patterns on components.

[0041] Further, one or more processors (101-1) (hereinafter “processors”) within the control unit (101) play a crucial role in analyzing the received data. Before combining the sensed parameters and sensory inputs, the processors (101-1) may perform pre-processing steps to remove any noise or inconsistencies present in the data which ensures the accuracy of the subsequent analysis.

[0042] Once pre-processing is complete, the processors (101-1) combine the two data sets i.e. sensed parameters and sensory inputs to form one or more datasets. These datasets are then analyzed by the processors (101-1). The analysis involves employing a fault identification model (104) housed within the control unit (101). The fault identification model (104) can be any or a combination of various advanced algorithms, including machine learning (ML) models, artificial intelligence (Al) models, neural networks, deep learning models, explainable Al (XAI) models, federated learning models, or reinforcement learning models. By analyzing the combined datasets, the fault identification model (104) identifies and classifies any potential faults present in the vehicle (500).

[0043] Furthermore, based on the identified faults, the processors (101-1) predict the potential occurrence of malfunctions within the vehicle (500). This prediction capability allows the device (100) to anticipate problems before they escalate into major breakdowns.

[0044] Finally, leveraging the predicted malfunctions, the processors (101-1) generate one or more recommendations for the users. The recommendations can be displayed through various user interfaces depending on the specific implementation. For instance, the recommendations might be presented on an in-vehicle display system, a dedicated mobile application, or a web interface accessible remotely. The recommendations aim to guide userstoward potential solutions for fixing the identified or predicted malfunctions in the vehicle (500).

[0045] In essence, the device (100) acts as a comprehensive diagnostic tool. It gathers data from the vehicle (500) through sensors (102) and user inputs, analyzes this data using advanced algorithms within the control unit (101), and generates actionable recommendations to address potential vehicle malfunctions. This approach can empower users to maintain their vehicles proactively and prevent major problems.

[0046] The present disclosure contemplates various embodiments of the device (100). The specific implementation details, such as the types of sensors (102) used, the chosen fault identification model (104), and the user interface for displaying recommendations, can be tailored based on the intended application and target users.

[0047] In another embodiment, the control unit (101) may include engine control unit (ECU), electronic throttle unit (ETU), telematics control unit (TCU), body control module (BCM), transmission control module (TCM), engine control module (ECM), powertrain control module (PCM), embedded device, or any other control unit that is capable to perform one or more operations in the device to detect faults in the vehicle (500) and generate the recommendations accordingly.

[0048] In these embodiments, one or more sensory inputs are received through a user interface (103) from one or more users. The user interface (103) is operatively coupled to one or more sensors (102). The sensory inputs may include- auditory symptoms, tactile symptoms, visual symptoms, and olfactory symptoms.

[0049] The auditory symptoms may include sounds such as engine knocking, brake squeals, or unusual vibrations detected by the human ear. Auditory symptoms are important as they can indicate issues like misfiring, worn brake pads, or bearing failures. The tactile symptoms may include physical sensations such as vibrations felt through the steering wheel, brake pedal, or seat. The symptoms can indicate problems like unbalanced tires, suspension issues, or engine misfires. The visual symptoms may include visual observations such as smoke emissions, dashboard warning lights, fluid leaks, or abnormal wear patterns. The symptoms can point to issues like overheating, fluid leaks, or electrical system faults. The olfactory symptoms may include unusual smells such as exhaust odors, burning smells, or the scent of leaking fluids. The symptoms can signal problems like fuel leaks, overheating components, or electrical shorts.

[0050] In these embodiments, the sensory inputs are collected and pre-processed along with data from onboard vehicle (500) sensors. Further, the collected data is cleaned to remove noise and inconsistencies, ensuring reliable input for further analysis. Identified and extracted key features from sensory and sensor data that are indicative of potential malfunctions which include temporal and frequency domain features. The collected sensory and sensor data are fused to create a unified, comprehensive dataset. The device (100) involves scaling data from different sources to a common range to ensure compatibility and accurate comparison. It combines data from various sensors to enhance the reliability and accuracy of the detected symptoms. Combining of data takes place using a set of instructions for data fusion such as “Kalman filters” or “Bayesian networks”.

[0051] Further, the integrated data undergoes analysis to identify and classify symptoms that may indicate potential vehicle issues. Using statistical methods and machine learning techniques to detect deviations from normal operating conditions. The ML techniques include clustering, principal component analysis (PCA), and autoencoders are used. Categorizes the identified anomalies into specific symptoms or fault classes based on the characteristics, using machine learning (ML) classifiers such as support vector machines (SVM), decision trees, or neural networks. The machine learning technique comprising a set of instructions is applied to the classified symptoms to predict the likelihood of specific part malfunctions. Training and deploying models on historical data and real-time inputs to recognize patterns associated with various types of malfunctions. Training models such as random forests, gradient boosting machines, and deep learning networks are employed to recognize the patterns.

[0052] In these embodiments, the device (100) includes the control unit (101) and comprises the fault identification model (104) coupled to the control unit (101), the Fault identification model (104) is configured to identify and classify the one or more faults present in the vehicle (500), the fault identification model is using more techniques selected from any or a combination of a machine learning (ML) model, an artificial intelligence (Al) model, a neural networks, a deep learning model, an explainable Al (XAI), federated learning model, or a reinforcement learning model, or other techniques capable to identify and classify the one or more faults of the vehicle (500).

[0053] In these embodiments, the device (100) suggests specific inspection or replacement actions for vehicle (500) parts identified as at risk of malfunction. The suggestions include prioritizing tasks based on the severity and urgency of the predictedissues. The device (100) offers detailed insights and guidance to users such as technicians or vehicle owners to help them prioritize and schedule maintenance tasks. The device (100) may provide visual dashboards, alerts, and reports to aid decision-making.

[0054] In these embodiments, the device (100) integrates historical data to enhance prediction accuracy and maintenance recommendations. The device (100) utilizes historical diagnostic trouble codes (DTCs) to identify recurring issues and predict future malfunctions. The device (100) analyzes past maintenance activities and their outcomes to refine the prediction models and improve recommendation accuracy.

[0055] In these embodiments, one or more connecting channels for users to access the device (100) may include various communication methods including but not limited to, In- vehicle connections through vehicle display and controls, dedicated diagnostic app, cellular network connection through mobile app or web interface, or telematics integration through telematics system installed in the vehicle (500).

[0056] In an embodiment, the device may be used for various applications such as engine performance monitoring tools, brake analysis systems, and battery heal monitoring.

[0057] In an embodiment, the device (100) may be used for an engine performance monitoring tool. The device (100) uses sensory inputs such as engine knocking sounds, vibrations felt through the steering wheel, and sensor data such as engine load percentage, and exhaust gas temperature. The device (100) identifies knocking sounds and excessive vibrations as symptoms of potential fuel combustion issues. Elevated engine load and exhaust gas temperatures further corroborate the likelihood of malfunction. The device (100) may provide recommendations such as inspection and adjustment of fuel injection parameters, engine timing, or cooling system components. The recommendations may include actions such as checking for fuel injector clogs, adjusting ignition timing, or inspecting the coolant levels and radiator condition.

[0058] In another embodiment, the device (100) may be used for a brake analysis system. The device (100) captures sensory inputs such as brake squeals, pulsations in the brake pedal, and sensor data such as brake fluid level, and brake pad wear sensor readings. The device (100) detects symptoms such as brake squeals and pedal pulsations as indicators of brake pad wear or rotor damage. Low brake fluid levels and worn brake pad sensors suggest imminent brake system failure. The device (100) generates recommendations that may include but are not limited to, inspection and replacement of worn brake components, bleeding the brakesystem, and checking for hydraulic leaks. The specific actions that may be taken by the user include replacing brake pads and rotors, inspecting brake lines for leaks, and ensuring proper brake fluid levels.

[0059] In another embodiment, the device (100) may be used for battery heal monitoring. The device (100) obtains sensory inputs such as unusual smells indicating battery acid leakage, and sensor data such as battery voltage, current, temperature, and state of charge. The device (100) identifies the smell of leaking acid and low voltage as symptoms of potential battery health issues. Irregular charging patterns and temperature fluctuations indicate internal cell degradation. The device (100) may generate recommendations including but not limited to, the inspection and potential replacement of the battery, checking charging system components for faults, and conducting a battery load test. Specific actions may be taken by the user such as cleaning battery terminals, testing the alternator output, and ensuring that the battery is securely mounted and properly ventilated.

[0060] In a nutshell, the device (100) represents a significant advancement in vehicle diagnostics, combining human sensory inputs with sophisticated machine-learning techniques to predict part malfunctions accurately. By integrating sensory data with sensor readings and employing advanced data processing and predictive analytics, the device (100) provides comprehensive and actionable maintenance recommendations. The proposed device (100) enhances vehicle safety, reliability, and performance, offering a robust solution for proactive vehicle maintenance.

[0061] In an embodiment, the device (100) uses various parameters involved in collecting, processing, and analyzing data to identify potential vehicle malfunctions. The parameters may include but not limited to auditory symptoms, tactile symptoms, visual symptoms, olfactory symptoms, sensor data, environmental conditions, historical data, realtime metrics, diagnostic codes, and machine-learning features. The parameters and their workable ranges are as below.

[0062] Auditory symptoms include engine noises ranging between 50-100 decibels (dB) depending on engine type and condition. The brake squeals range in between 70-90 dB during braking, may vary with brake pad material, and abnormal vibrations may be detected through tactile feedback, varying in intensity. Tactile symptoms include steering wheel vibrations that can be detectable at frequencies between 5-10 Hz. Visual symptoms can provide important clues about vehicle issues such as smoke emissions ranging from light grayto black can indicate varying levels of engine or exhaust system problems. Olfactory symptoms may include unusual odors, even at low concentrations (parts per million, ppm), and can sometimes indicate problems depending on the specific odor and individual sensitivity.

[0063] The monitoring of various sensor data can help diagnose vehicle performance. The engine temperature typically ranges from 80°C to 110°C during normal operation. Idle speed ranges from 600-1000 RPM, while maximum RPM varies by engine type. The brake fluid level should be within the optimal range between the minimum and maximum markers on the reservoir. Battery voltage is typically 12-14 volts when the engine is running, dropping to 10-12 volts when the engine is off. Exhaust emissions can be measured in parts per million (ppm) for gases such as carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NOx).

[0064] Further, environmental factors can also impact vehicle performance. Ambient temperature can vary widely, typically ranging from -20°C to 40°C, depending on location and season. Humidity levels, ranging from 0% to 100%, can affect air density and combustion efficiency. Road conditions, such as surface type, traction, and obstacles, can also vary depending on location and weather. Historical data such as maintenance records, including time intervals between service events, mileage, and types of maintenance performed, as well as previous fault codes, can provide valuable insights into the vehicle's history and potential issues. Monitoring real-time metrics can help diagnose vehicle performance. Engine load percentage varies from 0% (idle) to 100% (full throttle). The throttle position ranges from 0% (closed throttle) to 100% (wide open throttle). Fuel consumption rates are typically measured in liters per hour (L / h) or gallons per hour (GPH).

[0065] The diagnostic codes such as error codes, and specific alphanumeric codes corresponding to different types of faults or malfunctions, can be used to diagnose vehicle issues. Further, the advanced diagnostic techniques may utilize various machine learning features, such as statistical features (mean, standard deviation, skewness, kurtosis, etc.), frequency analysis (dominant frequencies, spectral power density, etc.), and time-series patterns such as trends, cycles, periodicity, etc.

[0066] The above parameters range of the vehicle (500) are general guidelines and may vary depending on specific vehicle models, sensor types, environmental conditions, and other factors.

[0067] In these embodiments, the collected sensory and digital data is processed to identify and classify symptoms indicative of potential vehicle issues. Machine learning algorithms analyze patterns and anomalies in the inputs to classify symptoms based on their severity and relevance to a particular vehicle part’s malfunction.

[0068] In these embodiments, the noisy data may include any information received by the device (100) that can potentially mislead the analysis and generation of recommendations. The noisy data may originate from the components such as sensors (102) or user inputs. Further, noisy data may include electrical interference, sensor malfunction, environmental factors, user inputs such as misinterpretation of sounds, inaccurate descriptions, memory lapses or any other additional data comes from the received parameters associated with the vehicle (500) and the one or more sensory inputs obtained from the users.

[0069] FIG. 2 illustrates an exemplary flow diagram that describes the step-wise illustration of the method (200) for generating repair recommendations to fix the malfunctions of the vehicle (500), in accordance with an embodiment of the present disclosure.

[0070] The method (200) leverages a combination of vehicle (500) sensor data and user inputs to identify potential problems and suggest solutions. Below is the illustration of the method (200) with various steps included.

[0071] At block 201, a transceiver (103) within the device (100) receives one or more sensed parameters associated with the vehicle (500). These parameters are captured by various sensors (102) installed at predetermined locations within the vehicle (500).

[0072] At block 202, following the receipt of sensed parameters, the method (200) proceeds to the receiving step. In the block 202, the transceiver (103) also receives one or more sensory inputs from users. The sensory inputs represent user observations or experiences with the vehicle (500).

[0073] At block 203, once both sensed parameters and sensory inputs are received, the method (200) progresses to the combining step. Here, one or more processors (101-1) within the device (100) combine the received data sets. This combined data set serves as the basis for further analysis.

[0074] At block 204, the method (200) then proceeds to the analyzing step. During this step, one or more processors (101-1) analyze the combined data set to identify potential faults within the vehicle (500).

[0075] At block 205, following the analysis, the method (200) progresses to the predicting step. Here, leveraging the identified faults, the processors (101-1) predict the potential occurrence of malfunctions within the vehicle (500). This predictive capability allows for proactive identification of problems before they escalate into major breakdowns.

[0076] At block 206, finally, the method (200) culminates in the generating step. The processors (101-1) generate one or more recommendations for the user(s) based on the predicted malfunctions. The recommendations can be displayed through various user interfaces depending on the specific implementation. For instance, the recommendations might be presented on an in-vehicle display system, a dedicated mobile application, or a web interface accessible remotely. The recommendations guide users toward potential solutions for fixing the identified or predicted malfunctions in the vehicle (500).

[0077] Further, the method (200) can incorporate a pre-processing of data at block 207, before combining the data sets. At block 207, the processors (101-1) perform operations to remove any noise or inconsistencies present in the received parameters and sensory inputs. The pre-processing step helps ensure the accuracy of the subsequent analysis.

[0078] In a nutshell, the method (200) provides a systematic approach to vehicle diagnostics. It gathers data from both the vehicle (500) itself through embedded one or more sensors and the user, analyzes the data using advanced techniques, and generates actionable recommendations to address potential vehicle issues. The method (200) can empower users to maintain their vehicles proactively and prevent major problems.

[0079] In these embodiments, ML (machine learning) based diagnostic method follows a four-stage process to correlate symptoms and sensor data, extract meaningful patterns, and predict faults.Stage 1 : Multi-Symptom Correlation ModelStage 2: Symptom-Sensor Matching via Pearson’s Correlation CoefficientStage 3: Fault Prediction via Random Forest ML ModelStage 4: Gradient Boosting Model for fault severity estimation

[0080] Stage 1 : Multi-Symptom Correlation ModelHuman-reported symptoms often exhibit interdependencies (e.g., "engine misfire" and "rough idling" are strongly correlated). A robust model is needed to quantify and use these relationships in predictive analytics.A Bayesian Network (BN) is constructed where symptoms act as nodes, and edges represent probabilistic dependencies.Bayesian Network FormulationLet S- , S2, S3... Snbe the set of reported symptoms. The joint probability distribution is defined as: (s Parents(S;))Where Parents (S;) denotes the set of direct dependencies for symptom St..ExampleIf a user reports “high engine vibration” (SI), the system estimates probabilities for related symptoms:P(S2= "rough idling" | S1= "high vibration") = 0.85 This helps in refining the diagnostic model by adding latent symptoms that users might not explicitly report.

[0081] Stage 2: Symptom-Sensor Matching via Pearson’s Correlation CoefficientTo validate the relationship between driver-reported symptoms and sensor deviations, Pearson’s Correlation Coefficient (r) is calculated. The formula to calculate r is as under:Where:• x = Symptom-derived feature (e.g., Steering Pull Deviation)• y = Sensor data (e.g., ABS Activation)Example: Steering Pull and Brake System CorrelationUsing Pearson’s Correlation, we calculate: r = 0.86 (Strong Positive Correlation)Since r >0.7, the system validates the symptom-sensor relationship and assigns it a higher weight in the Random Forest model.

[0082] Stage 3: Fault Prediction via Random Forest ML ModelA Random Forest (RF) classifier is trained on historical fault records, where features include:• Symptoms (S 1, S2, . . . ): Mapped to numerical values via one-hot encoding.• Sensor Readings (XI, X2, . . .): Real-time vehicle telemetry.Each decision tree in the Random Forest splits data using the Gini impurity criterion:Where pkis the proportion of samples belonging to class kRandom Forest Algorithm Implementation1. Data Pre-processing: Convert categorical symptoms into numerical values and standardize sensor data.2. Bootstrapping: Create multiple random subsets of the dataset.3. Tree Construction: Train multiple decision trees using different feature sets.4. Voting Mechanism: Aggregate predictions from all trees to determine the final classification.ExampleA vehicle exhibits high vibration + fluctuating RPM + ignition delay. The classifier predicts:P(Fault = “Spark Plug Failure”) = 0.87 indicating a likely issue with the ignition system.Advantages of Random Forest in Fault Prediction• Handles Non-Linearity: Can identify complex relationships between symptoms and sensor readings.• Feature Importance Analysis: Helps identify the most critical symptoms and sensor readings affecting fault detection.• Resistant to Overfitting: Since multiple trees contribute to the decision, the model remains stable even with noise in the data.

[0083] Stage 4: Fault Severity Estimation via Gradient BoostingThe Fault Severity Estimation stage in the present invention’s vehicle diagnostic method plays a crucial role in determining how critical a detected fault is, thereby enabling predictive maintenance recommendations. The present invention employs Gradient Boosting as a key machine learning technique in this stage, leveraging its ability to improve predictive accuracy by sequentially correcting errors in previous models.Once a fault is identified using multi-symptom correlation and the Random Forest classifier, the next step is to determine:• Severity Level: How critical is the fault? (Minor, Moderate, Severe, or Critical)• Urgency: Whether immediate attention is required or if the repair can be scheduled later.• Impact Analysis: The potential consequences of ignoring the fault over time.This estimation helps vehicle owners make informed decisions about repair urgency and cost.Gradient Boosting (GB) is chosen for fault severity estimation because:1. Handles Complex Non-Linearity: Fault severity depends on multiple interacting factors such as: o Sensor readings (vibrations, temperature, emissions) o Categorical symptoms (user-reported issues) o Historical failure data o Vehicle-specific parameters (age, brand, model)Gradient Boosting efficiently captures these relationships.. Improves Predictive Accuracy Over Time: Unlike traditional models, GB builds trees sequentially, where each new tree corrects the errors of the previous ones, leading to high accuracy.3. Handles Imbalanced Data Well: Some faults (e.g., catastrophic engine failure) occur rarely but are highly severe. GB assigns higher importance to such cases.4. Robust against Noise: Vehicle data often has variability due to driving conditions and user-reported inaccuracies. GB minimizes the impact of such inconsistencies.The input features for severity estimation are categorized into:• Numerical Sensor Data (quantitative): o Engine temperature, RPM, oil pressure, fuel efficiency changes, battery voltage, vibration frequencies.• Categorical Symptoms (converted into numerical values): o Noise level (low, medium, high mapped to 1, 2, 3) o Smoke color (black, white, blue mapped to different severity levels) o Braking response (normal, delayed, failed converted into numerical severity indicators)• Historical Data: o Fault resolution patterns from similar vehicle models. o Manufacturer-reported failure thresholds.The GB model follows these steps:1. Initialize with a Weak ModelThe process starts with a simple decision tree that predicts an initial severity score S0(x)S_0(x)S0(x) (say, based on sensor thresholds).2. Compute Residuals (Errors) The error between the predicted severity and actual historical severity is calculated: ft = Vi - S0(x;) where: yi = actual severity label (from historical data)So Oft)=initial model’s predictionrt= residual error to be corrected in the next iteration.3. Train New Trees on Residuals A new decision tree is trained on the residuals, aiming to minimize them.4. Update the Severity Estimation Model The next severity estimation function is:5i(x) = SQ(X) + ot -^x) where: h^x) is the new tree trained to correct residuals. a is a learning rate (controls the contribution of each tree).5. Iterate Until Convergence This process repeats for multiple iterations:Sr(x) = ST-1x) + ahT(x)After several iterations, the final function accurately predicts the severity of the fault.

[0084] FIG. 3 illustrates an exemplary block diagram of the system (300) to generate recommendations for fixing malfunctions in the vehicle (500), in accordance with an exemplary embodiment of the present disclosure.

[0085] In an embodiment, the present disclosure includes the system (300) to generate recommendations for fixing the malfunctions present in a vehicle (500). The system (300) comprises the sensors (102) installed at a pre-defined position of the vehicle (500). The sensors (102) are configured to sense one or more parameters associated with the vehicle (500). The system (300) includes a control unit (101) having a transceiver (103) coupled to the sensors (102). The transceiver (103) receives the sensed parameters associated with the vehicle (500), and receives one or more sensory inputs from the users. The processors (101- 1) are configured to- combine the one or more received parameters and the one or more received sensory inputs to obtain one or more datasets, analyze one or more datasets to identify one or more faults, predict one or more malfunctions in the vehicle (500), based on the one or more analyzed faults, and generate the one or more recommendations for the one or more users based on the one or more predicted faults of the vehicle (500).

[0086] In these embodiments, the sensors capture one or more parameters (hereinafter “parameters”) of the vehicle that includes engine noises, brake squeals, vibrations, exhaust emissions, battery voltage, engine load percentage, and other relevant parameters.

[0087] In these embodiments, the system (300) comprises one or more sensors (102) operatively coupled to the control unit (101). The sensors (102) are configured to sense one or more parameters of the vehicle (500).

[0088] In these embodiments, the sensors (102) may be selected from any or combination of engine sensors such as mass airflow (MAF) sensor, oxygen sensors (02 sensors), coolant temperature sensors, knock sensor, emission sensors such as exhaust gas temperature (EGT) sensor, nitrogen oxide (NOx) sensor, and powertrain Sensors such as throttle position sensor (TPS), vehicle speed sensor (VSS), manifold absolute pressure (MAP) sensor, breaking system sensors such as wheel speed sensor (WSS), brake pad wear sensor, and other sensors such as battery voltage sensor, oil pressure sensor, airbag sensor, rain sensor, or any other sensors capable to detect the faults in the vehicle (500). The sensors (103) may include other sensors for retrofit such as OBD-II port add-on sensors including oil analysis sensors, fuel efficiency sensors, tire pressure monitoring system (TPMS) sensors, advanced vibration sensors, in-cabin air quality sensors, or any other sensors capable to detect the faults in the vehicle (500).

[0089] FIG. 4 illustrates the exemplary representation (400) of the components of the device (100) for vehicle (500) inspections, in accordance with an exemplary embodiment of the present disclosure.

[0090] At step 1, data is collected from a variety of sensors installed within the vehicle (500). At step 2, raw sensor data is pre-processed, and features relevant to vehicle health and performance are extracted for further analysis. At step 3, the different types of sensory data are fused and integrated to create a comprehensive understanding of the vehicle's condition, combining both direct sensory inputs and derived metrics. At step 4, the fault identification model (104) analyzes the integrated sensory data to identify and classify symptoms indicative of potential vehicle issues, such as engine anomalies, brake system malfunctions, or battery health degradation. At step 5, utilizing the identified symptoms and data patterns, the device (100) predicts the likelihood of specific malfunctions occurring within the vehicle. Finally, at step 6, based on the predicted likelihood of malfunction and the severity of identified symptoms, the device (100) provides recommendations for maintenance and repair actions, assisting technicians or vehicle owners in making informed decisions to address potential issues.

[0091] In an embodiment, the device (100) may be employed in various types of vehicles for diagnosis of malfunctions including but not limited to, passenger cars (ICE and hybrid), scooters, motorcycles, trucks, or buses.

[0092] In these embodiments, communication port(s) are configured to enable communication between the device (100) and external entities. The ports may include RS- 232 ports, Ethernet ports, Gigabit ports, serial ports, parallel ports, or other interfaces tailored to networking requirements, depending on the specific network configuration. The bus acts as a communication channel that facilitates the exchange of data and instructions between the processor and other components within the system, such as memory, storage, and communication blocks. Common bus standards may encompass PCI / PCI-X, SCSI, USB, or similar protocols, ensuring compatibility and interoperability.

[0093] In these embodiments, typically, the device (100), system (300), or any module thereof is configured to communicate with one or more external computing devices, which may include a computer processor of a local computing device such as a smartphone, a tablet device, and / or a personal computer, and / or a remote computing device or a remote server, e.g., a cloud-based remote server. For example, external computing devices may receive data from the user system and may then process the data through the processing unit (104), for example, using techniques as described herein. Alternatively, or additionally, the external computing device may send data and / or instructions through the device (100). The present disclosure includes references to certain functionalities being performed by a computing device. Typically, such functionalities are performed by the processing unit, external computing devices, and / or a combination thereof.

[0094] Further, elements and / or features of different example embodiments may be combined with each other and / or substituted for each other within the scope of this disclosure and appended claims.

[0095] Example embodiments being thus described, it will be obvious that the same may be varied in many ways. Such variations are not to be regarded as a departure from the scope of the present invention, and all such modifications as would be obvious to one skilled in the art are intended to be included within the scope of the following claims.ADVANTAGES OF THE INVENTION:

[0096] The proposed invention provides a device to generate repair recommendations to fix malfunctions in the vehicle.

[0097] The proposed invention provides a device, a system, and a method implemented to facilitate users by providing recommendations to fix the issues in the vehicle.

[0098] The proposed invention provides a device that represents a significant advancement in the field of automotive diagnostics.

[0099] The proposed invention provides a system that offers a comprehensive solution for preemptive maintenance and ensuring optimal performance of vehicles in diverse operational environments.

Claims

We Claim:

1. A device (100) to generate one or more recommendations for fixing one or more malfunctions present in a vehicle (500), the device (100) comprising: a control unit (101) having: a transceiver (103) to receive one or more sensed parameters associated with the vehicle (500), and to receive one or more sensory inputs from one or more users, wherein the one or more sensory inputs are received through a user interface from one or more users, a symptomatic detection framework to consider a wide range of symptoms exhibited by the vehicle including mechanical components, wherein the symptomatic detection framework operates on a symptom-based approach, minimizing the collection and transmission of sensitive information about vehicle usage and performance, wherein the one or more sensory inputs include auditory symptoms, tactile symptoms, visual symptoms, and olfactory symptoms; a fault identification model (104) to identify and classify the one or more faults present in the vehicle (500), wherein the fault identification model is a machine learning (ML) model such as a Random Forest (RF) classifier; and one or more processors (101-1) coupled to the transceiver (103), wherein the one or more processors (101-1) are configured to: combines the one or more received parameters and the one or more received sensory inputs to obtain one or more datasets, wherein a Bayesian Network (BN) is configured to calculate a joint probability distribution of said combined parameters and is defines as:where symptoms act as nodes, and edges represent probabilistic dependencies and where Parents (S,) denotes the set of direct dependencies for symptom St.; analyse the one or more datasets to identify the one or more faults, wherein the Random Forest (RF) classifier is configured to identify the one or more faults, Random Forest (RF) classifier is trained on historical fault records, where featuresinclude symptoms and sensor readings, wherein each decision tree in the Random Forest splits data using the Gini impurity criterion:Where pkis the proportion of samples belonging to class k; predict the one or more malfunctions in the vehicle (500), based on the one or more analysed faults; and generate the one or more recommendations for the one or more users based on the one or more predicted malfunctions of the vehicle (500), wherein Gradient Boosting is employed to determine the criticality of the one or more predicted malfunctions, thereby enabling predictive maintenance recommendations.

2. The device (100) as claimed in claim 1, wherein the fault identification model is selected from any or a combination of a machine learning (ML) model, an artificial intelligence (Al) model, a neural network, a deep learning model, an explainable Al (XAI), federated learning model, or a reinforcement learning model.

3. The device (100) as claimed in claim 1, wherein: the one or more parameters associated with the vehicle (500) are sensed by one or more sensors (102) installed at a pre-defined position of the vehicle (500); and the one or more received parameters and the one or more received sensory inputs are pre-processed, by the one or more processors (101-1), before combining the one or more received parameters and the one or more received sensory inputs, to remove one or more noisy data.

4. The device (100) as claimed in claim 1, wherein the one or more sensed parameters of the vehicle (500) are selected from any or combination of engine revolutions per minute (RPM), coolant temperature, oil pressure, air intake temperature, exhaust gas temperature, engine noise, brake squeal, vibration, exhaust emission, battery voltage, engine load percentage, or any other parameters which are capable to predict the one or more malfunctions in the vehicle (500).

5. The device (100) as claimed in claim 1, wherein the one or more sensory inputs received from the one or more users are selected from any, or combination of misfiring, wornbrake pads, bearing failure, physical sensations, unbalanced tire, suspension issue, engine misfire, smoke emission, dashboard warning light, fluid leak, abnormal wear pattern, unusual smells such as exhaust odors, burning smells, or the scent of leaking fluid, or any other sensory input that is capable to sense by the one or more users through one or more senses.

6. The device as claimed in claim 1, wherein the one or more users are present inside the vehicle (500), at a surrounding of the vehicle (500), or remotely located having access to the device (100) through one or more connecting channels.

7. The device (100) as claimed in claim 1, wherein the one or more predicted malfunctions are one or more issues present in the vehicle (500), wherein the one or more predicted malfunctions are selected from any or combination of engine malfunction, transmission malfunction, brake malfunction, or any other malfunctions that are capable to identify through the fault identification model (104).

8. A method (200) for generating one or more recommendations to fix one or more malfunctions present in a vehicle (500), the method (200) comprising: receiving (201), through a transceiver (103), one or more sensed parameters of the vehicle (500), wherein the one or more parameters associated with the vehicle (500) are sensed by one or more sensors (102) installed at a pre-defined position of the vehicle (500); receiving (202), through the transceiver (103), one or more sensory inputs from one or more users, wherein the one or more sensory inputs are received through a user interface from one or more users, considering, through a symptomatic detection framework, a wide range of symptoms exhibited by the vehicle including mechanical components, wherein the symptomatic detection framework operates on a symptom-based approach, minimizing the collection and transmission of sensitive information about vehicle usage and performance, wherein the one or more sensory inputs include auditory symptoms, tactile symptoms, visual symptoms, and olfactory symptoms; combining (203), through one or more processors (101-1), the one or more received parameters of the vehicle (500) and the one or more received sensory inputs to obtain one or more datasets, wherein a Bayesian Network (BN) is configured to calculate a joint probability distribution of said combined parameters and is defines as:nP s1,s2. S„) = ]~^ P(s Parents(SD) i=l where symptoms act as nodes, and edges represent probabilistic dependencies and where Parents(Sj) denotes the set of direct dependencies for symptom Stanalysing (204), through the one or more processors (101-1), the one or more datasets to identify the one or more faults, wherein the Random Forest (RF) classifier is configured to identify the one or more faults, Random Forest (RF) classifier is trained on historical fault records, where features include symptoms and sensor readings, wherein each decision tree in the Random Forest splits data using the Gini impurity criterion:Where pkis the proportion of samples belonging to class k; predicting (205), through the one or more processors (101-1), the one or more malfunctions in the vehicle (500), based on the one or more analysed faults; and generating (206), through the one or more processors (101-1), the one or more recommendations for the one or more users based on the one or more predicted malfunctions of the vehicle (500), wherein Gradient Boosting is employed to determine the criticality of the one or more predicted malfunctions, thereby enabling predictive maintenance recommendations .

9. The method (200) as claimed in claim 8, wherein the method (200) comprising: pre-processing (207), by one or more processors (101-1), the one or more received parameters and one or more received sensory inputs to remove one or more noisy data.

Images