Description:METHOD AND SYSTEM FOR DETERMINING FIRE-PRONE CONDITIONS FOR VEHICLES
DESCRIPTION
Technical Field
[001] This disclosure relates generally to Machine Learning (ML), and more particularly to a method and system for determining a fire-prone condition in a vehicle.
Background
[002] Automobiles are complicated machines and include both mechanical and electronics parts. The automobiles have an abundance of frictional components, flammable substances, and complicated electric wiring. Therefore, fire is one of the most frequent risks associated with automobiles. Hence, fire safety becomes one of the critical aspects for a vehicle owner. Traditionally, sensors are used for fire detection. These sensors respond to various effects resulting from fire, and consequently generate output signals proportional to such effects. However, the sensors detect such effects when a flame or fire occurs in the vehicle. As a result of the subsequent fire and its suppression, the engine compartment components or the other components get damaged. Even though the existing sensors determine fire occurrence, but lack in determining conditions which may lead to a fire. Further, any inaccuracy in detection of fire by the sensors may lead to irreparable loss.
[003] The present invention is directed to overcome one or more limitations stated above or any other limitations associated with the known arts.
SUMMARY
[004] In one embodiment, a method for determining a fire-prone condition in a vehicle is disclosed. In one example, the method may include receiving real-time data corresponding to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle. The plurality of pre-defined parameters may correspond to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle. The method may further include generating a real-time condition vector based on the real-time data using a machine learning (ML) model. The method may further include determining a deviation in the real-time condition vector from a safe-condition vector. It should be noted that the ML model may be trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle. The method may further include determining the fire-prone condition in the vehicle based on the deviation and a predefined threshold deviation.
[005] In another embodiment, a system for determining a fire-prone condition in a vehicle is disclosed. In one example, the system may include a monitoring device. The monitoring device may include a processor and a memory communicatively coupled to the processor. The memory may store processor-executable instructions, which, on execution, may cause the processor to receive real-time data corresponding to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle. The plurality of pre-defined parameters may correspond to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle. The processor-executable instructions, on execution, may further cause the processor to generate a real-time condition vector based on the real-time data using a machine learning (ML) model. The processor-executable instructions, on execution, may further cause the processor to determine a deviation in the real-time condition vector from a safe-condition vector. It should be noted that the ML model may be trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle. The processor-executable instructions, on execution, may further cause the processor to determine the fire-prone condition in the vehicle based on the deviation and a predefined threshold deviation.
[006] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed. Indicatively
BRIEF DESCRIPTION OF THE DRAWINGS
[007] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, serve to explain the disclosed principles.
[008] FIG. 1 illustrates an environmental diagram of a system for determining a fire-prone condition in a vehicle, in accordance with some embodiments of the present disclosure.
[009] FIG. 2 illustrates a block diagram of various modules within a monitoring device configured to determine a fire-prone condition in a vehicle, in accordance with some embodiments of the present disclosure.
[010] FIG. 3 illustrates a flow diagram of a method for determining a fire-prone condition in a vehicle, in accordance with some embodiments of the present disclosure.
[011] FIG. 4 illustrates a flow diagram of a method for generating an alert, in accordance with some embodiments of the present disclosure.
[012] FIG. 5 illustrates a flow diagram of a detailed method for generating an alert in response to determining a fire-prone condition, in accordance with some embodiments of the present disclosure.
[013] FIG. 6 illustrates an exemplary system for generating an alert upon determining a fire-prone condition in a vehicle, in accordance with some embodiments of the present disclosure.
[014] FIG. 7 illustrates an exemplary Machine Learning (ML) model trained to determine a fire-prone condition in a vehicle, in accordance with some embodiments of the present disclosure.
DETAILED DESCRIPTION
[015] The following description is presented to enable a person of ordinary skill in the art to make and use the invention and is provided in the context of particular applications and their requirements. Various modifications to the embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the invention. Moreover, in the following description, numerous details are set forth for the purpose of explanation. However, one of ordinary skill in the art will realize that the invention might be practiced without the use of these specific details. In other instances, well-known structures and devices are shown in block diagram form in order not to obscure the description of the invention with unnecessary detail. Thus, the invention is not intended to be limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and features disclosed herein.
[016] While the invention is described in terms of particular examples and illustrative figures, those of ordinary skill in the art will recognize that the invention is not limited to the examples or figures described. Those skilled in the art will recognize that the operations of the various embodiments may be implemented using hardware, software, firmware, or combinations thereof, as appropriate. For example, some processes can be carried out using processors or other digital circuitry under the control of software, firmware, or hard-wired logic. (The term “logic” herein refers to fixed hardware, programmable logic and/or an appropriate combination thereof, as would be recognized by one skilled in the art to carry out the recited functions.) Software and firmware can be stored on computer-readable storage media. Some other processes can be implemented using analog circuitry, as is well known to one of ordinary skill in the art. Additionally, memory or other storage, as well as communication components, may be employed in embodiments of the invention.
[017] Referring now to FIG. 1, an environmental diagram of a system 100 for determining a fire-prone condition in a vehicle is illustrated, in accordance with some embodiment of the present disclosure. The system 100 detects a condition which may eventually result in a fire occurrence. To detect this type of condition, the system 100 monitors various parameters associated with fire-prone zones of the vehicles. Further, upon detection of the condition, the system 100 may generate an alert. The system 100 may include a monitoring device 102. In particular, the system 100 may implement the monitoring device 102 so as to determine a fire-prone condition in vehicles. As will be appreciated, the monitoring device 102 may be any computing device (for example, a server, a desktop, a laptop, a notebook, a netbook, a tablet, a smartphone, a mobile phone, a vehicle dashboard, or the like).
[018] The monitoring device 102 may include one or more processors 104, a computer readable medium (for example, a memory) 106, and an input/output (I/O) device 108. The computer readable medium 106 may store instructions that, when executed by the processors 104, may cause the processors 104 to determine the fire-prone condition in the vehicles. As will be described in greater detail in conjunction with FIG. 2 to FIG. 7, in order to determine the fire-prone condition, the processor 104 in conjunction with the computer readable medium 106 may perform various functions including receiving data, determining real-time condition vector, determining deviation, generating alerts, and the like.
[019] The computer-readable medium 106 may also store various data (for example, historical safe-condition vector, historical safe-condition data, real-time condition vector, real-time data, values related to various parameters, or the like) that may be captured, processed, and/or required by the monitoring device 102. The computer-readable medium 106 may be a non-volatile memory (e.g., flash memory, Read Only Memory (ROM), Programmable ROM (PROM), Erasable PROM (EPROM), Electrically EPROM (EEPROM) memory, etc.) or a volatile memory (e.g., Dynamic Random Access Memory (DRAM), Static Random-Access memory (SRAM), etc.).
[020] The monitoring device 102 may interact with a user via a user interface accessible via the I/O devices 108. The user, or an administrator may interact with the monitoring device 102 and vice versa through the I/O device 108. The I/O device 108 may include a display, and the user interface. By way of an example, the display may be used to display results of analysis performed by the monitoring device 102 (such as, for displaying notifications to alert and the like), to the user. By way of another example, the user interface may be used by the user to provide inputs to the monitoring device 102.
[021] The monitoring device 102 may also interact with one or more external devices, a vehicle 110, or a server 114 over a communication network 112 for sending or receiving various data. The external devices may include, but may not be limited to, a remote server, a digital device, or another computing system. Examples of the vehicle 110 may include, but are not limited to, motor vehicles, motorcycles, cars, trucks, buses, mobility scooters, railed vehicles, watercraft, amphibious vehicles, aircraft, and the like. The vehicle 110 may include various sensors 110a, which may be used to acquire data associated with various parameters corresponding to fire-prone zones of the vehicle 110. Examples of the sensors 110a may include, temperature sensors, pressure sensors, speed sensors, torque sensors, load sensors, exhaust sensors, Tire Pressure Monitoring System (TPMS) sensors, cabin sensors, Heating Ventilation and Air Conditioning (HVAC) sensors, engine sensors, and transmission sensors. As will be appreciated, the load sensors may be electrical load sensors (current sensors, voltage sensors, etc.) or engine load sensors (i.e., a Throttle Position Sensor (TPS), a Mass Air Flow (MAF) sensor, a Manifold Air Pressure (MAP) sensor, etc.).
[022] The communication network 112, for example, may be any wired or wireless communication network and the examples may include, but may be not limited to, the Internet, Wireless Local Area Network (WLAN), Wi-Fi, Long Term Evolution (LTE), Worldwide Interoperability for Microwave Access (WiMAX), and General Packet Radio Service (GPRS). By way of an example, in some embodiments, the monitoring device 102 may receive information from the vehicle 110 or the server 114. The server 114 may further include a database, which may store information.
[023] Referring now to FIG. 2, various modules within a monitoring device 200 (same as the monitoring device 102) configured to determine the fire-prone condition in the vehicle (for example, the vehicle 110) are illustrated, in accordance with some embodiments of the present disclosure. FIG. 2 is explained in conjunction with FIG. 1. In order to determine the fire-prone condition, the monitoring device 200 may receive sensor data 202 from various sensors associated with the vehicle. The sensors may include temperature sensors, pressure sensors, speed sensors, torque sensors, and load sensors. The temperature sensor may monitor temperature parameters, for example, an engine coolant temperature, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and an ambient temperature. The pressure sensor may monitor pressure related parameters, such as a barometric pressure, a fuel injection pressure, and an engine oil pressure. The speed sensors may monitor speed related parameters, such as a vehicle speed. In some embodiments, the load sensors may be engine load sensors (e.g., TPS sensor, MAF sensor, MAP sensor, etc.) that monitor load on the engine of the vehicle. Further, in some embodiments, the load sensors may be electrical load sensors (i.e., a current sensor, a voltage sensor, etc.) that monitor electrical load on various electrical components within the vehicle.
[024] Further, various operations may be performed based on the sensor data 202 by the monitoring device 200. To perform the operations, the monitoring device 200 may include a vector generation module 204, a deviation determination module 206, a fire-prone condition determination module 208, and an alert generation module 210. Further, the monitoring device 200 may also include a data store 212 to store various data and intermediate results generated by the modules 204-210.
[025] The vector generation module 204 may be configured to receive the sensor data 202. In some embodiments, the sensor data 202 may be referred to as real-time data. Thus, the vector generation module 204 may receive the real-time data corresponding to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle. The plurality of pre-defined parameters may correspond to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle.
[026] It should be noted that the plurality of predefined parameters may include the engine coolant temperature, the ambient temperature, the barometric pressure, the fuel injection pressure, the engine oil pressure, the engine torque, the engine load, the vehicle speed, the exhaust after-treatment temperature, the temperature-rise during a regen cycle, and a radiator fan operational time.
[027] Further, in some embodiments, the vector generation module 204 may be configured to generate a real-time condition vector based on the sensor data 202 or the real-time data. In particular, the vector generation module 204 may include a Machine Learning (ML) model 204a or associated with the ML model 204a to generate the real-time condition vector. The real-time condition vector may be generated by an encoder of the ML model in a hidden embedding space. Further, the vector generation module 204 may transmit the real-time condition vector to communicatively coupled to the deviation determination module 206 and/or the data store 212.
[028] In one embodiment, the deviation determination module 206 may receive the real-time condition vector from the vector generation module 204. In another embodiment, the deviation determination module 206 may extract the real-time condition vector from the data store 212. In one embodiment, the deviation determination module 206 may extract a safe-condition vector from the data store 212. Further, the deviation determination module 206 may be configured to determine a deviation in the real-time condition vector from the safe-condition vector. Further, the deviation determination module 206 may be communicatively coupled to the datastore 212 and fire-prone condition determination module 208.
[029] It should be noted that the ML model 204a may be trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle. Also, it should be noted that the ML model 204a may be an encoder-decoder model. Examples of the encoder-decoder model may include, but are not limited to, a Recurrent Neural Network (RNN) encoder-decoder model, a Long short-term memory (LSTM) encoder-decoder model, and a Convolutional Neural Network (CNN) encoder-decoder model. In some embodiments, the encoder of the ML model 204a may be trained with the historical safe-condition data to generate the safe-condition vector in a hidden embedding space. In some other embodiments, the encoder may be trained to generate the real-time condition vector based on the real-time data. Further, a decoder of the ML model 204a may be trained to validate if the safe-condition vector corresponds to the historical safe-condition data. An exemplary ML model is explained further in detail in conjunction with Fig. 7.
[030] The fire-prone condition determination module 208 may be configured to determine the fire-prone condition in the vehicle. It should be noted that the fire-prone condition determination module 208 may consider a predefined threshold deviation for determining the fire-prone condition. In other words, when the deviation in the real-time condition vector exceeds the predefined threshold deviation, the condition may be considered as the fire-prone condition by the fire-prone condition determination module 208. Further, upon determining the fire-prone condition, the fire-prone condition determination module 208 may transmit a signal to the alert generation module 210. The alert generation module 210 may generate an alert for a user/administrator 214 (e.g., owner, driver, service station, etc.) associated with the vehicle. The alert is provided to the user/administrator 214 by the alert generation module 210 in a predefined time window based on the time of occurrence of the fire. The alert may be a visual alert, an alarm, a text or a voice message. Further, the alert may be provided to at least one of a driver of the vehicle so that the driver may come out of the vehicle, an emergency contact (for example, to a family member of the driver), a nearest service station with location coordinates of the vehicle so that the vehicle may be towed for checking, and emergency service providers (for example, fire station, highway patrol unit, etc.) in case the service station is far away.
[031] In some embodiments, a confidence score may also be generated based on the deviation. Based on the confidence score, chances of occurrence of the fire may be determined. Further, in some embodiments, frequency of transmitting alerts may be changed based on the confidence score.
[032] By way of an example, consider a situation where a radiator fan of an engine is not working properly which may be due to any of a blown-out fuse, a faulty relay or a broken wire. In that case, heat may be prevented from escaping an engine compartment and the temperature of the engine may rise. In that case, the monitoring device 200 may receive sensor data including data captured by temperature sensor, generate a real-time condition vector based on the sensor data, determine a deviation in the real-time condition vector from the safe-condition vector, and based on the deviation determine a fire-prone condition. The fire-prone condition may be detected as a value of a parameter associated with the engine may not be optimal and/or out of a safe limit. Further, the monitoring device 200 may send an alert to the driver as soon as the fire-prone condition is detected, so that the driver may take an action accordingly. For example, the alert may be sent within few seconds of the detection of the fire-prone condition.
[033] As will be appreciated by one skilled in the art, a variety of processes may be employed for determining the fire-prone condition in the vehicle. For example, the system 100 and associated monitoring device 102 or 200 may determine the fire-prone condition by the processes discussed herein. In particular, as will be appreciated by those of ordinary skill in the art, control logic and/or automated routines for performing the techniques and steps described herein may be implemented by the system 100 and associated monitoring device 102, or 200 either by hardware, software, or combinations of hardware and software. For example, suitable code may be accessed and executed by the system 100 and the associated monitoring device 102, or 200, to perform some or all of the techniques described herein. Similarly, application specific integrated circuits (ASICs) configured to perform some, or all of the processes described herein may be included in the system 100 and the associated monitoring device 102, 200.
[034] Referring now to FIG. 3, a method 300 for determining a fire-prone condition in a vehicle is depicted via a flow diagram, in accordance with some embodiments of the present disclosure. FIG. 3 is explained in conjunction with FIG. 1-2. All the steps 302-308 of the flow diagram may be performed by the monitoring device 102, 200 of the system 100.
[035] At step 302, real-time data may be received. The real-time data may correspond to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle. For example, the one or more fire-prone zones may include zones related to battery area, engine compartment, connections in the vehicle. It should be noted that the plurality of pre-defined parameters may correspond to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle.
[036] For example, the plurality of predefined parameters may include an engine coolant temperature, an ambient temperature, a barometric pressure, a fuel injection pressure, an engine oil pressure, an engine torque, an engine load, a vehicle speed, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and a radiator fan operational time. In some embodiments, the real-time data corresponding to the plurality of predefined parameters may be acquired through a plurality of sensors. The plurality of sensors may include temperature sensors, pressure sensors, speed sensors, torque sensors, and load sensors.
[037] At step 304, a real-time condition vector may be gendered based on the real-time data using the Machine Learning (ML) model 204a. This step may be performed using the vector generation module 204. At step 306, a deviation in the real-time condition vector from a safe-condition vector may be determined using the deviation determination module 206. The ML model 204a may be an encoder-decoder model. Examples of the encoder-decoder model may include, but are not limited to, a Recurrent Neural Network (RNN) encoder-decoder model, a Long short-term memory (LSTM) encoder-decoder model, and a Convolutional Neural Network (CNN) encoder-decoder model. It may be noted that The ML model 204a may be trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle.
[038] In particular, an encoder of the ML model 204a may be trained with the historical safe-condition data to generate the safe-condition vector in a hidden embedding space and to generate the real-time condition vector based on the real-time data. Further, a decoder of the ML model 204a may be trained to validate if the safe-condition vector corresponds to the historical safe-condition data. Moreover, the ML model 204a may be updated from time-to time at vehicle service centers.
[039] At step 308, the fire-prone condition in the vehicle may be determined based on the deviation and a predefined threshold deviation. To perform this step the fire-prone condition determination module 208 may be employed. For example, a condition may be considered as the fire-prone condition if the deviation in the real-time condition vector is greater than the predefined deviation, otherwise the condition is a safe condition.
[040] Referring now to FIG. 4, a method 400 for generating an alarm is depicted via a flow diagram, in accordance with some embodiments of the present disclosure. FIG. 4 is explained in conjunction with FIG. 3. At step 402, the fire-prone condition may be determined based on a deviation in a real-time condition vector and a predefined threshold deviation. It should be noted that the fire-prone condition may be determined through the fire-prone condition determination module 208. Determination of the fire-prone condition has already been explained in conjunction with FIG. 3.
[041] Further, at step 404, an alert may be generated upon determining the fire-prone condition for a user associated with the vehicle using the alert generation module 210. It should be noted that the alert is provided in a predefined time window based on the time of occurrence of the fire. For example, the alert may be a visual alert, an alarm, a text or a voice message and provided to at least one of a driver of the vehicle so that the driver may come out of the vehicle, an emergency contact (for example, to a family member of the driver), a nearest service station with location coordinates of the vehicle so that the vehicle may be towed for checking, and emergency service providers (for example, fire station, highway patrol unit, etc.) in case the service station is far away.
[042] Further, in some embodiments, a confidence score may be determined based on the deviation. By way of an example, when the confidence score is ‘80’, then there may be high chances of occurrence of possible fire. Additionally, in some embodiments, frequent alerts may be transmitted until a fault within the vehicle is cleared and the deviation in the real-time condition vector from the safe-condition vector drops to or below the predefined threshold.
[043] Referring now to FIG. 5, a method 500 for generating an alarm in response to determining a fire-prone condition is depicted via a flow diagram, in accordance with some embodiments of the present disclosure. FIG. 5 is explained in conjunction with FIGs. 1-4. All the steps 502-512 of the flow diagram may be performed by the monitoring device 102, 200 of the system 100.
[044] At step 502, real-time data corresponding to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle may be received using the vector generation module 204. It should be noted that the plurality of pre-defined parameters may correspond to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle. By way of an example, the plurality of predefined parameters may include an engine coolant temperature, an ambient temperature, a barometric pressure, a fuel injection pressure, an engine oil pressure, an engine torque, an engine load, a vehicle speed, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and a radiator fan operational time.
[045] Thereafter, a real-time condition vector may be gendered based on the real-time data using the ML model 204a of the vector generation module 204, at step 504. At step 506, a deviation in the real-time condition vector from a safe-condition vector may be determined using the deviation determination module 206. The ML model may be an encoder-decoder model. Examples of the encoder-decoder model may include, but are not limited to, a Recurrent Neural Network (RNN) encoder-decoder model, a Long short-term memory (LSTM) encoder-decoder model, and a Convolutional Neural Network (CNN) encoder-decoder model. It may be noted that the ML model may be trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle.
[046] At step 508, a condition whether the deviation in the real-time vector is greater than a threshold deviation may be checked. Further, at step 510, when the condition is true, the condition may be considered as a fire-prone condition and an alert may be generated for a user associated with the vehicle. Otherwise, at step 512, when the condition is false, the condition may be considered a safe condition, and the process may terminate.
[047] Referring now to FIG. 6, an exemplary system 600 for generating an alert upon determining a fire-prone condition in a vehicle 602 is illustrated, in accordance with some embodiments of the present disclosure. FIG. 6 is explained in conjunction with FIGs. 1-5. The vehicle 602 is a car. Further, the car may include various sensors. For example, in some embodiments, the car may include temperature sensors 602a, pressure sensors 602b, speed sensors 602c, torque sensors 602d, and load sensors 602e. The temperature sensor may monitor temperature parameters, for example, an engine coolant temperature, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and an ambient temperature. The pressure sensor may monitor pressure related parameters, such as a barometric pressure, a fuel injection pressure, and an engine oil pressure. The speed sensors may monitor speed related parameters, such as a vehicle speed. In some embodiments, the load sensors may be engine load sensors (e.g., a TPS sensor, an MAF sensor, an MAP sensor, etc.) that monitor load on the engine of the vehicle. Further, in some embodiments, the load sensors may be electrical load sensors (i.e., a current sensor, a voltage sensor, etc.) that monitor electrical load on various electrical components within the vehicle.
[048] By way of an example, currently, values corresponding to one or more of parameters associated with a battery, an engine, and connections of the vehicle 602 are not optimal or not acceptable safe values due to a fault. Further, the sensors responsible for monitoring the parameters may continuously monitor the parameters associated with fire-prone zones of the vehicle 602 and acquire the values corresponding to the parameters as real-time data. The real-time data may correspond to sensor data 604.
[049] Further, the sensor data 604 may be transmitted to a monitoring device 606. In some embodiments, the monitoring device 606 may be within the vehicle 602. Alternatively, in some embodiments, the monitoring device 606 may be communicatively coupled to the vehicle 602 through a wireless connection. Examples of the monitoring device 606 may include, but are not limited to, a smartphone, a laptop, a mobile phone, a smart watch, smart-band, a smart wearable, and a vehicle dashboard. Further, the monitoring device 606 may include or may be associated with an ML model 606a. The ML model 606a may be updated time-to-time when the monitoring device 606 is employed within vehicle 602.
[050] The ML model 606 may be an encoder-decoder model and trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle. Examples of the encoder-decoder model may include, but are not limited to, a Recurrent Neural Network (RNN) encoder-decoder model, a Long short-term memory (LSTM) encoder-decoder model, and a Convolutional Neural Network (CNN) encoder-decoder model.
[051] The monitoring device 606 may generate a real-time condition vector based on the sensor data 604 using a ML model 606a. Further, the monitoring device 606 may determine a deviation in the real-time condition vector from a safe-condition vector based on the sensor data 604. In some embodiments, a confidence score for occurrence of fire-prone condition may be determined based on the deviation. Thereafter, the monitoring device 606 may determine if there is a fire-prone condition in the vehicle 602 based on the deviation and a predefined threshold deviation.
[052] In case the fire-prone condition is determined, the monitoring device 606 may generate an alert 608 and transmit the alert to the driver in a predefined time window. In the current example, the alert 608 may be a visual alert represented by a fire sign and rendered on a display device 610. Additionally, in some embodiments, alerts may be transmitted frequently until a fault within the vehicle is cleared and the deviation in the real-time condition vector from the safe-condition vector drops below the predefined threshold. Frequency of the alerts may be increased when the fault is not cleared in a predefined time.
[053] Referring now to FIG. 7, an exemplary ML model 700 trained for determining the fire-prone condition is illustrated, in accordance with some embodiments of the present disclosure. Fig. 7 is explained in conjunction with FIGs. 1-6. The ML model 700 is similar to the ML model 204a. The ML model 700 is an encoder-decoder architecture, which means the ML model 700 includes an encoder 702 and a decoder 704. Examples of the encoder-decoder model may include, but are not limited to, a Recurrent Neural Network (RNN) encoder-decoder model, a Long short-term memory (LSTM) encoder-decoder model, and a Convolutional Neural Network (CNN) encoder-decoder model. The encoder 702 may be trained using historical safe-condition data 706. For example, the historical safe-condition data 706 may include acceptable safe values corresponding to parameters associated with one or more fire-prone zones of the vehicle. These parameters may correspond to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle.
[054] By way of an example, the parameters may include an engine coolant temperature, an ambient temperature, a barometric pressure, a fuel injection pressure, an engine oil pressure, an engine torque, an engine load, a vehicle speed, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and a radiator fan operational time. Further the encoder 702 may generate a safe-condition vector 708 in a hidden embedding space 710. Once the encoder 702 is trained with the historical safe-condition data 706, the encoder may be capable of generating a real-time condition vector when real-time data is inputted to the encoder 702. In real-time data, values corresponding to one or more of the parameters may not be acceptable safe values or out of tolerance limit. In that case, the real-time condition vector may be generated based on these real-time values, and further to find the fire-prone condition, a deviation from the safe-condition vector may be determined.
[055] The decoder 704 may be trained to validate if the safe-condition vector 708 corresponds to the historical safe-condition data 706. Output of the decoder 704 may be decoded data 712 with same values. Further, in some embodiments, a confidence score may be determined based on the deviation by the ML model to determine chances of occurrence of possible fire.
[056] As will be also appreciated, the above-described techniques may take the form of computer or controller implemented processes and apparatuses for practicing those processes. The disclosure can also be embodied in the form of computer program code containing instructions embodied in tangible media, such as floppy diskettes, solid state drives, CD-ROMs, hard drives, or any other computer-readable storage medium, wherein, when the computer program code is loaded into and executed by a computer or controller, the computer becomes an apparatus for practicing the invention. The disclosure may also be embodied in the form of computer program code or signal, for example, whether stored in a storage medium, loaded into and/or executed by a computer or controller, or transmitted over some transmission medium, such as over electrical wiring or cabling, through fiber optics, or via electromagnetic radiation, wherein, when the computer program code is loaded into and executed by a computer, the computer becomes an apparatus for practicing the invention. When implemented on a general-purpose microprocessor, the computer program code segments configure the microprocessor to create specific logic circuits.
[057] Thus, the disclosed method and system try to overcome the existing technical problems. As mentioned earlier, the techniques disclosed herein include determining real-time condition vector, determining a deviation in the real-time condition vector with respect to safe-condition vector, and determining a fire-prone condition based on the deviation. The techniques help in timely detecting conditions which may eventually lead to the fire, thereby providing sufficient time to resolve the underlying problems. Thus, timely detection of the fire-prone condition may help in protecting the engine, battery, and other components from getting damaged. Further, the techniques may provide high reliability and accuracy in the detection of the fire-prone condition, as the ML model is used for detection of such condition. Additionally, the ML model employed is an encoder-decoder model that is trained with huge historical safe-condition data. As will be appreciated, the safe-condition data is much larger and easily available in comparison to the data that may result in a fire. Therefore, it is easier to build an ML model with the safe-condition data. In particular, it is easier to encode the safe-condition data in an encoder-decoder model and to determine a deviation in the real-time data with respect to the safe-condition data in order to detect a fire-prone condition. The techniques further provide generation of an alert in a predefined time window before actual occurrence of the fire-, thereby providing adequate time for the driver to take necessary actions for his safety.
[058] In light of the above-mentioned advantages and the technical advancements provided by the disclosed method and system, the claimed steps as discussed above are not routine, conventional, or well understood in the art, as the claimed steps enable the following solutions to the existing problems in conventional technologies. Further, the claimed steps clearly bring an improvement in the functioning of the device itself as the claimed steps provide a technical solution to a technical problem.
[059] The specification has described system and method for determining a fire-prone condition in a vehicle. The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed. Alternatives (including equivalents, extensions, variations, deviations, etc., of those described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope and spirit of the disclosed embodiments.
[060] Furthermore, one or more computer-readable storage media may be utilized in implementing embodiments consistent with the present disclosure. A computer-readable storage medium refers to any type of physical memory on which information or data readable by a processor may be stored. Thus, a computer-readable storage medium may store instructions for execution by one or more processors, including instructions for causing the processor(s) to perform steps or stages consistent with the embodiments described herein. The term “computer-readable medium” should be understood to include tangible items and exclude carrier waves and transient signals, i.e., be non-transitory. Examples include random access memory (RAM), read-only memory (ROM), volatile memory, nonvolatile memory, hard drives, CD ROMs, DVDs, flash drives, disks, and any other known physical storage media.
[061] It is intended that the disclosure and examples be considered as exemplary only, with a true scope and spirit of disclosed embodiments being indicated by the following claims.
, Claims:CLAIMS
We Claim:
1. A method (300) of determining a fire-prone condition in a vehicle, the method (300) comprising:
receiving (302), by a monitoring device (200), real-time data corresponding to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle, wherein the plurality of pre-defined parameters corresponds to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle;
generating (304), by the monitoring device (200), a real-time condition vector based on the real-time data using a machine learning (ML) model;
determining (306), by the monitoring device (200), a deviation in the real-time condition vector from a safe-condition vector, wherein the ML model is trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle; and
determining (308), by the monitoring device (200), the fire-prone condition in the vehicle based on the deviation and a predefined threshold deviation.

2. The method (300) as claimed in claim 1, comprising acquiring the real-time data corresponding to the plurality of predefined parameters through a plurality of sensors, wherein the plurality of sensors comprises temperature sensors, pressure sensors, speed sensors, torque sensors, and load sensors.

3. The method (300) as claimed in claim 1, wherein the plurality of predefined parameters comprises an engine coolant temperature, an ambient temperature, a barometric pressure, a fuel injection pressure, an engine oil pressure, an engine torque, an engine load, a vehicle speed, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and a radiator fan operational time.

4. The method (300) as claimed in claim 1, comprising generating (404), upon determining the fire-prone condition, an alert for a user associated with the vehicle, and wherein the alert is provided in a predefined time window based on the time of occurrence of the fire.

5. The method (300) as claimed in claim 1, wherein the ML model is an encoder-decoder model comprising:
an encoder trained with the historical safe-condition data to generate the safe-condition vector in a hidden embedding space and to generate the real-time condition vector based on the real-time data; and
a decoder trained to validate if the safe-condition vector corresponds to the historical safe-condition data.

6. A system (100) for determining a fire-prone condition in a vehicle, the system (100) comprising:
a monitoring device (102), wherein the monitoring device (102) comprises:
a processor (104); and
a computer-readable medium (106) communicatively coupled to the processor (104), wherein the computer-readable medium stores processor-executable instructions, which, on execution, cause the processor (104) to:
receive (302) real-time data corresponding to a plurality of predefined parameters associated with one or more fire-prone zones of the vehicle, wherein the plurality of pre-defined parameters corresponds to one or more parameters of at least one of a traction power source of the vehicle, a Heating Ventilation and Air Conditioning (HVAC) system of the vehicle, an auxiliary power source of the vehicle, a drivetrain of the vehicle, and an exhaust of the vehicle;
generate (304) a real-time condition vector based on the real-time data using a machine learning (ML) model;
determine (306) a deviation in the real-time condition vector from a safe-condition vector, wherein the ML model is trained to generate the safe-condition vector based on historical safe-condition data corresponding to the plurality of predefined parameters associated with the one or more fire-prone zones of the vehicle; and
determine (308) the fire-prone condition in the vehicle based on the deviation and a predefined threshold deviation.

7. The system (100) as claimed in claim 6, wherein the processor-executable instructions cause the processor (104) to acquire the real-time data corresponding to the plurality of predefined parameters through a plurality of sensors, wherein the plurality of sensors comprises temperature sensors, pressure sensors, speed sensors, torque sensors, and load sensors.

8. The system (100) as claimed in claim 6, wherein the plurality of predefined parameters comprises an engine coolant temperature, an ambient temperature, a barometric pressure, a fuel injection pressure, an engine oil pressure, an engine torque, an engine load, a vehicle speed, an exhaust after-treatment temperature, a temperature-rise during a regen cycle, and a radiator fan operational time.

9. The system (100) as claimed in claim 6, wherein the processor-executable instructions cause the processor (104) to generate (404), upon determining the fire-prone condition, an alert for a user associated with the vehicle, and wherein the alert is provided in a predefined time window based on the time of occurrence of the fire.

10. The system (100) as claimed in claim 6, wherein the ML model is an encoder-decoder model comprising:
an encoder trained with the historical safe-condition data to generate the safe-condition vector in a hidden embedding space and to generate the real-time condition vector based on the real-time data; and
a decoder trained to validate if the safe-condition vector corresponds to the historical safe-condition data.