Description: 2 SYSTEM AND METHOD FOR CLOGGED FILTER DETECTION AND NOTIFICATION IN A CONNECTED RESIDENTIAL AIR CONDITIONERS TECHNICAL FIELD 5 [001] The present disclosure relates to a system and an associated method for clogged filter detection and notification in a connected residential air conditioners. In particular, the present disclosure relates to the system and the method, where the system enables filter clogging 10 identified by Pulse Width Modulation (PWM) values generated from indoor Fan. PWM values can be impacted by various fan speeds like silent, low, medium, high, and turbo. BACKGROUND OF THE INVENTION 15 [002] Background description includes information that may be useful in understanding the present invention. [003] Air filters are considered to be one of the important components of air conditioners as they filter out dust and dirt along with things like pet dander and allergens. When the air filters are dirty, they may not filter the 20 dust from indoor air as expected. Hence, the people who stay indoors with filter clogged air conditioner for a longer time might face several health issues. [004] Apart from health issues, the clogged air filter causes the following problems: 25 • Overheating • Increased energy usage • Poor indoor air quality • Inadequate cooling 3 [005] Typically, the clogged air filters are cleaned during scheduled maintenance by the service engineers or by the customers themselves to address the above-mentioned problems. However, customers will come to know the filter clogging status only when they open the filter space or observe significant deviation in the air conditioner's performance. Hence, 5 it is important to identify the clogged air filter on-time to avoid unnecessary issues related to cooling, unnecessary energy consumption and potential health issues due to dirty filters. Prior art technologies merely allow a predefined filter clog notification sent after 480 running hours. [006] This approach may fail, if the filter clog happens before or after 480 10 hours. Hence, there is a need to identify the filter clog only when it occurs based on the systematic data analysis. [007] US9366448B2 discloses method and apparatus for configuring a filter change notification of an HVAC controller. The invention follows the threshold method to trigger a filter change notification. This system 15 identifies the need for air filter change by measuring pressure across the air filter in maintenance mode. Further, this invention also allows user to customize the filter change threshold. However, the drawbacks of the invention are disclosed below- • Threshold based approaches are not accurate. 20 • Additional overhead of initial calibration. • Users might not be able to provide correct threshold as they do not have much technical knowledge [008] Further, CN109028460B discloses an air conditioner filter screen filth blockage determination method and device and air conditioner. The 25 invention combines the PWM and fan rotating speed to detect the filter clog. The drawbacks of the invention are depicted below- • Depends on the calibrated threshold value. 4 • May not be able to capture the continuous degrade or intermediate filter clog. • Doesn’t cover the connected products. • Alert is not delivered in the mobile app. [009] Towards this direction, the present disclosure proposes a system 5 and an associated method to a data-driven and artificial intelligence driven solution to identify the clogged air filter and inform the customer on time. OBJECTS OF THE INVENTION [010] Some of the objects of the present disclosure, which at least one 10 embodiment herein satisfy, are listed herein below. [011] It is an object of the present subject matter to provide a system in order to enable a data-driven and Artificial Intelligence (AI) driven method to identify filter clogging without adding extra sensors. [012] It is another object of the present subject matter to provide a system 15 to identify the filter clogging by PWM values reported from indoor fan. [013] It is yet another object of the present subject matter to provide a system that is capable precise detection of filter clogging without need of dedicated and expensive sensors like airflow or air pressure sensors. [014] It is an yet another object of the present subject matter to provide a 20 system that is capable of timely notification to users and service teams to reduce service costs. [015] It is yet another object of the present subject matter to provide a system to enable to detect intermediate clogging in AC filters like 50%, 60% and 75%. 25 [016] These and other objects and advantages will become more apparent when reference is made to the following description and accompanying drawings. 5 SUMMARY OF THE INVENTION [017] This summary is provided to introduce concepts related to a method and a system that enables filter clogging identified by Pulse Width Modulation (PWM) values generated from indoor Fan wherein the drop in PWM values implies the clogged filter. This summary is not 5 intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. [018] In an aspect of the present disclosure, there is provided a method, and a system for detecting state of an air filter in air 10 conditioning or purification based electrical/electronic device based on artificial intelligence technique. The method comprising performing statistical detection through a remote server by the steps of capturing a PWM pattern of power supply for a blower covered by an air filter, capturing an RPM of the blower corresponding to the PWM pattern 15 and representing at least one of the PWM pattern and the corresponding RPM into a time series data. The method further includes detecting a change in statistical properties of the time series data based on a change point detection technique and performing a neural network enabled classification for categorizing the air filter as 20 clogged or unclogged, said classification based on said detection of statistical properties and a relationship between time series data of PWM and time series data of RPM. [019] In another aspect of the present disclosure, the method further includes sending a notification about classification through either an on-25 device display or a mobile app notification wherein the said notification comprising at least one of a clogging level and a clogging alert and receiving a user feedback in respect of the notification, wherein the said feedback certifying said clogging level and/or said clogging alert as 6 correct or incorrect. [020] In yet another aspect of the present disclosure, the change point detection technique detects change of statistical properties of the time series data. [021] In yet another aspect of the present disclosure, the method 5 further includes calculating a baseline data from the first detected change point in the time series data by the change point technique. [022] In yet another aspect of the present disclosure, the converting into the time series data further comprises storing the time series data as at least one of, high, low, medium values of RPM, and minimum, 10 maximum, median values, and standard deviation values of PWM. [023] In yet another aspect of the present disclosure, the detecting comprises comparing the represented time series data with historical time series data for detecting the change in statistical properties and thereby categorizing the air filter. 15 [024] In yet another aspect of the present disclosure, categorizing the air filter comprises: comparing the represented time series data with historical time series data; and detecting clogging by checking if the statistical properties of the represented time series data are less than the historical time series data. 20 [025] In yet another aspect of the present disclosure, the method further comprises determining a level of the clogging of the air filter during the neural network enabled classification performed over the represented time series data. [026] In yet another aspect of the present disclosure, the method 25 further comprises accumulating results of neural network classification over the time series data, said results comprising a labelled data duly verified based on user feedback; and directly executing the neural network enabled classification over the time series, upon said labelled 7 data exceeding a threshold. [027] 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 5 components. BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS [028] The illustrated embodiments of the subject matter will be understood by reference to the drawings, wherein like parts are 10 designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of devices, systems, and methods that are consistent with the subject matter as claimed herein, wherein: [029] Figure 1 illustrates an exemplary system facilitating detection 15 of filter clogging identified by Pulse Width Modulation (PWM) values in accordance with an exemplary embodiment of the present disclosure; [030] Figure 2 illustrates exemplary components of the system in accordance with an exemplary embodiment of the present disclosure; [031] Figure 3 illustrates an example method for working of the 20 system in accordance with an exemplary embodiment of the present disclosure; [032] Figure 4(i) illustrates a graph depicting daily PWM values for 392 days calculated during working of an AC device in accordance with an exemplary embodiment of the present disclosure; 25 [033] Figure 4(ii) illustrates the graph after applying the change point detection algorithm calculated during working of an AC device for 392 days in accordance with an exemplary embodiment of the present 8 disclosure; [034] Figure 5 illustrates the images depicting different clogging stages in the AC device in accordance with an exemplary embodiment of the present disclosure; [035] Figure 6 illustrate a working example scenario referring method of 5 Fig. 3 in accordance with an embodiment of the present subject matter; and [036] Figure 7 illustrates another example method for working of the system in accordance with an exemplary embodiment of the present disclosure. [037] The figures depict embodiments of the present subject matter for 10 the purposes of illustration only. A person skilled in the art will easily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein. 15 DETAILED DESCRIPTION [038] 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 20 contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosure as defined by the appended claims. [039] While the embodiments of the disclosure are subject to various modifications and alternative forms, specific embodiment thereof have been 25 shown by way of example in the figures and will be described below. It should be understood, however, that it is not intended to limit the disclosure to the particular forms disclosed, but on the contrary, the disclosure is to cover all 9 modifications, equivalents, and alternative falling within the scope of the disclosure. [040] The terms “comprises”, “comprising”, or any other variations thereof used in the disclosure, are intended to cover a non-exclusive inclusion, such that a device, system, assembly that comprises a list of components does not 5 include only those components but may include other components not expressly listed or inherent to such system, or assembly, or device. In other words, one or more elements in a system or device proceeded by “comprises… a” does not, without more constraints, preclude the existence of other elements or additional elements in the system or device. 10 [041] The present disclosure relates to a method, and a system for detecting state of an air filter in air conditioning or purification based electrical/electronic device based on artificial intelligence technique. The method comprising performing statistical detection through a remote server by the steps of capturing a PWM pattern of power supply for a blower covered by an air filter, 15 capturing an RPM of the blower corresponding to the PWM pattern and representing at least one of the PWM pattern and the corresponding RPM into a time series data. The method further includes detecting a change in statistical properties of the time series data based on a change point detection technique and performing a neural network enabled classification for categorizing the air 20 filter as clogged or unclogged, said classification based on said detection of statistical properties and a relationship between time series data of PWM and time series data of RPM. The corresponding structural and functional attributes of the system 101 are detailed in subsequent sections. Exemplary Implementations 25 [042] Figure 1 illustrates an exemplary system facilitating detection of filter clogging identified by Pulse Width Modulation (PWM) values in accordance with an exemplary embodiment of the present disclosure. The 10 filter clogging may be detected in respect of an air conditioner or AC by a remote system (101) which may be connected to the AC through cloud, as a part of IoT framework or by wired connection. Further, a user interface (109) or a user device (109) may be provided for operating over the AC and sending feedback to the system (101). 5 [043] In an aspect, the system (101) can be a single board computer built on a single circuit board, with microprocessor(s), memory, input/output (I/O) interface, and other features required for detection of filter clogging identified by Pulse Width Modulation (PWM) values. [044] In an aspect, the system (101) may comprise of different subunits, 10 for example, a reception engine (102), a conversion engine (103), an identification engine (104) and a selection engine (105). The functionalities of the different sub-units are discussed in subsequent sections with reference to Figure 2. [045] In an aspect, as can be evidenced from Figure 1, the selection engine 15 (105) may transmit data (i.e. for categorizing the air filter as clogged or unclogged) to a database (108) present in a central server (107) via a network (106). The central server (107) processes the information from the database (108) and may notify a user through a user interface (109). In an example, the communication network (106) includes, but not limited to, 2G network, 3G 20 network, 4G network, LTE network, 5G network, 6G network, and so forth. In other example the system (101), the database (108), the central server may be integrated network node or a single integrated unit. [046] The system (101) may be part of a larger computer system and/or maybe operatively coupled to the network (106) with the aid of a 25 communication interface to facilitate the transmission of and sharing data and predictive results. The computer network (106) may be a local area network (LAN), an intranet and/or extranet, an intranet and/or extranet that is in communication with the Internet, or the Internet. The network (106) in 11 some cases is a telecommunication and/or a data network, and may include one or more computer servers. The network (106), in some cases with the aid of a computer system, may implement a peer-to-peer network, which may enable devices coupled to the computer system to behave as a client or a server. 5 [047] The system (101) may also include memory or memory locations (e.g., random-access memory, read-only memory, flash memory), electronic storage units (e.g., hard disks) communication interfaces (e.g., network adapters) for communicating with one or more other systems, and peripheral devices, such as cache, other memory, data storage, and/or 10 electronic display adapters. [048] The system 101 also comprises one or more IO Managers as software instructions that may run on the one or more processors and implement various communication protocols such as User Datagram Protocol (UDP), Modbus, MQ Telemetry Transport (MQTT), Open 15 Platform Communications Unified Architecture (OPC UA), Semiconductor's equipment interface protocol for equipment-to-host data communications (SECS/GEM), Profinet, or any other protocol, to access data in real-time from disparate data sources via any communication network, such as Ethernet, Wi-Fi, Universal Serial Bus (USB), Zigbee, 20 Cellular or 5G connectivity, etc., or indirectly through a device’s primary controller, through a Programmable Logic Controller (PLC) or through a Data Acquisition System (DAQ), or any other such mechanism. [049] Figure 2 illustrates various exemplary components present in a system (101) in accordance with an exemplary embodiment of the present 25 disclosure. The system (101) includes, but not limited to, processors, memory elements, one or more sensors and the processing devices. [050] In an aspect, the processor(s) (201) may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal 12 processors, central processing units, logic circuitries, and/or any devices that manipulate data based on operational instructions. Among other capabilities, the one or more processor(s) (201) are configured to fetch and execute computer-readable instructions stored in the memory of the system (101). 5 [051] In an aspect, the memory (202) may store one or more computer-readable instructions or routines, which may be fetched and executed to create or share data units over a network service. The memory may include any non-transitory storage device including, for example, volatile memory such as RAM, or non-volatile memory such as EPROM, flash memory, and 10 the like. [052] In an aspect, the processing devices(s) may be implemented as a combination of hardware and programming device(s) (for example, programmable instructions) to implement one or more functionalities of the processing device(s). In examples described herein, such combinations of 15 hardware and programming may be implemented in several different ways. In one example, the programming for the processing device(s) may be processor executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processing device(s) may include a processing resource (for example, one or more processors), to execute such 20 instructions. In other examples, the processing devices(s) may be implemented by electronic circuitry. [053] In an aspect, the interface(s) (203) facilitate communication of the system (101) with various devices coupled to the system (101). The interface(s) also provide a communication pathway for one or more 25 components of the system (101). Examples of such components include, but are not limited to, processing device(s) and data storage. [054] In an aspect, the processing device (206) comprises of a reception engine (102), where the reception engine (102) fetches static 13 as well as dynamic inputs. The static inputs include capturing a PWM pattern of power supply for a blower covered by an air filter and capturing an RPM of the blower corresponding to the PWM pattern at a given time period. [055] In an aspect, the dynamic inputs, on the other hand include, 5 continuously over a period of time, capturing a PWM pattern of power supply for a blower covered by an air filter and capturing an RPM of the blower corresponding to the PWM pattern. [056] In an aspect, the conversion engine (103) receives input from the reception engine (102) and represents at least one of the PWM pattern 10 and the corresponding RPM into a time series data. The converting into the time series data further comprises storing the time series data as at least one of: high, low, medium values of RPM; and minimum, maximum, median values, and standard deviation values of PWM. [057] In an aspect, the identification engine (104) upon receiving 15 inputs from the conversion engine (103) and the pre-defined data sets (static and dynamic) from the reception engine (102), identifies or detects a change in statistical properties of the time series data based on a change point detection technique. [058] In as aspect, the change point detection technique by the 20 identification engine (104) detects change of statistical properties of the time series data. The change point detection technique further comprises of calculating a baseline data from the first detected change point in the time series data by the change point technique. [059] In an aspect, the detecting of change in statistical properties by 25 the identification engine (104) comprises comparing the represented time series data with historical time series data for detecting the change in statistical properties and thereby categorizing the air filter. [060] In an aspect, output of the identification engine (104) is 14 communicated to a selection engine (105) to perform a neural network enabled classification for categorizing the air filter as clogged or unclogged. [061] In an aspect, the said classification by the selection engine (105) is based on detection of statistical properties and a relationship between 5 time series data of PWM and time series data of RPM. [062] In an aspect, the detection of statistical properties by the selection engine (105) comprises of comparing the represented time series data with historical time series data; and detecting clogging by checking if the statistical properties of the represented time series data 10 are less than the historical time series data. [063] In an aspect, the detection by the selection engine (105) further comprises determining a level of the clogging of the air filter during the neural network enabled classification performed over the represented time series data. 15 [064] In an aspect, the detection by the selection engine (105) involves accumulation of results of neural network classification over the time series data, said results comprising a labelled data duly verified based on user feedback. In such a scenario, the engine (105) may directly execute the neural network enabled classification over the time series 20 data collected by the reception engine (102), when said labelled data exceeds a threshold. [065] The system (101) also comprises of a storage device (208) to store the data as received from different components of the processing device (206) in order to enable access of the information as and when 25 required. [066] In an aspect, the system (101) through a communication module is configured to send a notification about classification through either an on-device display within the air conditioner. Otherwise, the 15 notification may be communicated as a mobile app notification to a user device (109). The notification comprises at least one of a clogging level and a clogging alert and to receive a user feedback in respect of the notification, said feedback certifying said clogging level and/or said clogging alert as correct or incorrect. 5 [067] Figure 3 illustrates an example method implemented in the system (101) in accordance with an exemplary embodiment in the present disclosure. The order in which the method is described is not intended to be construed as a limitation, and any number of the described method blocks may be combined in any order to implement 10 the method, or an alternative method. Furthermore, the method may be implemented by processing device(s) or computing device(s) through any suitable hardware. [068] At block (301), the method includes capturing a PWM pattern of power supply for a blower covered by an air filter. 15 [069] At block (302), the method includes capturing an RPM of the blower corresponding to the PWM pattern. The step 301 and 302 refer the operation of a reception engine 102 as discussed in preceding Fig. 2. [070] At block (303), the method includes representing at least one of 20 the PWM pattern and the corresponding RPM into a time series data. The step 303 refer the operation of an conversion engine 103 as discussed in preceding Fig. 2. [071] At block (304), which refers an operation of the identification engine (104), the method includes detecting a change in statistical 25 properties of the time series data based on a change point detection technique. The change point detection is one of the key components in the time series analysis as it identifies the changes in the underlying data. In particular, the change point detection detects the changes in 16 statistical properties in the time series data. In an aspect, the input data has been modelled as a time series data and the change point detection is applied to identify the changes in PWM's statistical properties for each device. [072] The baseline is generated after identifying the change point from 5 the dataset for the first time. A first encounter of change signifies the possibility of filter clogging. Hence, in an aspect, the statistical property will be calculated of data just before the first change point and store it as a baseline. [073] The baselines of the AC will be stored in Azure Blob as the .json 10 file. In an aspect, the following schema format can be used to log the baseline values: "fan_speed" : { "minimum" : Numeric, "q1" : Numeric, 15 "median" : Numeric, "q3" : Numeric, "maximum" : Numeric } [074] Further, in an aspect, the following keys can be used to store the 20 baseline: • Fan_speed: It is just a place holder text. This will be replaced by any of 'medium', 'high', and 'turbo'. [075] Each key ('medium', 'high', 'turbo') will have another 'dict' as value as follows: 25 • "minimum": minimum_PWM (for a given period) • "q1": lower quartile • "median": median PWM value 17 • "q3": upper quartile • "maximum": maximum_PWM(for a given period) [076] Furthermore, in an aspect, for the AC device, the baseline values will be stored as per the following samples: medium.json "medium" : { "minimum" : 20, "q1" : 30, "median" : 40, "q3" : 50, "maximum" : 60 } high.json "high" : { "minimum" : 25, "q1" : 35, "median" : 45, "q3" : 55, "maximum" : 65 } turbo.json "turbo" : { "minimum" : 30, "q1" : 40, "median" : 50, "q3" : 60, "maximum" : 70 } 5 [077] At block (305), which refers an operation of the selection engine (105), the method includes performing a neural network enabled classification for categorizing the air filter as clogged or unclogged, said classification based on said detection of statistical properties and a relationship between time series data of PWM and time series data of 10 RPM. [078] The overall code flow and infrastructure requirements would be depicted in the subsequent paragraphs. In an aspect, the following Pseudocode explains the overall flow- FETCH List of devices with PWM FROM Lookup Table FOR each device: IF `baseline_stats` not exists in BLOB: FETCH available data from BLOB IF duration(data_available) >= 60 days: PERFORM `change_point_detection` on the available data IF 'change_point' detected FETCH data till first `change_point` CALCULATE stats STORE stats as `baseline_stats` in BLOB 18 ELSE CONTINUE to next device ELSE: CONTINUE to next device ELSE: LOAD `baseline_stats` from BLOB FETCH `daily_telemetry_data` from BLOB CALCULATE median from `daily_telemetry_data` IF `daily median` is out of range of `baseline_stats`: SEND notification ELSE: CONTINUE to next device [079] In an aspect, as described in the Pseudocode above paragraph, the PWM reported devices can be fetched from Lookup table, and telemetry data will be fetched only for those devices. Further, the data can be preprocessed and decisions about fault will be taken as per the 5 Pseudocode. [080] In this manner, the system (101) operates and provide the method for to enable detection of filter clogging identified by Pulse Width Modulation (PWM) values generated from indoor Fan. 10 Working of the invention [081] The working of the present system (101) has been elucidated considering a few examples as mentioned below: [082] Example 1 [083] Consider an AC device that has Daily PWM data for 392 days. 15 The AC fan was operated in Turbo mode. [084] In an aspect, a sample of data collected by reception engine (102) is given by the table below: 19 [085] Figure 4(i) illustrates a graph depicting daily PWM values for 392 days calculated by the conversion engine (103) during working of an AC device in accordance with an exemplary embodiment of the present disclosure. [086] It was observed that the AC device operated with PWM values 5 of around 1700 since the day 0 for around 130 days. After 130 days, there was a drop in the PWM value and the PWM value fell to around 1530. The said fall indicated partial filter clogging. The AC was allowed to operate normally after day 130 and the same PWM value, i.e. around 1530 was observed until day 255. Thus, indicating the same partial 10 clogging. [087] After the day 255, there was again a drop in the PWM values and it fell below the 1400 mark. The falling of the PWM values below the 1400 mark depicted the AC in a fully clogged condition. Thus, the analysis of figure 4(i) is analyzed below- 15 Normal operation till Day 130 Partial filter clogged from Day 130 to 255 Fully Clogged Operation from Day 255 onwards [088] Figure 4(ii) illustrates the graph after applying the change point 20 detection algorithm by the identification engine (104). Th engine (104) executes based on working of an AC device for 392 days in accordance with an exemplary embodiment of the present disclosure. As can be observed from the 20 figure 4(ii), the detected change points are 130 and 255. After detecting the said change points, the change point detection algorithm comprises of the below mentioned steps- • The data for 392 days is stored in a 1-d matrix, D. • Consider the data is indexed as D0 to D391 5 • A matrix, M of size 392 x 392 is generated. • To populate the value of Mi, j • The Euclidean distance between the data points Di and Dj is calculated. • Radial basis function (RBF) is applied on the distance value to discriminate between closer and farther points. 10 • Closer points have a low value and farther points have higher value. • The above matrix becomes the partition cost matrix. Through an iterative search method, the partition scheme with the least cost is found. • If M0, j + Mj, k + Mk, 392 is the lowest among all possible combinations in the matrix, then j, k are identified as change points. 15 • In the above example, j is identified as 130 and k is identified as 295. [089] Further, it is pertinent to note that the number of change points can vary based on the input data. In the above example only two change points are found, hence, M0, j + Mj, k + Mk, 392 is less. [090] If multiple change points are identified,