Description:FORM 2

THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003

COMPLETE SPECIFICATION
(See Section 10 and Rule 13)

Title of invention:
DETECTING DEFECTS IN A SOLAR PLANT

Applicant:
Fototentia Diagnostics Private Limited
B-702, Satellite Garden phase 1,
Gen AKV Marg, Goregaon East,
Mumbai-400063, Maharashtra

The following specification describes the invention and the manner in which it is to be performed.

PRIORITY INFORMATION
[001] The present application does not claim a priority from any other application.
TECHNICAL FIELD
[002] The present subject matter described herein, in general, relates to a system and a method for detecting defects in a solar plant.
BACKGROUND
[003] A solar plant comprises numerous components. The components are of different types and each component has a specific function. For the solar plant to perform at high efficiency, all the components of the solar plant need to be in proper working condition. Any damage to one component could lead to a problem in the functioning of the solar plant. Each type of component has different working conditions and is prone to different type of damage. Manually inspecting all the different types of components in a solar plant is a difficult, man-power intensive and a time-consuming job.
SUMMARY
[004] Before the present system(s) and method(s), are described, it is to be understood that this application is not limited to the particular system(s), and methodologies described, as there can be multiple possible embodiments which are not expressly illustrated in the present disclosures. It is also to be understood that the terminology used in the description is for the purpose of describing the particular implementations or versions or embodiments only and is not intended to limit the scope of the present application. This summary is provided to introduce aspects related to a system and a method for detecting defects in a solar plant. This summary is not intended to identify essential features of the claimed subject matter nor is it intended for use in determining or limiting the scope of the claimed subject matter.
[005] In one implementation, a method for detecting defects in a solar plant is disclosed. Initially, a set of images comprising thermal images and visual images of a plurality components of a solar plant may be received from a set of sensors on a platform. Each image from the set of images corresponds to an image of one component. Further, the set of images may be processed using a machine learning algorithm to recognise a type of each component of the plurality of components. The type of each component may be at least one of a photovoltaic module, a string combiner box, an inverter, and a transformer. Further, one or more sub-components of each component are identified using a machine learning model. Subsequently, a temperature of the one or more sub-components may be determined based on the thermal image of the component. The temperature may be determined by processing the thermal image of the component based on a pixel value of one or more pixels of the thermal images. The pixel value may correspond to a colour of a pixel. Further, one or more defects in each component may be detected based on the temperature of the one or more sub-components using a machine learning model. Each sub-component corresponding to the one or more defects may be highlighted. The sub-component may be highlighted by labelling the sub-component with the defect. The sub-component may be labelled with the defect by creating a bounding box around the sub-component. Further, a location of each component from a list of defective components may be identified based on GPS information associated with the set of images. The list of defective components may comprise each component with one or more defects detected. Furthermore, a strategy to fix the one or more defects may be recommended using reinforcement learning and artificial intelligence techniques. Finally, a health report of the solar plant may be generated based on the one or more defects in the plurality of components. In one aspect, the aforementioned method for detecting defects in a solar plant may be performed by a processor using programmed instructions stored in a memory.
[006] In another implementation, a non-transitory computer readable medium embodying a program executable in a computing device for detecting defects in a solar plant is disclosed. The program may comprise a program code for receiving, from a set of sensors, a set of images comprising thermal images and visual images of a plurality of components of a solar plant. Each image from the set of images may correspond to an image of one component. Further, the program may comprise a program code for processing the set of images using a machine learning model for recognising a type of the plurality of components in the set of images. The type may be at least one of a photovoltaic module, a string combiner box, an inverter, and a transformer. Furthermore, the program may comprise a program code for identifying one or more sub-components of each component using a machine learning model. Subsequently, the program may comprise a program code for determining a temperature of the one or more sub-components based on a thermal image of each component. The temperature may be determined by processing the thermal image of the component based on a pixel value of one or more pixels of the thermal images. The pixel value may correspond to a colour of a pixel. Further, the program may comprise a program code for detecting one or more defects in each component based on the temperature of the one or more sub-components using a machine learning model. Detecting one or more defects in each component may comprise highlighting each sub-component corresponding to the one or more defects. Highlighting the sub-component may comprise labelling the sub-component with the defect. Labelling the sub-component with the defect may comprise creating a bounding box around the sub-component. Further, the program may comprise a program code for identifying a location of each component from a list of defective components based on GPS information associated with the set of images. The list of defective components may comprise each component with one or more defects detected. Furthermore, the program may comprise a program code for recommending a strategy to fix the one or more defects using reinforcement learning and artificial intelligence techniques. Finally, the program may comprise a program code for generating a health report of the solar plant based on the one or more defects in the plurality of components. Generating the health report may comprise a list of defective components, the one or more defects corresponding to each component in the list of defective components, the location of the defective components, and the strategy to fix the one or more defects of each component.
BRIEF DESCRIPTION OF THE DRAWINGS
[007] The foregoing detailed description of embodiments is better understood when read in conjunction with the appended drawings. For the purpose of illustrating of the present subject matter, an example of a construction of the present subject matter is provided as figures, however, the invention is not limited to the specific method and system for detecting defects in a solar plant disclosed in the document and the figures.
[008] The present subject matter is described in detail with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the drawings to refer to various features of the present subject matter.
[009] Figure 1 illustrates a network implementation detecting defects in a solar plant, in accordance with an embodiment of the present subject matter.
[010] Figure 2 illustrates a method detecting defects in a solar plant, in accordance with an embodiment of the present subject matter.
[011] Figure 3 illustrates an example of a thermal image and a visual image used by the system.
[012] Figure 4 illustrates an image of a component highlighted with a defect.
[013] The figures depict an embodiment of the present disclosure for purposes of illustration only. One skilled in the art will readily recognize from the following discussion that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.

DETAILED DESCRIPTION
[014] Some embodiments of this disclosure, illustrating all its features, will now be discussed in detail. The words "receiving," "recognising," "identifying," "determining," "detecting," "generating," and other forms thereof, are intended to be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the listed item or items. It must also be noted that as used herein and in the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Although any system and methods similar or equivalent to those described herein can be used in the practice or testing of embodiments of the present disclosure, the exemplary, system and methods are now described.
[015] The disclosed embodiments are merely examples of the disclosure, which may be embodied in various forms. Various modifications to the embodiment will be readily apparent to those skilled in the art and the generic principles herein may be applied to other embodiments. However, one of ordinary skill in the art will readily recognize that the present disclosure is not intended to be limited to the embodiments described but is to be accorded the widest scope consistent with the principles and features described herein.
[016] The present subject matter discloses a method and a system for detecting defects in a solar plant. The defects correspond to an anomaly in various components of the solar plant namely, a photovoltaic module, a string combiner box, a transformer, and an inverter. The goal of the invention is to identify any anomaly in the solar plant and provide a health report of the solar plant such that the anomalies may be fixed. The invention also discloses a method to recommend strategies to fix the anomalies. The invention aims at automating and increasing efficiency of evaluating health of the solar plant. The system receives a set of images corresponding to one or more components in the solar plant. The system identifies a type of the one or more components by processing the set of images based on a machine learning algorithm. Further, the system detects defects based on the type of component and the temperature of the one or more sub-components. Finally, the system provides a comprehensive health report of the solar plant. The health report comprises a list of defective components, corresponding defects, location of the components in the solar plant, and strategies to fix the one or more defects.
[017] In an embodiment, the system detects defects in a solar plant. It may be noted that overall health of the solar plant may be evaluated using the system. In another embodiment, the system may recommend strategies to fix the defects detected in the solar plant. The current invention helps to identify defects in a solar plant, recommend strategies to fix the defects, and generate a health report of the solar plant.
[018] Certain technical challenges exist for achieving the goal of detecting defects in a solar plant. One technical challenge includes identifying defects in different types of components of a solar plant. The solution presented by the embodiments disclosed herein to address the above challenge is a machine learning model trained to recognise a type of component in an image by image-processing. Further, the machine learning model is trained to detect different defects in different types of components. It may be noted that use of one or more machine learning models is required to detect defects in different types of components. The machine learning models may process images to recognise a type of a component in an image. The machine learning models may identify sub-components of a component in the images based on the type of component. The machine learning models may detect a defect in the component based on a temperature of the sub-components by processing thermal images using thermographic analysis. Another technical challenge includes, recommending a strategy to fix various defects that may be detected in different types of components.
[019] The solution presented by the embodiments disclosed herein to address the above challenge is through analysing defects based on the type of component and using reinforcement learning and artificial intelligence techniques to calculate the strategy to fix the defects. The system may use historic data of maintenance, repairs, and defects of the components to train a machine learning model using reinforcement learning for calculating a strategy to fix the defects.
[020] Referring now to Figure 1, a network implementation 100 of a system 102 for detecting defects in a solar plant is disclosed. Initially, the system 102 receives a set of images of a plurality of components. In an example, the software may be installed on a user device 104-1. It may be noted that the one or more users may access the system 102 through one or more user devices 104-2, 104-3…104-N, collectively referred to as user devices 104, hereinafter, or applications residing on the user devices 104. The system 102 receives information from one or more sensor devices 104.
[021] Although the present disclosure is explained considering that the system 102 is implemented on a server, it may be understood that the system 102 may be implemented in a variety of computing systems, such as a laptop computer, a desktop computer, a notebook, a workstation, a virtual environment, a mainframe computer, a server, a network server, a cloud-based computing environment. It will be understood that the system 102 may be accessed by multiple users through one or more user devices 104-1, 104-2…104-N. In one implementation, the system 102 may comprise the cloud-based computing environment in which the user may operate individual computing systems configured to execute remotely located applications. Examples of the user devices 104 may include, but are not limited to, a portable computer, a personal digital assistant, a handheld device, and a workstation. The user devices 104 are communicatively coupled to the system 102 through a network 106.
[022] In one implementation, the network 106 may be a wireless network, a wired network, or a combination thereof. The network 106 can be implemented as one of the different types of networks, such as intranet, local area network (LAN), wide area network (WAN), the internet, and the like. The network 106 may either be a dedicated network or a shared network. The shared network represents an association of the different types of networks that use a variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), Wireless Application Protocol (WAP), and the like, to communicate with one another. Further the network 106 may include a variety of network devices, including routers, bridges, servers, computing devices, storage devices, and the like.
[023] In one embodiment, the system 102 may include at least one processor 108, an input/output (I/O) interface 110, and a memory 112. The at least one processor 108 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, Central Processing Units (CPUs), state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions. Among other capabilities, the at least one processor 108 is configured to fetch and execute computer-readable instructions stored in the memory 112.
[024] The I/O interface 110 may include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like. The I/O interface 110 may allow the system 102 to interact with the user directly or through the client devices 104. Further, the I/O interface 110 may enable the system 102 to communicate with other computing devices, such as web servers and external data servers (not shown). The I/O interface 110 can facilitate multiple communications within a wide variety of networks and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular, or satellite. The I/O interface 110 may include one or more ports for connecting a number of devices to one another or to another server.
[025] The memory 112 may include any computer-readable medium or computer program product known in the art including, for example, volatile memory, such as static random-access memory (SRAM) and dynamic random-access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, Solid State Disks (SSD), optical disks, and magnetic tapes. The memory 112 may include routines, programs, objects, components, data structures, etc., which perform particular tasks or implement particular abstract data types.
[026] The memory 112 may include programs or coded instructions that supplement applications and functions of the system 102. In one embodiment, the memory 112, amongst other things, serves as a repository for storing data processed, received, and generated by one or more of the programs or the coded instructions.
[027] As there are various challenges observed in the existing art, the challenges necessitate the need to build the system 102 for detecting defects in a solar plant. At first, a user may use the user device 104 to access the system 102 via the I/O interface 110. The user may register the user devices 104 using the I/O interface 110 in order to use the system 102. In one aspect, the user may access the I/O interface 110 of the system 102. The detail functioning of the system 102 is described below with the help of figures.
[028] The present subject matter describes the system 102 for detecting defects in a solar plant. In an embodiment, the system 102 receives a set of images comprising thermal images and visual images of a plurality of components of a solar plant. Each image from the set of images may be an image of one component. The set of images may be received from sensor devices such as a vision camera, and a thermal camera by a drone, a human, a remote-controlled robot, and the like. In an embodiment, the system may continuously receive the set of images at certain intervals of time. The solar plant may comprise different types of components namely, (a) a photovoltaic module, (b) a string combiner box, (c) an inverter, and (d) a transformer.
[029] In another embodiment, the system may receive a set of thermal images in a zip format. The system extracts visual images from the set of thermal images using deep learning, machine learning algorithms, and image processing techniques.
[030] The system may use a machine learning model to recognise a type of component in each image of the set of images. The machine learning model may be trained using a dataset of labelled set of images. The labelled set of images may comprise one or more images of each type of component labelled with the type. The machine learning model may be trained to recognise the type of the component in an unlabelled visual image. Each type of component may have one or more sub-components. The one or more sub-components of the component may be identified using a machine learning model trained using a dataset comprising one or more images of different types of components labelled with one or more sub-components of each of the different types of components. The one or more sub-components of the different types of components may be as shown in Table A.
Type of Component Sub-components
Photovoltaic Module 1) Solar Cell
2) Frame
3) Protective Glass
4) Encapsulant
5) Back Sheet
String Combiner Box 1) MC4 Connector
2) SCB main o/g cable
3) String Fuse
4) Fuse Holders
5) DC switch
6) String Cable
Inverters 1) DC Cable
2) AC Cable
3) DC Fuse
4) AC Fuse
Transformers 1) HT Cable
2) LT Cable
Table A
[031] The system may comprise a thermographic engine for determining a temperature of the one or more sub-components, also called as thermal data hereafter. The thermographic engine may determine the temperature by processing the corresponding thermal image using machine learning models for thermographic analysis. Thermographic analysis may be used to extract temperature readings from thermal images using a pixel value of one or more pixels of the thermal image. The pixel value may correspond to a colour. The machine learning model may be trained to analyse the pixel value of one or more pixels and determine the temperature based on the pixel value. Each pixel value corresponds to a different colour. The machine learning model may be trained using a matrix of pixel values and corresponding temperatures.
[032] In an embodiment, the system may comprise an image processing engine for converting the visual images and the thermal data into text format to generate structured, tabular data. Further, the image processing engine may process the structured, tabular data to recognise duplicate images, and waste images. The waste images may be images comprising only a part of one component. The image processing engine may update the structured, tabular data by removing the data corresponding to duplicate images and waste images.
[033] The temperature of the one or more sub-components may be compared with an ideal temperature of the one or more sub-components. Further, the system may comprise an error engine for detecting one or more defects in the component based on the comparison and the type of the component. The defects that may be detected in the different types of components are as shown in Table B.

Type of Component Defects
Photovoltaic Module 1) Hotspot
2) Crack
3) Misaligned Frame
String Combiner Box 1) MC4 (MC4 Connector)
2) SOC (SCB main o/g cable)
3) SFH (String Fuse Holder)
4) DDS (DC Disconnect Switch)
5) SCT (String Cable Termination)
Inverters 1) DCT (DC Cable Termination)
2) ACT (AC Cable Termination)
3) DCF (DC Fuse)
4) ACF (AC Fuse)
Transformers 1) HTCT (Transformer HT Cable termination)
2) LTCT (Transformer LT Cable termination)
Table B
[034] In an embodiment, the one or more defects may be classified into four categories namely, low, moderate, high, and extreme based on a temperature difference (dT) between the temperature and the ideal temperature of the sub-component associated with each defect. The defects may be classified as shown below:

Category Temperature Difference
LOW dT<=10
MODERATE 10 < dT <= 20
HIGH 20 < dT <= 40
EXTREME dT > 40

[035] In an embodiment, the system may comprise an annotation engine for highlighting the one or more defects in the images of components. The one or more sub-components having a temperature greater than the ideal temperature of the one or more sub-components may be highlighted in the image of a defective component. The sub-component may be labelled with the detected defect using image overlay. In an embodiment, the one or more sub-components may be labelled by creating a bounding box, in the corresponding images, around the one or more sub-components having the temperature higher than the ideal temperature. In an embodiment, one or more pixels surrounding the sub-components to be highlighted may be modified by at least one of colour inversion, saturation, increasing contrast, and changing colour.
[036] In an embodiment, the system may comprise a repair and maintenance engine for facilitating easy repair of the defective components of the solar plant. The repair and maintenance engine may extract GPS information associated with each image from the set of images to determine location of the component, present in the image, in the solar plant. In an embodiment, only the defective components may be located.
[037] Further to detecting the one or more defects in the component, the system may use the repair and maintenance engine for recommending a strategy to fix the one or more defects. The repair and maintenance engine may calculate the strategy using reinforcement learning and artificial intelligence. The repair and maintenance engine may use a machine learning model trained using reinforcement learning. The machine learning model may be trained using a data set comprising historic data about repair and maintenance of the one or more components in the solar plant, and one or more defects that generally occur in the one or more components. The machine learning model may be used to recommend the strategy to fix the one or more defects. The machine learning model may compare a particular defect, with the defects in the historic data to find the repair or maintenance undertaken for that particular defect. The historic data may be shown as Table C.
Date Defect Repair/Maintenance
02/04/2004 LTCT Tightened Transformer cable
02/06/2014 DCF Replaced DC Fuse
Table C
[038] The system may generate a health report of the solar panel based on the one or more defects detected in each component of the plurality of components. The solar plant may have one or more inverters, transformers, photovoltaic modules, string combiner boxes. In an embodiment the one or more components may be numbered or labelled with alphabets. The health report may comprise a list of defective components, the one or more defects corresponding to each component in the list of defective components, the categories of the one or more defects, the location of the defective components, and the strategy to fix the one or more defects of each component. The health report may be as shown in Table D.

Component Defect Category Location Strategy
Transformer D HTCT LOW 41.40338, 2.17403 Replace HT Cable
Inverter B DCT EXTREME 21.88652, 4.35623 Tighten DC cable
Inverter F DCF HIGH 26.85552, 3.33523 Replace DC fuse
Table D
[039] The system may calculate a health rating for the solar plant based on the health report. The system may compare the total number of components with the number of defective components to calculate the health rating. In an embodiment, the system may calculate a ratio to determine the health rating. The ration may be calculated as – Ratio = (number of defects classified as high + number of defects classified as extreme) / (total number of defects). The health rating of the solar plant may be determined as shown below:
Health Rating Ratio
Good Ratio <= 0.1
Average 0.1 < Ratio <= 0.2
Low Ratio > 0.2

[040] Let us assume the health report of the solar panel is as shown in Table In the present invention, the machine learning models may work in real time during continuous monitoring of a solar plant. Further, the machine learning models may be continuously learning using reinforcement learning. Lastly, the system may identify defects based on temperatures of sub-components of different type of components. A traditional monitoring system can only identify discrepancies or anomalies in temperature which may be flagged as defects.
[041] It may be understood that the present invention uses fully automated system 102 with no human involvement in operations. Therefore, efficiency and accuracy of the machine learning model may be bound to increase manifold as compared to the undeniable limitations of the traditional monitoring methods. In view of the above, the present invention may be understood to be an advancement over the human intelligence and thus the steps may not be performed by the individuals in the traditional monitoring methods.
[042] Referring now to figure 2, a method 200 for detecting defects in a solar plant is shown, in accordance with an embodiment of the present subject matter. The method 200 may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, functions, etc., that perform particular functions or implement particular abstract data types.
[043] The order in which the method 200 is described is not intended to be construed as a limitation, and any number of the described method blocks can be combined in any order to implement the method 200 or alternate methods for detecting defects in a solar plant. Additionally, individual blocks may be deleted from the method 200 without departing from the scope of the subject matter described herein. Furthermore, the method 200 for detecting defects in a solar plant can be implemented in any suitable hardware, software, firmware, or combination thereof. However, for ease of explanation, in the embodiments described below, the method 200 may be considered to be implemented in the above-described system 102.
[044] At block 202, a set of images comprising thermal images and visual images of a plurality of components of a solar plant may be received.
[045] At block 204, a type of the plurality of components in the set of images may be recognised. The type may be at least one of a photovoltaic module, a string combiner box, an inverter, and a transformer.
[046] At block 206, one or more sub-components of each component may be identified using a machine learning model.
[047] At block 208, a temperature of the one or more sub-components may be determined based on a thermal image of each component. The temperature may be determined using a machine learning model and a pixel value of one or more pixels of the thermal image.
[048] At block 210, one or more defects in each component may be detected based on the type of component, and the temperature of the one or more sub-components using a machine learning model.
[049] At block 212, a health report of the solar plant may be generated based on the one or more defects in the plurality of components. The health report may comprise a list of defective components, the one or more defects corresponding to each component in a list of defective components, location of the defective components, and a strategy to fix the one or more defects of each component.
[050] Figure 3 illustrates an example of a visual image 306 and a thermal image 302 received by the system. The block 304 illustrates a colour gradient between a minimum temperature in the image and a maximum temperature in the image. In an embodiment, the block 304 represents a matrix of temperatures and corresponding pixel values.
[051] Figure 4 illustrates an image of a defective component. The defect is marked with a bounding box 402.
[052] Exemplary embodiments discussed above may provide certain advantages. Though not required to practice aspects of the disclosure, these advantages may include those provided by the following features.
[053] Some embodiments of the system and the method save time in monitoring large solar plants.
[054] Some embodiments of the system and the method provide accurate data regarding health of a solar plant.
[055] Some embodiments of the system and the method helps in reducing manpower required to monitor a solar plant.
[056] Some embodiments of the system and the method save time required to assess and fix defects in a solar plant.
[057] Although implementations for methods and system for detecting defects in a solar plant have been described in language specific to structural features and/or methods, it is to be understood that the appended claims are not necessarily limited to the specific features or methods described. Rather, the specific features and methods are disclosed as examples of implementations for detecting defects in a solar plant. , C , Claims:WE CLAIM:
1. A method, implemented by a system for detecting defects in a solar plant, comprising:
receiving a set of images comprising thermal images and visual images of a plurality of components of a solar plant;
recognising a type of the plurality of components in the set of images, wherein the type is at least one of a photovoltaic module, a string combiner box, an inverter, and a transformer;
identifying one or more sub-components of each component using a machine learning model;
determining a temperature of the one or more sub-components based on a thermal image of each component;
detecting one or more defects in each component based on the type of component, and the temperature of the one or more sub-components using a machine learning model; and
generating a health report of the solar plant based on the one or more defects in the plurality of components.
2. The method in claim 1, wherein each image from the set of images is an image of one component.
3. The method in claim 1, wherein recognizing the type of component comprises processing visual images using a machine learning algorithm.
4. The method in claim 1, wherein detecting one or more defects in each component comprises comparing the temperature of each sub-component with an ideal temperature of each sub-component.
5. The method in claim 1, wherein determining the temperature comprises processing the thermal images using thermographic analysis. The temperature is determined based on a pixel value of one or more pixels of the thermal images, wherein the pixel value corresponds to a colour of a pixel.
6. The method in claim 1, wherein detecting one or more defects in each component comprises highlighting each sub-component corresponding to the one or more defects, and wherein highlighting the sub-component comprises labelling the sub-component with the defect.
7. The method in claim 6, wherein highlighting the sub-component comprises modifying one or more pixels surrounding the sub-components to be highlighted by at least one of colour inversion, saturation, increasing contrast, image overlay and changing colour.
8. The method in claim 6, wherein labelling the sub-component with the defect comprises creating a bounding box around the sub-component.
9. The method in claim 1, wherein the one or more defects are detected based on the type of each component.
10. The method in claim 1, wherein a location, in the solar plant, of each component from a list of defective components is identified based on GPS information associated with the set of images, and wherein the list of defective components comprises each component with one or more defects detected.
11. The method in claim 1, wherein a strategy to fix the one or more defects is recommended using reinforcement learning and artificial intelligence techniques.
12. The method in claim 1, wherein generating the health report comprises generating a list of defective components, the one or more defects corresponding to each component in the list of defective components, the location of the defective components, and the strategy to fix the one or more defects of each component.
13. The method in claim 1, wherein a health rating of the solar plant is calculated based on the health report.

14. A system for detecting defects in a solar plant, the system compromising:
a memory; and
a processor coupled to the memory, wherein the processor is configured to execute program instructions stored in the memory for:
receiving, by the processor, a set of images comprising thermal images and visual images of a plurality of components of a solar plant;
recognising, by the processor, a type of the plurality of components in the set of images, wherein the type is at least one of a photovoltaic module, a string combiner box, an inverter, and a transformer;
identifying, by the processor, one or more sub-components of each component using a machine learning model;
determining, by the processor, a temperature of the one or more sub-components based on a thermal image of each component;
detecting, by the processor, one or more defects in each component based on the type of component, and the temperature of the one or more sub-components using a machine learning model; and
generating, by the processor, a health report of the solar plant based on the one or more defects in the plurality of components.

15. A non-transitory computer program product having embodied thereon a computer program for detecting defects in a solar plant, the computer program product storing instructions, the instructions comprising instructions for:
receiving a set of images comprising thermal images and visual images of a plurality of components of a solar plant;
recognising a type of the plurality of components in the set of images, wherein the type is at least one of a photovoltaic module, a string combiner box, an inverter, and a transformer;
identifying one or more sub-components of each component using a machine learning model;
determining a temperature of the one or more sub-components based on a thermal image of each component;
detecting one or more defects in each component based on the type of component, and the temperature of the one or more sub-components using a machine learning model; and
generating a health report of the solar plant based on the one or more defects in the plurality of components.