Elevator system and elevator control method

Abstract

To perform operation in consideration of the feeling of an elevator user.SOLUTION: An elevator system of an embodiment includes: a face detection unit that detects a face of an elevator user reflected on an image captured by a camera installed in a landing or a car of an elevator; an expression analysis unit that analyzes an expression of the face of the elevator user to generate an analysis result in which various emotions included in the expression are digitized; an information management unit that manages management information indicating a correspondence relation between the various emotions of the elevator user and various elevator controls; and a control determination unit that, based on the analysis result and the management information, determines an elevator control to be executed for the elevator user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] FIELD OF THE INVENTION Embodiments of the present invention relate to elevator systems and elevator control methods. [Background technology]

[0002] Elevator operation has been improved from various perspectives, for example, by using an image sensor to detect a child's facial expression and movements to determine whether they are sleeping, and by activating a sleeping mode that reduces the volume and slows down the door opening and closing speed if the child is sleeping, or by efficiently storing photographic data, including facial expressions, taken by a security camera to deter crimes inside the elevator. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-121690 [Patent Document 2] Japanese Patent Publication No. 2023-054161 Summary of the Invention [Problem to be solved by the invention]

[0004] However, there are still many proposals for improving elevator operation that take into consideration the emotions of elevator users. For example, conventional elevator systems do not provide operation that takes into consideration the emotions of passengers who are using the elevator for the first time and are confused about how to operate it, or when passengers are feeling anxious.

[0005] The problem to be solved by the invention is to provide an elevator system and an elevator control method that can operate the elevator while taking into consideration the feelings of elevator users. [Means for solving the problem]

[0006] The elevator system of the embodiment includes a face detection unit that detects the face of an elevator user captured in an image from a camera installed at the elevator landing or inside the elevator car; an expression analysis unit that analyzes the facial expressions of the elevator user to generate analysis results that quantify various emotions contained in the expressions; an information management unit that manages management information that indicates the correspondence between the various emotions of the elevator user and various elevator controls; and a control decision unit that decides the elevator control to be executed for the elevator user based on the analysis results and the management information. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram illustrating an example of the configuration of an elevator system according to an embodiment. [Figure 2] FIG. 2 is a diagram showing an example of the configuration of a hall. [Figure 3] FIG. 3 is a diagram showing an example of the configuration of the area around the entrance and exit in the elevator car. [Figure 4] FIG. 4 is a diagram showing an example of an analysis result regarding the facial expressions of elevator users. [Figure 5] FIG. 5 is a diagram showing an example of management information relating to the same elevator users as the analysis results of FIG. [Figure 6] FIG. 6 is a flowchart showing an example of the operation of the facial expression analysis system. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, embodiments will be described with reference to the drawings.

[0009] <Elevator system configuration> FIG. 1 is a diagram illustrating an example of the configuration of an elevator system according to an embodiment.

[0010] The elevator system shown in Figure 1 is a system for managing elevator groups, and includes an elevator group management control device 20, elevator control devices 11a to 11c and cars 12a to 12c that make up multiple elevators, and a hall call registration device 15 and cameras (imaging devices) 16a to 16c installed in a hall (elevator hall) 14 on each floor.

[0011] Here, an example is shown in which three elevators, A to C, are managed as a group. The elevators A to C are respectively associated with three elevator control devices 11a to 11c, three cars 12a to 12c, and three cameras 16a to 16c.

[0012] Elevator control devices 11a to 11c are provided for each of the elevator cars 12a to 12c, and perform operation control of the corresponding elevator car. Specifically, elevator control devices 11a to 11c control motors (hoists, not shown) for raising and lowering the elevator cars 12a to 12c, and control the opening and closing of doors. These elevator control devices 11a to 11c are configured using computers.

[0013] The cars 12a to 12c are driven by motors to move up and down in the elevator shaft. Load sensors 13a to 13c are installed in the cars 12a to 12c, respectively, to detect the load inside the car. Load data detected by the load sensors 13a to 13c is transmitted to the group management control device 20 via the elevator control devices 11a to 11c.

[0014] Furthermore, cameras (imaging devices) 18a to 18c are installed in the elevator cars 12a to 12c, respectively. The cameras 18a to 18c photograph elevator users in the cars. The images obtained by photographing (photographed images) are transmitted to a facial expression analysis system 22 in the group management control device 20 via the elevator control devices 11a to 11c.

[0015] The group management control device 20 is a device for group management and control of the operation of multiple elevator cars 12a to 12c, and is configured using a computer, similar to the elevator control devices 11a to 11c. The group management control device 20 is equipped with a call storage unit 21, a facial expression analysis system 22, and an operation control unit 23.

[0016] The call storage unit 21 stores hall calls registered by operating the hall call registration device 15.

[0017] The facial expression analysis system 22 is a device that presents elevator operation suggestions that take into consideration the emotions of elevator users. The facial expression analysis system 22 is configured to be able to communicate with the information terminal 10.

[0018] The information terminal 10 allows the operator to view and change some or all of the management information managed by the facial expression analysis system 22 using applications and various information provided by the facial expression analysis system 22.

[0019] The facial expression analysis system 22 is configured using one or more computers, and various functions provided by the facial expression analysis system 22 are realized as programs executed by the one or more computers. Furthermore, at least some of the various functions provided by the facial expression analysis system 22 may be located in a location other than the facial expression analysis system 22, for example, in a server on a communication network (on the cloud) that can communicate with the facial expression analysis system 22. The various functions provided by the facial expression analysis system 22 will be described in detail later.

[0020] The operation control unit 23 controls the operation of multiple elevators through the elevator control devices 11a to 11c. The operation control unit 23 performs general group management control unless instructed to do so by the control decision unit 22d of the facial expression analysis system 22, but if instructed to execute a certain control by the control decision unit 22d of the facial expression analysis system 22, the operation control unit 23 gives priority to the execution of that control.

[0021] When a new hall call is registered, the operation control unit 23 selects a car from among the cars 12a to 12c to which the hall call should be assigned and causes the car to respond to the corresponding hall 14 (causing the car to head to the hall 14 on the floor where the hall call is registered).

[0022] 2 is a diagram showing an example of the configuration of the hall 14. However, this is just an example and the present invention is not limited to this example.

[0023] A hall call registration device 15 for registering hall calls is installed at the hall (elevator hall) 14 on each floor. In the example of FIG. 2, for convenience, a hall call registration device 15 installed on an arbitrary floor is shown, but in reality, at least one hall call registration device 15 is installed on each floor. The hall call registration device 15 is connected to the group management control device 20 via a transmission cable (not shown). The hall call registration device 15 has direction buttons (also called "hall call buttons") that allow elevator users to specify their destination direction (upward / downward).

[0024] Here, a "hall call" is a call signal registered by operating a direction button at a platform, and includes information on the registered floor and destination direction. In contrast, a "car call" is a call signal registered by operating a destination floor button inside a car, and includes information on the car and destination floor.

[0025] Furthermore, cameras 16a to 16c are installed for each elevator car in the landing (elevator hall) 14 on each floor. The cameras 16a to 16c are installed near the landing doors 17a to 17c, respectively, and capture images including the faces of multiple elevator users waiting in front of the arrival gates of the elevator cars 12a to 12c. The images obtained by the capture are transmitted to a facial expression analysis system 22 in the group management control device 20.

[0026] Next, one of the cars 12a to 12c will be referred to as the car 12 for convenience, and the internal configuration of the car 12 will be described.

[0027] 3 is a diagram showing an example of the configuration of the area around the entrance / exit inside the car 12. However, this is just an example and the present invention is not limited to this example.

[0028] A car door 40 is provided at the entrance of the car 12 so as to be able to open and close freely. The example in Fig. 3 shows a double-door type car door 40, and two door panels 40a, 40b constituting the car door 40 open and close in opposite directions along the width direction (horizontal direction).

[0029] Entrance pillars 41a and 41b are provided on both sides of the entrance of the car 12. In the example of Fig. 3, when the car door 40 is opened, one door panel 40a is stored in a door pocket 42a provided on the back side of the entrance pillar 41a, and the other door panel 40b is stored in a door pocket 42b provided on the back side of the entrance pillar 41b.

[0030] Furthermore, one or both of the entrance pillars 41a, 41b are equipped with a display device 43 having a display panel, an operation panel 45 on which destination floor buttons 44 and the like are arranged, and a speaker 46. In the example of Fig. 3, the speaker 46 is installed on the entrance pillar 41a, and the display device 43 and operation panel 45 are installed on the entrance pillar 41b. The ceiling is equipped with a camera 18, which is one of the above-mentioned cameras 18a to 18c, as well as lighting 47, air conditioning equipment 48, and the like.

[0031] When an elevator user in a car 12 presses a destination floor button 44 on an operation panel 45, the destination floor specified by the destination floor button 44 is registered as a car call. The car call registered in the car 12 is sent to the group management control device 20 via the elevator control device 11 corresponding to the car 12.

[0032] <Functional configuration of facial expression analysis system 22> 1 has, as various functions, a face detection unit 22a, a facial expression analysis unit 22b, an information management unit 22c, a control decision unit 22d, a control evaluation unit 22e, and an evaluation reflection unit 22f. These functions will be described below.

[0033] In this system, elevator control can be customized for each elevator user to ensure comfortable operation for each elevator user. To avoid complicating the explanation, we will explain an example in which various functions process a single elevator user.

[0034] The face detection unit 22a has a function of detecting the face of an elevator user captured in an image captured by a camera installed at the elevator hall or inside the elevator car.

[0035] The facial expression analysis unit 22b has a function of generating an analysis result that quantifies various emotions of the elevator user by analyzing the facial expressions of the elevator user. The analysis result includes an analysis result before the elevator control described below is executed, as well as an analysis result after the elevator control is executed. Details of the analysis result will be described later with reference to FIG. 4.

[0036] The information management unit 22c has a function of managing management information indicating the correspondence between the various emotions of the elevator users and various elevator controls. Details of the management information will be described later with reference to FIG. 5.

[0037] The control decision unit 22d is a function that decides elevator control to be executed for the elevator user based on the analysis result and management information.

[0038] The control evaluation unit 22e is a function that evaluates the effectiveness of the elevator control that has been decided to be executed, based on the difference between the analysis result before the execution of the elevator control and the analysis result after the execution of the elevator control.

[0039] The evaluation reflecting unit 22f is a function for reflecting the result of evaluating the effectiveness of the elevator control in the management information.

[0040] The various functions described above will be explained in detail below.

[0041] Face detection unit 22a The face detection unit 22a inputs images captured by the cameras 16a to 16c installed at each floor's landing and each car, and also inputs images captured by the cameras 18a to 18c for each car, and detects the face of the elevator user if the face of the elevator user is captured in any of the images. Face detection may be performed using AI (Artificial Intelligence). For example, the accuracy of face detection can be improved by inputting images of many people's faces into a predetermined model used by the AI in advance and having the AI learn features such as the positions of the eyes, nose, and mouth.

[0042] Furthermore, the face detection unit 22a refers to history information that records past face detection results as a database, determines whether the detected elevator user's face matches any of the faces of elevator users detected in the past, and if they match, recognizes the user as a returning elevator user, while if they do not match, recognizes the user as a first-time elevator user. If the elevator user is a first-time elevator user, the face detection result of that elevator user is recorded in the history information.

[0043] The history information is assigned an identification number to each elevator user and is stored in the information management unit 22c (described later) as information including data showing facial features specific to the elevator user. By referring to the history information, the face detection unit 22a can identify whether the detected elevator user is a returning elevator user or a first-time elevator user.

[0044] Although an example of detecting a face is shown here, the present invention is not limited to this example, and elements other than the face (such as the movement of hands, feet, or the whole body) may be detected. For example, by detecting the movement of the whole body, an analysis result indicating an emotion such as restlessness can be obtained in the analysis described below.

[0045] ·Facial expression analysis section 22b The facial expression analysis unit 22b analyzes the facial expressions of elevator users detected by the face detection unit 22a. In this analysis, an emotion value indicating the degree of each emotion is calculated for each pre-defined emotion classification (type). The facial expression analysis unit 22b compiles the information obtained in this analysis to generate an analysis result. The analysis of facial expressions may be performed using AI. For example, the accuracy of the facial expression analysis can be improved by inputting facial images showing various emotions (e.g., various facial images showing various emotions such as "confused," "stressed," "relaxed," "in a hurry," "anxious," "enjoyed," "tired," "angry," "excited," etc.) into a predetermined model used by the AI and repeatedly learning the characteristics of the positions, shapes, and movements of the eyes, nose, and mouth for each emotion type.

[0046] An example of the analysis result of the facial expression of the elevator user detected by the face detection unit 22a is shown in Fig. 4. However, this is just an example and the present invention is not limited to this example.

[0047] The analysis results shown in FIG. 4 include analysis results before the execution of elevator control that has been decided to be executed by the control decision unit 22d described later, as well as analysis results after the execution of the elevator control.

[0048] Specifically, as shown in Figure 4, in addition to the item "emotion classification number," there are also items related to when getting on the elevator (i.e., before elevator control is executed), such as "type of emotion when getting on," "emotion value when getting on," and "emotion classification," as well as items related to when getting off the elevator (i.e., after elevator control is executed), such as "type of emotion when getting off," "emotion value when getting off," and "emotion classification."

[0049] In the "emotion classification number" column, a number for a predetermined emotion classification (type) is written.

[0050] The columns for the items related to riding, "Type of emotion while riding," "Emotion value while riding," and "Emotion classification," respectively, list the type of emotion while riding (for example, "confused," "stressed," "relaxed," "in a hurry," "anxious," "enjoyed," "tired," "angry," "excited," etc.), the emotion value indicating the degree of that emotion (for example, a value in the range from 0 to 100), and information indicating whether the emotion is "positive" or "negative."

[0051] The columns for the items related to disembarking, "Type of emotion when disembarking," "Emotion value when disembarking," and "Emotion classification," contain information corresponding to the columns for the items related to boarding, "Type of emotion when boarding," "Emotion value when boarding," and "Emotion classification," respectively, such as the type of emotion when disembarking (e.g., "confused," "stressed," "relaxed," "in a hurry," "anxious," "enjoyed," "tired," "angry," "excited," etc.), an emotion value indicating the degree of that emotion (e.g., a value in the range from 0 to 100), and information indicating whether the emotion is "positive" or "negative."

[0052] ·Information Management Department 22c The information management unit 22c manages, on a storage medium, management information indicating the correspondence between the various emotions shown in Fig. 4 and various appropriate elevator controls (hereinafter, may be abbreviated as "controls") that take these emotions into consideration. In addition, the information management unit 22c also manages, on the storage medium, history information of elevator users detected by the face detection unit 22a.

[0053] The storage medium may be located inside the group management control device 20 or outside (for example, in the above-mentioned server) the group management control device 20. The management information and history information managed by the information management unit 22c are used as necessary by the control decision unit 22d, control evaluation unit 22e, evaluation reflection unit 22f, etc., which will be described later.

[0054] In addition, the information management unit 22c collects various information indicating the operation status of the elevator (such as the location of each elevator, running condition, load, presence or absence of hall calls, floor where the hall call was made, and information indicating the assigned elevator that responded to the hall call) and manages it on a storage medium.

[0055] 4 include controls related to elevator operation (e.g., changing the assigned car number, changing the car door opening / closing speed, etc.), as well as controls related to the display device 43, speaker 46, lighting 47, or air conditioning equipment 48 inside the car. Examples include control to display and output predetermined information on the display device 43 inside the car, control to output predetermined information (announcements, music, etc.) as audio from the speaker 46 inside the car, and control to change the brightness or color tone of the lighting 47 inside the car. Other examples include control to change the air conditioning (air volume, etc.) of the air conditioning equipment 48 inside the car.

[0056] For example, for a tired elevator user, the elevator user may be made to relax by applying controls such as displaying soothing images on the display panel of the display device 43, playing soft music from the speaker 46, or changing the brightness or color tone of the lighting 47 to a more subdued one.

[0057] Also, for example, for an elevator user who is enjoying themselves, the mood of the elevator user may be boosted by applying controls such as displaying more enjoyable images on the display panel of the display device 43, playing cheerful music from the speaker 46, or brightening the lighting 47.

[0058] Furthermore, for example, for confused elevator users, a simple explanation of how to operate the control panel 45 may be displayed on the display panel of the display device 43, or guidance may be broadcast from the speaker 46, thereby providing a sense of security to the elevator user.

[0059] In addition, for elevator users who are sweating and feeling hot, control may be applied to increase the amount of cool air sent from the air conditioning equipment 48, so that the elevator users feel comfortable inside the car.

[0060] Fig. 5 shows an example of management information relating to elevator users that is the same as the analysis result of Fig. 4. However, this is just an example, and the present invention is not limited to this example.

[0061] The management information shown in FIG. 5 includes items such as "control number," "emotion type," "control," "execution probability coefficient," and "correlation value."

[0062] In the "control number" column, the elevator control number corresponding to the predetermined emotion classification (type) is written.

[0063] The "Type of Emotion" column is where you write down the type of emotion you are experiencing (e.g., "confused," "stressed," "relaxed," "hurried," "anxious," "enjoyed," "tired," "angry," "excited," etc.).

[0064] The "Control" column describes appropriate elevator control that takes into consideration the emotion described in the "Emotion Type" column. Specific examples of elevator control that correspond to each of the emotions described in the "Emotion Type" column include "automatically displaying simple operation instructions on the display panel," "broadcasting guidance from the speaker to explain how to operate the elevator," "adjusting the lighting softly and playing quiet music," "maintaining the current environmental settings," "increasing the door opening and closing speed," "reassuring announcements and playing relaxing music," "bright lighting and playing cheerful music," "calm lighting and playing calm music," "setting the car environment to a quiet one," and "calm lighting and playing calm music."

[0065] 5, for example, the "type of emotion" column may contain two "confused" options, and the corresponding "control" columns may contain different control contents, "automatically display simple operating instructions on the display panel" and "broadcast guidance from the speaker to explain how to operate." In this case, different values are entered in the "execution probability coefficient" column, which will be described later.

[0066] The elevator control to be executed for the elevator user may be selected from among various elevator controls, or multiple elevator controls may be selected if there are no problems. The elevator control to be executed is selected based on the "correlation value" or "execution probability coefficient" described later, or the value written in the "emotion value when boarding" column shown in Figure 4. The higher the value, the higher the priority of the corresponding elevator control to be selected.

[0067] The "Execution Probability Coefficient" column indicates the value (for example, a value in the range from 0 to 100) of the coefficient (execution probability coefficient) used in a predetermined arithmetic formula that calculates the probability that the corresponding elevator control will be executed. The initial value is, for example, 0. The larger the value of the execution probability coefficient, the higher the probability that the corresponding elevator control will be executed.

[0068] Each execution probability coefficient listed in the "Execution Probability Coefficient" column may be configured to be appropriately changed by the AI through learning. In this case, the execution probability coefficient may be determined depending on the number of times the corresponding elevator control has been executed in the past for the elevator user.

[0069] The "Correlation Value" column contains a correlation value (e.g., a value in the range from 0 to 100) that indicates the degree of correlation between the corresponding emotion and the elevator control. The initial value is, for example, 0. A predetermined value (e.g., 1) is added to the correlation value when the corresponding elevator control is successfully executed. Specifically, the correlation value increases according to the number of times the corresponding elevator control has been successful in the past, and the larger the value, the stronger the relationship that the corresponding elevator control has with a positive impact on the emotion of the elevator user.

[0070] Whether the corresponding elevator control was successful or not can be determined based on the analysis results shown in Figure 4, from the difference between the analysis results before and after the execution of the corresponding elevator control. For example, when the "emotion classification" is "negative," if the numerical value of the "emotion value when getting off" is lower than the numerical value of the "emotion value when getting on" by more than a certain amount, the corresponding elevator control can be considered successful; otherwise, the corresponding elevator control can be considered unsuccessful. On the other hand, when the "emotion classification" is "positive," if the numerical value of the "emotion value when getting off" is higher than the numerical value of the "emotion value when getting on" by more than a certain amount, the corresponding elevator control can be considered successful; otherwise, the corresponding elevator control can be considered unsuccessful.

[0071] Note that the various elevator controls listed in the "Control" column in Fig. 5 are not necessarily fixed, and may be configured so that the operator of the information terminal 10 can change them as needed, or so that the AI can change them appropriately by learning using a genetic algorithm or the like. For example, a more appropriate combination of elevator controls may be set by repeatedly having a predetermined model used by the AI learn successfully executed and unsuccessfully executed elevator controls for each type of emotion.

[0072] Control decision unit 22d The control decision unit 22d decides elevator control to be executed for the elevator user based on the analysis result generated by the facial expression analysis unit 22b and the management information managed by the information management unit 22c, that is, based on the "correlation" or "execution probability coefficient" shown in Fig. 5 as described above, or the numerical value written in the "emotion value when boarding" column shown in Fig. 4. Furthermore, the control decision unit 22d instructs the operation control unit 23 to execute the elevator control that has been decided to be executed.

[0073] In the case of a first-time elevator user, data on "correlation" and "execution probability coefficient" has not yet been formed, so the elevator control to be executed may be determined based on the value entered in the "emotion value at the time of boarding" column shown in FIG. 4, for example. In this case, one elevator control corresponding to the highest emotion value may be selected, or, if there are no problems, multiple elevator controls corresponding to emotion values above a predetermined value may be selected. In this way, even for a first-time elevator user for whom no past data exists, optimal elevator control that takes into account the current emotion can be executed.

[0074] Furthermore, for first-time elevator users, the elevator control may also be performed by audio guidance from the speaker 46 in the car, such as a first-time greeting, elevator operation instructions, and building information. This can provide a sense of security to first-time elevator users who are unfamiliar with the elevator equipment and feel uneasy.

[0075] In the case of a returning elevator user, the elevator control to be executed is determined based on, for example, the value entered in the "Correlation" column shown in Fig. 5. In this case, one elevator control corresponding to the highest correlation value may be selected, or, if there are no problems, multiple elevator controls corresponding to correlation values equal to or greater than a predetermined value. In this way, it is possible to execute the optimal elevator control for a returning elevator user according to the emotional tendencies of the elevator user from the past to the present.

[0076] Furthermore, if the highest correlation value cannot be found from the values listed in the "Correlation" column, the elevator control to be executed may be determined by taking into account the value listed in the "Execution Probability Coefficient" column. In this case, the elevator control corresponding to the highest execution probability coefficient may be selected. Furthermore, the elevator control to be executed may be determined based on the value listed in the "Emotion Value When Boarding" column shown in FIG. 4. In this case, the elevator control corresponding to the highest emotion value may be selected, or, if there are no problems, multiple elevator controls corresponding to emotion values equal to or greater than a predetermined value may be selected. In this way, even if the highest correlation value cannot be found from the values listed in the "Correlation" column, it is possible to execute an elevator control suitable for the elevator user.

[0077] Control evaluation unit 22e The control evaluation unit 22e evaluates the effectiveness of the elevator control by determining whether the elevator control is successful or not based on the difference between the analysis result before the elevator control that the control decision unit 22d has decided to execute and the analysis result after the elevator control has been executed. The following two examples are given as examples of the evaluation method in this case.

[0078] (Evaluation Method 1) This evaluation method uses information on the "emotion value when getting on" and "emotion value when getting off" corresponding to the elevator control that has been decided to be executed, and their respective "emotion classifications."

[0079] That is, as mentioned above, when the emotion classification corresponding to the elevator control shown in FIG. 4 is "negative," if the "emotion value when getting off" is lower than the "emotion value when getting on" by a certain amount or more, the elevator control is determined to be successful; otherwise, the elevator control is determined to have failed.

[0080] For example, if the elevator control that has been decided to be executed is "feeling anxious," the emotion classification is "negative," and the "emotional value when getting off" (=30) is lower than the "emotional value when getting on" (=70) by a certain amount, so the elevator control is determined to have been successful.

[0081] On the other hand, if the "emotion classification" is "positive," and the "emotion value when getting off" is higher than the "emotion value when getting on" by a certain amount, the corresponding elevator control is determined to have been successful; otherwise, the corresponding elevator control is determined to have failed.

[0082] (Evaluation Method 2) This evaluation method uses not only the "emotion value when getting on," "emotion value when getting off," and each "emotion classification" information corresponding to the elevator control that has been decided to be executed, but also the "emotion value when getting on," "emotion value when getting off," and each "emotion classification" information corresponding to the various elevator controls shown in Fig. 4. In other words, the entire information shown in Fig. 4 is used.

[0083] First, for the entire information shown in Figure 4, a value (hereinafter referred to as the "first evaluation value") is calculated by subtracting the sum of the "emotion values at the time of riding" of the various emotions whose "emotion classification" indicates "negative" from the sum of the "emotion values at the time of riding" of the various emotions whose "emotion classification" indicates "positive."

[0084] Next, for the entire information shown in Figure 4, a value (hereinafter referred to as the "second evaluation value") is calculated by subtracting the sum of the "emotion values at the time of getting off" of the various emotions whose "emotion classification" indicates "negative" from the sum of the "emotion values at the time of getting off" of the various emotions whose "emotion classification" indicates "positive."

[0085] Finally, if the value obtained by subtracting the "second evaluation value" from the "first evaluation value" is a positive value greater than or equal to a predetermined value, it is determined that the elevator control was successful; otherwise, it is determined that the corresponding elevator control failed.

[0086] Evaluation Reflection Section 22F The evaluation reflecting unit 22f reflects the result of the control evaluation unit 22e's evaluation of the effectiveness of the elevator control (whether the elevator control was successful or not) in the management information shown in Fig. 5. In addition, this result may be provided to an AI model provided in the information management unit 22c so that the optimal control combination can be re-learned. By reflecting the evaluation result in the management information or feeding it back to the AI model in this way, it is possible to execute elevator control that is appropriate for the elevator user's emotions at any given time, thereby realizing a comfortable elevator environment for the elevator user.

[0087] Specifically, the higher the evaluation, the higher the correlation value corresponding to the corresponding elevator control in the management information is set by the evaluation reflecting unit 22f. For example, if the elevator control is successful, the evaluation reflecting unit 22f updates the correlation value by adding a predetermined value (for example, 1) to the value written in the "correlation value" column shown in FIG. 5. The evaluation reflecting unit 22f also updates the value written in the "execution probability coefficient" column as appropriate. The updated management information will be used again the next time the same elevator user returns. In other words, the management information becomes information customized for each elevator user.

[0088] <Example of operation> An example of the operation of the facial expression analysis system 22 will be described with reference to the flowchart of FIG.

[0089] The face detection unit 22a monitors the images captured by the cameras 16a to 16c installed for each elevator at the landing of each floor, and continues monitoring until it detects the face of an elevator user (passenger) about to board the elevator from the images (No in step S1).

[0090] When the face detection unit 22a detects the face of the elevator user when getting on the elevator (Yes in step S1), the facial expression analysis unit 22b analyzes the facial expression of the elevator user when getting on the elevator (step S2) and generates an analysis result in which various emotions are quantified from the expression (step S3).

[0091] On the other hand, the information management unit 22c manages management information indicating the correspondence between the various emotions of the elevator users and various elevator controls.

[0092] Based on the analysis results generated by the facial expression analysis unit 22b and the management information managed by the information management unit 22c, the control decision unit 22d selects, for example, the elevator control with the highest correlation value (a value indicating the degree of correlation between the corresponding emotion and the elevator control) as the elevator control to be executed for the elevator user (step S4).

[0093] The operation control unit 23 executes the elevator control selected by the control determination unit 22d (step S5).

[0094] Thereafter, face detection unit 22a monitors images of the inside of the elevator car taken by cameras 18a to 18c for each elevator car, and continues to track the face of the elevator user included in the images until the elevator user gets off the elevator (No in step S6).

[0095] When the face detection unit 22a detects the face of the elevator user when getting off the elevator (Yes in step S6), the facial expression analysis unit 22b analyzes the facial expression of the elevator user when getting off the elevator (step S7) and generates an analysis result in which various emotions are quantified from the expression (step S8).

[0096] When the control evaluation unit 22e evaluates the effectiveness of the elevator control based on the difference between the analysis result before the elevator control that has been decided to be executed and the analysis result after the elevator control is executed, the evaluation reflection unit 22f reflects the result of the evaluation of the effectiveness of the elevator control in the management information and updates the evaluation value, etc. (step S9).

[0097] The processes of steps S1 to S9 are performed for each elevator user.

[0098] As described above in detail, the elevator system of the embodiment can operate the elevator in consideration of the emotions of the elevator user. In particular, by realizing elevator control customized for each elevator user and individually executing elevator control suited to the emotions of each elevator user at that time, it is possible to individually realize an elevator environment that is comfortable and convenient for each elevator user. [Explanation of symbols]

[0099] 10...information terminal, 11a to 11c...elevator control device, 12a to 12c...car, 14...landing (elevator hall), 15...landing call registration device, 16a to 16c...camera (imaging device), 17a to 17c...landing door, 18a to 18c...camera (imaging device), 20...group management control device, 21...call memory unit, 22...facial expression analysis system, 22a...face detection unit, 22b...facial expression analysis unit, 22c...information management unit, 22d...control decision unit, 22e...control evaluation unit, 22f...evaluation reflection unit, 23...operation control unit, 40...car door, 41a, 41b...entrance pillar, 42a, 42b...door pocket, 43...display device, 44...destination floor button, 45...operation panel, 46...speaker, 47...lighting, 48...air conditioning equipment.

Claims

1. a face detection unit that detects the face of an elevator user captured in an image captured by a camera installed at the elevator hall or inside the elevator car; an expression analysis unit that analyzes facial expressions of the elevator user to generate an analysis result that quantifies various emotions of the elevator user; an information management unit that manages management information indicating a correspondence between various emotions of the elevator user and various elevator controls; a control determination unit that determines elevator control to be performed for the elevator user based on the analysis result and the management information; An elevator system comprising:

2. The facial expression analysis unit Analyzing the facial expressions of the elevator users using AI (Artificial Intelligence); 10. The elevator system of claim 1.

3. The results of the analysis are as follows: The information includes information indicating various emotions of the elevator user, information indicating the degree of each of the various emotions, and information indicating whether each of the various emotions is positive or negative.

10. The elevator system of claim 1.

4. The management information is The information includes information indicating various emotions of the elevator user, information indicating various elevator controls corresponding to the various emotions, and information on correlation values indicating the degree of correlation between the various elevator controls and the emotions corresponding to the various emotions.

10. The elevator system of claim 1.

5. The control decision unit determining an elevator control associated with the highest correlation value as an elevator control to be executed for the elevator user; 5. The elevator system of claim 4.

6. The results of the analysis are as follows: The analysis result includes not only the analysis result before the execution of the elevator control that has been decided to be executed, but also the analysis result after the execution of the elevator control.

6. The elevator system of claim 5.

7. a control evaluation unit that evaluates effectiveness of the elevator control based on a difference between an analysis result before the elevator control is executed and an analysis result after the elevator control is executed; an evaluation reflecting unit that increases a correlation value corresponding to a corresponding elevator control in the management information as the evaluation is higher; 7. The elevator system of claim 6, further comprising:

8. At least one of the various elevator controls in the management information includes at least one of elevator control for displaying and outputting predetermined information on a display device in the car, elevator control for outputting predetermined information by voice from a speaker in the car, and elevator control for changing the brightness or color tone of lighting in the car.

10. The elevator system of claim 1.

9. At least one of the various elevator controls in the management information includes an elevator control for adjusting air conditioning in a car.

10. The elevator system of claim 1.

10. a face detection unit detecting a face of an elevator user captured in an image captured by a camera installed at the elevator hall or inside the elevator car; generating an analysis result in which various emotions of the elevator user are quantified by analyzing the facial expressions of the elevator user using an expression analysis unit; managing, by an information management unit, management information indicating a correspondence relationship between various emotions of the elevator user and various elevator controls; determining, by a control determination unit, elevator control to be performed for the elevator user based on the analysis result and the management information; An elevator control method comprising:

Images