Iot-based smart facility control automation processing method based on artificial intelligence model

Abstract

The device according to the embodiment collects environmental data and operation data of a facility from an Internet of Things sensor, determines the operation state of the facility by inputting into a pre-trained artificial intelligence model, judges the operation state of the facility according to the collected environmental data and operation data, generates a control command for the facility according to the judgment result, and sends the generated control command to the facility to control the facility according to the control command.

Description

Technical Field

[0001] The following embodiments relate to an intelligent facility control automation technology that combines real-time data collected by Internet of Things (IoT) sensors with learned artificial intelligence models to determine the facility's operating status and automatically optimize control strategies. Background Technology

[0002] Building facilities (such as air conditioning, ventilation, lighting, etc.) are crucial for maintaining indoor comfort and energy efficiency, and their control and management are becoming increasingly important. This is especially true in environments such as large commercial facilities, hospitals, and smart offices, where multiple facilities operate simultaneously. Regularly checking their status and controlling them in an optimal manner is directly related to managerial efficiency and cost reduction.

[0003] Therefore, technologies that collect facility status and environmental data through IoT sensors and automatically control them based on these data are being actively introduced. However, most systems rely on fixed conditions or rule-based logic, which makes them less adaptable to environmental changes or user conditions.

[0004] In recent years, artificial intelligence (AI)-based control methods have gained attention for their ability to enable efficient facility operation, prevent malfunctions, and create customized environments. In particular, AI models can learn from past facility operating history, sensor response patterns, and environmental change data to predict the likelihood of anomalies or control failures in specific facilities and autonomously generate optimal control commands. This technology offers the advantage of minimizing administrator intervention while achieving a high degree of autonomy and efficiency, and is expected to become a core technology for future building facility management.

[0005] Therefore, there is a need to develop an IoT-based intelligent facility control automation service based on an artificial intelligence model.

[0006] Existing technical documents

[0007] Patent documents

[0008] Patent Document 1: Korean Patent Registration No. 10-1899514 (Published on September 18, 2018)

[0009] Patent Document 2: Korean Patent Registration No. 10-2105589 (Published on April 28, 2020)

[0010] Patent document 3: Korean Patent Registration No. 10-1885720 (Announced on August 6, 2018)

[0011] Patent document 4: Korean Patent Registration No. 10-2435561 (Announced on August 23, 2022) Summary of the Invention

[0012] The problem that the invention aims to solve

[0013] In this embodiment, based on environmental and facility data collected from IoT sensors, the artificial intelligence model autonomously determines the operating status of the facility and automatically generates and executes optimal control commands accordingly.

[0014] The purpose of this embodiment is to provide a self-testing method in which the device self-diagnoses the condition and performs preliminary checks and preventative maintenance.

[0015] The purpose of this implementation is to automate the manager approval process based on the risk level and failure history of control commands, and to apply approval steps as appropriate to ensure the stability and reliability of facility control.

[0016] The purpose of this invention is not limited to the above-mentioned purposes, and other purposes not mentioned can be clearly understood from the following description.

[0017] Methods for solving problems

[0018] According to an embodiment of the present invention, the IoT-coupled smart facility control automation processing method based on an artificial intelligence model includes the steps of collecting environmental data of the space where the facility is located and the facility's operational data from a method executed by the device; determining the facility's operational state by inputting the facility into a pre-trained artificial intelligence model, and judging the facility's operational state based on the collected environmental data and operational data; generating a control command for the facility based on the judgment result; and transmitting the generated control command to the corresponding device, wherein the control command may include the step of controlling the device according to the control command.

[0019] The method executed by the device further includes a step of creating a self-test order based on the judgment result. The step of generating the self-test order involves obtaining an operating history, including the device's usage history and the inspection history; checking the facility's cumulative operating time, usage frequency, number of anomalies, anomaly occurrence cycle, manufacturing time, etc., based on the operating history; calculating the facility's aging degree within the space; checking a preset sensitivity level based on the space's usage; calculating the space's sensitivity by checking the number of people, dwell time, activity level, and noise level within the space from environmental data; obtaining the user's clothing information within the space from environmental data; determining the temperature experience mixing degree by checking the length, material, and structure of the user's clothing from the clothing information; checking the user's clothing style from the clothing information; determining the situation mixing degree based on the number of clothing styles determined above; determining the degree of personality diversity by checking the color of the user's clothing from the clothing information; calculating the user's sensitivity within the space based on the degree of temperature experience mixing, the degree of situation mixing, and personality diversity; and calculating the sensitivity of the situation and the user based on the sensitivity of the situation and the user's sensitivity. The process includes: determining the degree of collision in the space; determining the space's need level based on the aging level of the equipment and the degree of collision in the space; selecting the equipment to be inspected based on whether to perform a self-inspection according to the degree of inspection need; determining the basic inspection cycle based on the aging level of the equipment; calculating the degree of environmental change based on environmental data; calculating the degree of change in the values ​​of environmental data and operational data collected from IoT sensors; applying weights determined based on the degree of environmental change and changes in sensor data values ​​to the basic inspection cycle to determine the self-inspection cycle; determining the self-inspection time by considering the distribution of people in the space, the space usage schedule, and the time-based electricity price in order to select the time to perform the inspection according to the inspection cycle; determining a first weight inversely proportional to the aging level of the equipment to minimize the load on the equipment; determining a second weight inversely proportional to changes in sensor data values; and determining the self-inspection intensity, including duration and application intensity, by applying the first and second weights to the degree of inspection need. This may include a step of generating a self-inspection mode by applying the self-inspection intensity to the equipment based on preset stimulus types; and generating an inspection sequence for the inspected equipment based on the self-inspection cycle, self-inspection time, and self-inspection mode.

[0020] The steps of transmitting generated control commands to relevant devices and controlling the devices according to the control commands include: transmitting the generated control commands to the administrator's device; transmitting the control commands to the administrator's terminal upon receiving approval of the control commands from the administrator's terminal; obtaining the control history of devices in the space within a preset time period; confirming similar control commands created based on preset standards in the control history; determining the degree of impact by examining the number of received complaints, the increase in energy consumption, and the time required to restore the space environment after executing similar control commands from the control history; automatically determining the manager approval structure and aging level of the control commands based on the input of the space's situational sensitivity and user sensitivity; and transmitting the control commands according to the manager approval level. The steps leading to the administrator terminal include: if the approval received from the administrator meets the preset approval conditions, then the control command, including the step of controlling the device, is executed according to the control command. If the administrator approval level is 0, the administrator approval structure automatically transmits and executes the generated control command to the relevant device without administrator approval. If the administrator approval level is 1, the control command is transmitted to an administrator terminal. If approval is received from an administrator terminal, the control command is executed. If the administrator approval level is 2, the control command received by the administrator terminal is the same as the standard number determined based on the running status judgment result. If approval is received from the administrator terminal exceeding the preset minimum number of approvers, the control command is executed. If the administrator approval level is 3, it may include a step of receiving the same number of control commands as the standard personnel to the administrator terminal. If approval is received from the administrator terminal corresponding to all reference personnel, the control command is executed.

[0021] The steps for creating control commands for the device based on the above judgment results, and the subsequent steps for providing control change notifications to the administrator terminal, include: comparing existing control commands with created control commands; extracting changed control conditions, added control conditions, and deleted control conditions; receiving condition information about the control change notification from the manager terminal; comparing existing control commands with created control commands; generating control change notifications based on the changed control commands if the comparison results satisfy the condition information; calculating the degree of change of the control commands based on the comparison results; calculating the change time from the most recent execution time of the existing control commands to the time of execution of the created control commands; and determining the importance of the notification by applying a weight determined based on the change period to the degree of change. This may include a step of setting the notification strength of the control change notification based on the notification importance, and a step of providing the control change notification to the administrator terminal with the applied notification strength.

[0022] The steps for collecting control fault data from operational history include: equipment control fault history, management feedback records, sensor response anomaly logs, and fault history; generating fault control commands using a pre-trained generative AI model based on the fault control conditions, executed command sequences, applied facility groups, and environmental response data from the control fault data; calculating the similarity between the initial control command to be applied to the facility and the fault control command; predicting the probability of control failure for the corresponding initial control command based on the similarity; and generating a modification request notification to modify the initial control command if the predicted control failure probability is greater than a preset threshold, and providing it to the administrator's terminal.

[0023] The apparatus according to one embodiment can be combined with hardware and controlled by a computer program stored on a medium to perform any of the methods described above.

[0024] Invention Effects

[0025] The implementation can analyze the facility's operational status in real time and automatically control it to perform efficient facility operation without the need for management personnel intervention.

[0026] By self-diagnosing the condition of the equipment and optimizing the timing and intensity of inspections, the implementation can prevent malfunctions in advance and reduce maintenance costs.

[0027] By automatically adjusting the manager's approval process based on the risk and environmental sensitivity of control commands, the implementation can simultaneously ensure the security and flexibility of facility control.

[0028] On the other hand, the effects of the embodiments are not limited to those described above, and other unmentioned effects can be clearly understood by those skilled in the art from the following description. Attached Figure Description

[0029] Figure 1 This is a schematic diagram illustrating the system configuration according to one embodiment.

[0030] Figure 2 It is a flowchart illustrating, according to one embodiment, the automated processing of IoT-coupled smart facility control based on an artificial intelligence model.

[0031] Figure 3 It is a flowchart illustrating the process of generating a self-inspection order based on the judgment results of a single embodiment.

[0032] Figure 4 This is a flowchart illustrating the process of transmitting control commands to an administrator terminal according to a single embodiment.

[0033] Figure 5 This is a flowchart illustrating the process of providing control change notifications to an administrator terminal according to an embodiment.

[0034] Figure 6 To illustrate, according to one embodiment, the process of providing a modification request notification requesting modification of an initial control command based on control failure data.

[0035] Figure 7 This is an example of a device configuration according to one embodiment. Detailed Implementation

[0036] The embodiments have been described in detail herein with reference to the accompanying drawings. However, the scope of the patent application is not limited to these embodiments, as various changes can be made to the embodiments. All modifications, equivalents, or substitutions to the embodiments should be understood to be included within the scope of the claims.

[0037] The specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and can be modified and performed in various forms. Therefore, the embodiments are not limited to the specific form of disclosure, and the scope of this specification includes modifications, equivalents, or substitutions incorporated into the technical concept.

[0038] Terms such as "first" or "second" can be used to describe various components, but the interpretation of these terms should only be used to distinguish one component from another. For example, the first component can be named the second component, and similarly, the second component can be named the first component.

[0039] When a component is said to be "connected" to another component, it should be understood that it may be directly connected to another component or connected to another component, but there may be another component between them.

[0040] The terminology used in these embodiments is for illustrative purposes only and should not be construed as limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this alternative text, the terms “comprising” or “having” should be understood to mean the presence of any feature, number, step, movement, component, part, or combination thereof as described herein, and does not exclude the presence or addition of one or more other features, numbers, steps, movements, components, parts, or combinations thereof.

[0041] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of common sense in the art to which the embodiments pertain. Terms such as those defined in common dictionaries should be interpreted as having the meaning consistent with their meaning in the relevant descriptive context, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0042] Furthermore, when describing the embodiments with reference to the accompanying drawings, the same reference code will be provided for the same component regardless of the drawing code, and repeated descriptions will be omitted. When describing embodiments, detailed descriptions should be omitted if it is determined that a detailed description of the relevant notifying technology may unnecessarily obscure the essential points of the embodiment.

[0043] The embodiments can be implemented in various types of products, such as personal computers, laptops, tablets, smartphones, televisions, smart home appliances, smart cars, kiosks, wearable devices, etc.

[0044] In this invention, Artificial Intelligence (AI) refers to the technology that mimics human learning, reasoning, and perceptual abilities and is implemented in computers. It can include concepts such as machine learning and symbolic logic. Machine learning (ML) is an algorithmic technique that autonomously classifies or learns the features of input data. Artificial intelligence technology is a type of machine learning algorithm that analyzes input data, learns from the analysis results, and makes judgments or predictions based on those results. Furthermore, the technology of using machine learning algorithms to simulate the cognitive and judgmental functions of the human brain can also be understood as falling within the scope of artificial intelligence. For example, it may include technical fields such as language understanding, visual understanding, reasoning / prediction, knowledge representation, and motion control. Machine learning can refer to using experience in processing data to train a neural network model. Machine learning may mean that computer software has improved its ability to process data autonomously. Neural network models are built by modeling the correlations between data, and these correlations can be represented by multiple parameters. Neural network models extract and analyze features from given data to derive the correlations between the data. Repeating this process to optimize the parameters of the neural network model can be called machine learning. For example, a neural network model can learn the mapping (correlation) between the inputs and outputs of data given in the form of input / output pairs. Alternatively, neural network models can learn relationships by inferring patterns between given data, even if only the input data is given. AI learning models or neural network models can be designed to implement the structure of the human brain on a computer and can include multiple network nodes that are weighted by simulating neurons in a human neural network. Multiple network nodes can interconnect by simulating the synaptic activity of neurons that send and receive signals through synapses. In an AI learning model, multiple network nodes reside at different depths and can exchange data based on convolutional connections. For example, an AI learning model can be an artificial neural network, a convolutional neural network (CNN), etc. As an example, an AI learning model can perform machine learning based on supervised learning, unsupervised learning, reinforcement learning, etc. Machine learning algorithms used to perform machine learning can include decision trees, Bayesian networks, support vector machines, artificial neural networks, Ada-boost, perceptrons, genetic programming, and clustering. Among these, CNNs are a type of multilayer perceptron designed to use minimal preprocessing. A CNN consists of one or more convolutional layers and general neural network layers above them, with additional weights and pooling layers. Due to this structure, CNNs can fully utilize input data in a two-dimensional structure. Compared to other deep learning architectures, CNNs perform well with both video and audio. CNNs can also be trained using standard backtracking. CNNs are easier to train than other feedforward neural network techniques and have the advantage of using fewer parameters. A convolutional network is a neural network that contains a set of nodes with string parameters.The increased availability of training data and computational power, coupled with advancements in algorithms such as unique linear units and dropout training, have led to significant improvements in many computer vision tasks. Due to the sheer size of datasets—such as those available for many tasks today—outfitting is less critical; increasing network size can improve test accuracy. Optimal use of computational resources becomes a limiting factor. For this, distributed, scalable deep neural networks can be used. The embodiments described above can be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments can be implemented using one or more general-purpose or special-purpose computers, such as processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field-programmable gate arrays (FPGAs), programmable logic units (PLUs), microprocessors, or any other device capable of executing and responding to instructions. The processing unit can execute an operating system (OS) and one or more software applications running on the operating system. Furthermore, the processing unit can access, store, process, and generate data in response to software execution. For ease of understanding, a processing unit can be described as a single processing unit, but a person with general knowledge in the relevant field may know that a processing unit may contain multiple processing elements and / or various types of processing elements. For example, a processing unit may contain multiple processors or one processor and one controller.

[0045] Figure 1 This describes the system configuration based on one embodiment. (See also...) Figure 1 According to one embodiment, the system may include an administrator terminal 10 and a device 30, which can communicate with each other via a communication network. First, the communication network can be configured regardless of communication methods such as wired and wireless, and can be implemented in various forms to enable communication between servers and between servers and terminals. The administrator terminal 10 may be a terminal used by an administrator to monitor or approve AI-based facility control automation services. Yes. The administrator terminal 10 may be a desktop computer, laptop computer, tablet computer, smartphone, etc. For example, as... Figure 1As shown, the administrator's terminal 10 can be a smartphone, and may take different forms depending on the embodiment. The administrator's terminal 10 may be configured to perform all or part of the computing, storage / reference, input / output, and control functions of a common computer. The administrator's terminal 10 may be configured to communicate with device 30 via wired or wireless communication. The administrator's terminal 10 may access web pages created by individuals or entities using services provided by device 30, or may install applications developed and distributed by individuals or organizations using services provided by device 30. The administrator's terminal 10 may link to device 30 via web pages or applications. The administrator's terminal 10 may connect to web pages provided by device 30. Device 30 may be accessed via applications. The singular expression in the claims may be understood to include the plural. Using an artificial intelligence model, the operating status of a facility can be determined based on environmental data of the space and facility operation data collected from IoT sensors, and automatic control commands can be generated and transmitted accordingly. Specifically, device 30 can comprehensively analyze the facility's aging level, inspection history, environmental changes, and user sensitivities to determine control requirements or generate self-test commands using a pre-trained artificial intelligence model. Furthermore, device 30 can automatically branch the manager approval structure of the determined control commands and execute procedures requiring single or multi-level manager approval based on the approval level. Device 30 dynamically determines the intensity, timing, and frequency of self-tests based on environmental conditions and can automatically generate calibration control commands or issue notifications to management personnel for equipment requiring inspection. (This will be discussed later.) Figure 2 The configuration and operation steps of the IoT-coupled smart facility control automation processing method based on an artificial intelligence model for device 30 are described in detail. Device 30 can be its own server, owned by an individual or entity using the services provided by device 30, or it can be a cloud server, or it can be a peer-to-peer (P2P) collection of distributed nodes. Device 30 can be configured to perform all or part of the computing, storage / reference, I / O, and control functions of a regular computer. Device 30 can be configured to communicate with the administrator's terminal 10 via wired and wireless means, and can control the operation of the administrator's terminal 10 and the information displayed on the administrator's terminal 10 screen. On the other hand, for ease of explanation, Figure 1 Only terminal 10 of the manager is shown, but the number of terminals can vary depending on the embodiment. There is no particular limitation on the number of terminals, provided the processing power of device 30 allows. According to one embodiment, a database can be set up internally within device 30, but is not limited to this; the database can be configured separately from device 30. Device 30 may include many artificial neural networks for improving the performance of machine learning algorithms.

[0046] Figure 2 This is a flowchart illustrating an automated processing procedure for controlling IoT-coupled smart facilities based on an artificial intelligence model, according to one embodiment. (Reference) Figure 2 First, in step S201, device 30 can collect environmental data and facility operation data from the space where the facility is located from IoT sensors. Device 30 collects real-time data from one or more IoT sensors placed within the facility installation space. At this time, the IoT sensors can detect environmentally related data, such as temperature, humidity, illuminance, carbon dioxide concentration, airflow, and noise, and from sensors connected to the facility itself, it can obtain facility operation information, such as operating cycle, power consumption, output status, vibration value, etc.

[0047] In phase S202, device 30 can determine the facility's operational status by inputting data into a pre-trained artificial intelligence model, thereby determining the facility's operational status based on collected environmental and operational data. Device 30 can input data into a pre-trained artificial intelligence model based on the collected dataset to determine the facility's operational status. At this point, the AI ​​model can be configured to include at least one of supervised learning, unsupervised learning, and reinforcement learning, and can be trained to perform tasks such as identifying normal / abnormal equipment, predicting the probability of failure, and determining whether there is an overload.

[0048] In step S203, device 30 can generate control commands for the facility based on its judgment results. Device 30 can automatically generate control commands optimized for the facility based on the judgment results of the artificial intelligence model. When generating control commands, device 30 can consider complex conditions such as the current state of the facility, energy consumption status, user settings, and the sensitivity of the space. For example, if the CO2 level of the air conditioning facility exceeds a certain threshold, and the space is set as a conference room, a control command can be configured to ventilate the room at maximum airflow for 10 minutes, and then switch to medium mode.

[0049] In step S204, device 30 transmits the generated control command to the appropriate device, allowing control of the device according to the control command. Device 30 may use a communication method or protocol that can be connected to the facility's control system to send the command to control the facility based on the generated control command.

[0050] In step S305, device 30 can generate inspection commands for the facility under inspection based on the self-inspection cycle, timing, and mode. Device 30 can generate specific inspection orders for the facility by combining the determined self-inspection cycle, time, and mode, enabling efficient facility maintenance. The inspection command includes the operating conditions, start time, intensity, duration, and repetitiveness of the self-inspection work to be performed by the plant, and can generate customized commands for each plant. In this way, device 30 can automatically order inspection work optimized for each facility's conditions and space constraints.

[0051] In step S506, device 30 can determine the importance of a notification by applying a weight determined based on the duration of the change to the degree of change. Device 30 can calculate the final notification importance by combining the calculated degree of change and the duration of the change. Specifically, the greater the degree of change and the shorter the change period, the higher the notification importance can be set, and the notification importance can be represented using a scale of 0 to 1 or 0 to 100. This is not limited to this.

[0052] In step S507, device 30 can change the notification strength according to the notification importance setting control. If the notification importance is less than 0.3, device 30 can set the notification strength to "low"; if the notification importance is greater than 0.3 but less than 0.6, it can be set to "medium"; if it is greater than 0.6, it can be set to "high"; and if it is greater than 0.8, it can be set to "urgent". Furthermore, custom notification strength settings can be applied as needed, such as adjusting the strength for specific spaces and facilities.

[0053] Figure 7 This is an example of the composition of device 30 according to one embodiment.

[0054] The apparatus 30 according to one embodiment includes a processor 31 and a memory 32. The apparatus 30 according to one embodiment may be the server or terminal described above. The processor 31 can... Figures 1 to 6 Includes at least one of the aforementioned devices, or can be used Figures 1 to 6 At least one of the aforementioned methods may be performed. Memory 32 may store information related to the aforementioned methods, or it may store a program that implements the methods. Memory 32 may be volatile memory or non-volatile memory.

[0055] Processor 31 can execute programs and control device 30. The code of the program executed by processor 31 can be stored in memory 32. Device 30 is connected to external devices (e.g., personal computers or networks) via input / output devices (not shown) and can exchange data.

[0056] For example, the devices, methods, and components described in the embodiments are, for example, processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field-programmable gate arrays (FPGAs), programmable logic units (PLUs), microprocessors, or any other device capable of executing and responding to instructions. It can be implemented using one or more general-purpose or special-purpose computers. The processing unit can execute an operating system (OS) and one or more software applications executed on the operating system. Furthermore, the processing unit can access, store, process, and generate data in response to software execution. For ease of understanding, a processing unit can be described as a single processing unit, but those skilled in the art will recognize that a processing unit may contain multiple processing elements and / or various types of processing elements. For example, a processing unit may contain multiple processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

[0057] The method according to the embodiments is implemented in the form of program instructions that can be executed in various computer ways and can be recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., individually or in combination. The program instructions recorded in the medium may be specifically designed and configured for the embodiments or may be known and available to computer software vendors. Examples of computer-readable recording media include magnetic media (such as hard disks, floppy disks, and magnetic tapes), optical media (such as CD-ROMs and DVDs), magneto-optical media (such as floppy disks), and hardware devices (such as ROMs, RAMs, flash memory, etc.) specifically configured to store and execute program instructions. Examples of program instructions include machine code (e.g., code created by a compiler) and high-level language code (e.g., an interpreter) that can be executed by a computer. The hardware device may be configured to run as one or more software modules to perform the operations of the embodiments, and vice versa.

[0058] The software may include computer programs, code, instructions, or one or more identical combinations, and may be configured to run the processing unit as needed, or may command the processing unit independently or collectively. The software and / or data may be permanently or temporarily contained in any type of machine, component, physical device, virtual device, computer storage medium or apparatus, or in transmitted signal waves for interpretation by the processing unit or to provide instructions or data to the processing device. The software may be distributed through networked computer systems and stored or executed in a distributed manner. The software and data may be stored on one or more computer-readable recording media.

[0059] Although the above embodiments are described with limited drawings, those skilled in the art can make various technical modifications and alterations based on the above description. For example, appropriate results may be obtained if the described techniques are performed in a different order than the described methods, and / or if the components of the described systems, structures, devices, circuits, etc., are combined or arranged in a different form than the described methods, or if they are replaced or substituted by other components or equivalents.

[0060] Therefore, other embodiments, other embodiments, and equivalents having the patent claims also fall within the scope of the following claims.

Claims

1. A method for automating the control of IoT-based smart facilities based on an artificial intelligence model, wherein, include: Depending on how the equipment operates, environmental data of the installation space and operational data of the facilities are collected from IoT sensors; The facility's operational status is determined by inputting the facility into a pre-trained artificial intelligence model, based on the environmental and operational data collected above. Based on the results of the above judgment, control commands for the facilities are generated; and The generated control commands are transmitted to the device, and the device is controlled according to the control commands.

Images