Office layout design support method and office layout design support apparatus

Abstract

To design a layout in consideration of each seating environment of a plurality of workers in an office.SOLUTION: An office layout design support method supports layout design in an office, and includes a standing-up step S4 of standing up an evaluation function of privacy related to all workers and an evaluation function of proximity to the surrounding workers related to all the workers by using, as variables, layout variables including one or both of the arrangement of each of a plurality of desks on which each of a plurality of workers is seated to a plurality of candidate positions and the height of a partition provided for each of the plurality of desks, and a layout determining step S5 of determining each value of the layout variables for each of the plurality of desks by applying an evolutionary calculation method to the standing-up objective function.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to an office layout design support method and an office layout design support device. [Background technology]

[0002] For example, when opening a new office, it is necessary to decide on a layout for arranging desks, cabinets, etc. Or, when a company that has moved into an office undergoes reorganization, the office layout may need to be changed. In such cases, support for office layout design is being provided with the aim of making the office layout more appropriate and quickly deciding on the layout.

[0003] For example, Patent Document 1 discloses a layout design support system that includes a target range determination unit that determines a target range within a floor in accordance with information that defines the shape of the floor for arranging fixtures, a region allocation unit that allocates multiple rectangular regions to the target range in accordance with arrangement rules for each region to be arranged within the target range, a function allocation unit that calculates evaluation values ​​for one or more predetermined evaluation items from the allocation results of the multiple regions and determines the allocation of functions to the multiple regions based on the evaluation values, and an arrangement processing unit that arranges fixtures within the target range in accordance with predetermined arrangement rules. The arrangement processing unit arranges islands containing, for example, multiple desks for all of the target ranges that have been divided into several parts in accordance with the predetermined arrangement rules.

[0004] However, if an office worker feels that they have little privacy, they may be unable to concentrate fully on their work, which could result in reduced productivity. For this reason, when designing an office layout, it is desirable to take into consideration the seating environment for each of the multiple workers, including, for example, privacy. In this regard, in the layout design support system of Patent Document 1, the arrangement of each desk is carried out simply according to a predetermined arrangement rule, and therefore, in Patent Document 1, the desks may be arranged without considering the seating environment of each worker. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2021-157515 Summary of the Invention [Problem to be solved by the invention]

[0006] The problem to be solved by the present invention is to provide an office layout design support method and an office layout design support device that can design a layout in an office taking into account the seating environments of each of a plurality of workers. [Means for solving the problem]

[0007] The present invention employs the following means to solve the above-mentioned problems. Specifically, the present invention provides an office layout design support method for supporting layout design in an office, comprising: a formulation step of formulating a privacy evaluation function for all of the workers and an evaluation function of proximity for all of the workers to surrounding workers, using layout variables including one or both of the arrangement of each of a plurality of desks at which a plurality of workers are seated at a plurality of candidate positions and the height of a partition to be provided for each of the plurality of desks, and a layout determination step of applying an evolutionary computation method to an objective function formulated based on the privacy evaluation function and the proximity evaluation function, thereby determining the value of each of the layout variables for each of the plurality of desks. According to the above-described configuration, in the formulation step, a privacy evaluation function for all workers and an evaluation function for proximity to surrounding workers for all workers are formulated using layout variables including either or both of the arrangement of each of the multiple desks at which each of the multiple workers is seated in a plurality of candidate positions and the height of the partitions to be provided for each of the multiple desks. Then, in the layout determination step, an evolutionary computation method is applied to an objective function formulated based on the thus formulated privacy evaluation function and proximity evaluation function. As described above, the privacy evaluation function and proximity evaluation function constituting the objective function use, as variables, layout variables including either or both of the arrangement of each of the multiple desks at which each of the multiple workers is seated in a plurality of candidate positions and the height of the partitions to be provided for each of the multiple desks. Therefore, by applying an evolutionary computation method to the objective function and finding a solution that optimizes the objective function, it is possible to determine the values ​​of each layout variable, that is, the desk layout if the layout variables include desk layout, the partition height if the layout variables include partition height, and both the desk layout and partition height if the layout variables include both, so that both privacy and proximity are appropriate for all workers. In this way, it becomes possible to design an office layout that takes into account the seating environments of each of a plurality of workers.

[0008] In one aspect of the present invention, the layout variables include both the placement of each of the multiple desks at the multiple candidate positions and the height of the partitions, the objective functions include a first objective function based on the privacy evaluation function and a second objective function based on the proximity evaluation function, and in the layout determination step, the values ​​of each of the layout variables are determined by applying an evolutionary computing method to the first objective function and the second objective function to solve a multi-objective optimization problem. According to the above configuration, objective functions (first objective function and second objective function) are provided for each of the privacy evaluation function and the proximity evaluation function, and the values ​​of each layout variable are determined by applying an evolutionary computing method to each of these to solve a multi-objective optimization problem. In this way, since there are multiple objective functions to which the evolutionary computing method is applied, multiple solutions can be obtained as Pareto solutions. Since an appropriate solution can be selected as needed from the multiple solutions obtained in this way, the degree of freedom in layout design can be improved.

[0009] In another aspect of the present invention, a structural equation modeling step applies structural equation modeling to results of a questionnaire about the working environment to extract a plurality of factors including privacy and proximity, derives an overall effect about privacy and an overall effect about proximity for each of a plurality of factors other than privacy and proximity, and derives a factor score about privacy and a factor score about proximity for each of respondents to the questionnaire; and based on the results of a survey of each of the plurality of respondents about an isovist area, which is the area of ​​an area visible to the respondent within a circle extending in a horizontal plane centered on the viewpoint of the respondent when seated at his / her desk, for each of a plurality of radii and a plurality of eye heights, calculates a correlation between the factor score about privacy and the survey results for each of the plurality of respondents, and the correlation between the factor score about privacy and the survey results for each of the plurality of respondents, and the correlation between the factor score about proximity and the survey results for each of the plurality of respondents. and a correlation analysis step of performing a correlation analysis between each of the factor scores related to privacy and the results of the survey, and deriving a privacy correlation coefficient which is the correlation coefficient between the isovist area and privacy, and a proximity correlation coefficient which is the correlation coefficient between the isovist area and proximity, for each of the plurality of radii and the plurality of eye level heights, wherein in the equation formulation step, a combination of the radii and the eye level heights for which the absolute value of the privacy correlation coefficient is large is extracted, and an evaluation function for privacy is formulated using the radius and the isovist area at the eye level for the combination as variables, and a combination of the radii and the eye level heights for which the absolute value of the proximity correlation coefficient is large is extracted, and an evaluation function for proximity is formulated using the radius and the isovist area at the eye level for the combination as variables. According to the above configuration, structural equation modeling is applied to the results of a survey on the workplace environment to derive a privacy factor score and a proximity factor score for each of the survey respondents. Furthermore, based on the results of a survey of each of the respondents for each of a plurality of radii and a plurality of eye heights, the isovist area, which is the area of ​​the area visible to the respondent within a circle extending in a horizontal plane centered on the respondent's viewpoint when seated at their own desk, is analyzed. A correlation analysis is then performed between the privacy factor score for each of the respondents and the survey results to derive a privacy correlation coefficient, which is the correlation coefficient between the isovist area and privacy for each of the plurality of radii and a plurality of eye heights. The absolute value of the privacy correlation coefficient thus derived increases as the correlation between the radius and eye height at which the isovist area was obtained and privacy increases. Therefore, by extracting a combination of radius and eye level height with a large absolute value of the privacy correlation coefficient and formulating a privacy evaluation function using the isovist area at the radius and eye level height of that combination as a variable, it is possible to realize a privacy evaluation function while suppressing a decrease in accuracy while narrowing down the number of variables used, thereby enabling multi-objective optimization problems to be solved efficiently. Similarly, a correlation analysis is performed between the proximity factor scores for each of the multiple respondents and the survey results to derive a proximity correlation coefficient, which is the correlation coefficient between the isovist area and proximity for each of the multiple radii and multiple eye heights. The absolute value of the proximity correlation coefficient derived in this manner increases as the correlation between the radius and eye height at which the isovist area is obtained and the proximity is strengthened. Therefore, by extracting a combination of radius and eye height with a large absolute value of the proximity correlation coefficient and formulating a proximity evaluation function using the isovist area at the radius and eye height of that combination as a variable, it is possible to realize a proximity evaluation function while suppressing a decrease in accuracy while narrowing the number of variables used. This allows for efficient solution of multi-objective optimization problems.

[0010] In another aspect of the present invention, in the layout determination step, the overall effect on privacy calculated for one factor selected from the multiple factors is set as a first coefficient, the overall effect on proximity calculated for the one factor is set as a second coefficient, and an overall evaluation function is constructed by adding a privacy term obtained by multiplying the first coefficient and the privacy evaluation function and a proximity term obtained by multiplying the second coefficient and the proximity evaluation function, and among multiple solutions obtained by solving the multi-objective optimization problem, the solution with the largest overall evaluation function is proposed as the layout result. According to the above configuration, in the structural equation modeling step, structural equation modeling is applied to extract multiple factors including privacy and proximity, and an overall effect on privacy and an overall effect on proximity are derived for each of the multiple factors other than privacy and proximity. The overall effect on privacy for each of the multiple factors derived in this way represents the strength of the influence that privacy has on that factor. Similarly, the overall effect on proximity for each of the multiple factors represents the strength of the influence that proximity has on that factor. Here, the overall evaluation function is constructed by adding a privacy term obtained by multiplying the first coefficient, which is the overall effect on privacy calculated for one factor selected from among multiple factors, and a proximity term obtained by multiplying the second coefficient, which is the overall effect on proximity calculated for the one factor, and a proximity term obtained by multiplying the second coefficient, which is the evaluation function for proximity. That is, the first coefficient and the second coefficient represent the strength of the influence of privacy and the strength of the influence of proximity for the one factor, respectively. Therefore, by multiplying the first coefficient and the second coefficient by the privacy evaluation function and the proximity evaluation function, respectively, the first coefficient and the second coefficient act as weights for the privacy evaluation function and the proximity evaluation function. In this way, the evaluation function can be appropriately expressed.

[0011] In another aspect of the present invention, the objective function is constructed by adding a privacy term obtained by multiplying a first coefficient and the privacy evaluation function and a proximity term obtained by multiplying a second coefficient and the proximity evaluation function, and in the layout determination step, an evolutionary computation method is applied to the objective function to solve a single-objective optimization problem, thereby determining the values ​​of each of the layout variables. According to the above configuration, the office layout design support method can be appropriately implemented.

[0012] The present invention also provides an office layout design support device that supports layout design within an office, comprising: an equation formulation unit that formulates a privacy evaluation function for all of the workers and an evaluation function for proximity to surrounding workers for all of the workers, using layout variables including either or both of the arrangement of each of a plurality of desks at which each of a plurality of workers is seated at a plurality of candidate positions and the height of a partition to be provided for each of the plurality of desks as variables; and a layout determination unit that determines the value of each of the layout variables for each of the plurality of desks by applying an evolutionary computation method to an objective function formulated based on the privacy evaluation function and the proximity evaluation function. According to the above-described configuration, similar to the explanation of the office layout design support method, it becomes possible to design a layout in an office taking into consideration the seating environments of each of a plurality of workers. [Effects of the Invention]

[0013] According to the present invention, it is possible to provide an office layout design support method and an office layout design support device that can perform layout design in an office taking into consideration the seating environments of each of a plurality of workers. [Brief explanation of the drawings]

[0014] [Figure 1] FIG. 10 is a diagram showing an example of candidate positions for placing desks. [Figure 2] 2 is a diagram showing an example of desk arrangement for each candidate position shown in FIG. 1. FIG. [Figure 3] 1 is a block diagram of an office layout design support device according to an embodiment of the present invention. [Figure 4] 10 is a flowchart of an office layout design support method using the office layout design support device. [Figure 5] FIG. 10 is an explanatory diagram regarding an isovist area. [Figure 6] FIG. 1 is an explanatory diagram of the isovist area for each of seven radii. [Figure 7] This is an explanatory diagram of the isovist area for each of three types of eye height. [Figure 8] FIG. 10 is a diagram showing an example of the results of a preliminary factor analysis. [Figure 9] FIG. 10 is a diagram showing an example of the results of structural equation modeling. [Figure 10] FIG. 10 is a diagram showing an example of the total effect calculated as a result of structural equation modeling. [Figure 11] FIG. 10 is a diagram showing examples of 21 types of isovist area, factor scores related to privacy, and factor scores related to proximity for each of the respondents to the questionnaire. [Figure 12] FIG. 10 is a diagram showing an example of a privacy correlation coefficient derived as a result of a correlation analysis between the isovist area and the factor score related to privacy at each of a plurality of radii and a plurality of eye-level heights. [Figure 13] FIG. 10 is a diagram showing an example of a proximity correlation coefficient derived as a result of a correlation analysis between isovist area and factor scores related to proximity at each of a plurality of radii and a plurality of eye-level heights. [Figure 14] FIG. 10 is a diagram showing an example of a Pareto solution obtained by solving a multi-objective optimization problem for a first objective function based on an evaluation function for privacy and a second objective function based on an evaluation function for proximity. [Figure 15] 15A and 15B are diagrams illustrating examples of layouts corresponding to each of the Pareto solutions shown in FIG. 14. [Figure 16] FIG. 10 illustrates an example of an optimal single solution selected for each of multiple factors. [Figure 17] FIG. 10 is a diagram showing an example of multiple solutions obtained by widening the range from which a solution is selected for each of multiple factors. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. First, an example used in explaining this embodiment will be described. It goes without saying that this embodiment can be applied to various examples other than the example described below. Fig. 1 is a diagram showing an example of candidate positions where desks are to be placed, and Fig. 2 is a diagram showing an example of the placement of desks at each candidate position shown in Fig. 1. In FIG. 1, a plurality of, for example, 15, candidate positions CP are provided in an office. Two trapezoidal pieces of furniture DA are provided at each of these candidate positions CP to form a desk D at which one worker can sit. In FIG. 2, six types of furniture DA arrangement patterns for the candidate positions CP are depicted as desks D1 to D6. One of these desks D1 to D6 is selected and placed as the desk D for each candidate position CP. That is, in the example of this embodiment, the type of desk D for each candidate position CP is determined from among the desks D1 to D6. In Fig. 2, an arrow is drawn on each of the desks D1 to D6. The starting point of the arrow is the seating position of the worker corresponding to desk D. The direction of the arrow is the line of sight of the worker.

[0016] A partition PA is provided at each desk D placed at each candidate position CP. The partition PA is provided along the edge of the fixture DA so as to surround the direction in which the worker's line of sight is directed from the outside. In FIG. 2, the partition PA is only drawn on the desk D1 shown in the upper left, but similar partitions PA are provided on the other desks D2 to D6. In the example of this embodiment, the height of the partition PA to be installed for the desk D positioned at each candidate position CP is selected from three types: 720 mm, which is the same height as the furniture DA (i.e., equivalent to when no partition PA is installed), 1150 mm, and 1550 mm.

[0017] In this way, the office layout design support device and office layout design support method of this embodiment determine both the placement of each of the multiple desks D, at which each of the multiple workers is seated, at the multiple candidate positions CP, and the height of the partitions PA to be provided for each of the multiple desks D. In this way, the office layout design support device and office layout design support method of this embodiment support the layout design within an office. As will be described later as a modified example, the office layout design support device and office layout design support method may be configured to determine only the placement of each of a plurality of desks D at a plurality of candidate positions CP, where each of a plurality of workers is seated. Alternatively, the device and method may be configured to determine only the height of the partitions PA to be provided for each of the plurality of desks D, assuming that each of the plurality of desks D at which a plurality of workers is seated has been placed at a plurality of candidate positions CP. Although FIG. 1 shows, for example, 15 candidate positions CP, if the number of desks D is less than the number of candidate positions CP, it is not necessary to place desks D at all of these candidate positions CP, and it is acceptable for there to be candidate positions CP at which no desks D are placed.

[0018] FIG. 3 is a block diagram of the office layout design support device. The office layout design support device 1 is comprised of a computer terminal such as a server, a personal computer, or a tablet terminal, and performs required functions by executing a preset program. The office layout design support device 1 functionally comprises a pre-factor analysis unit 11, a structural equation modeling unit 12, a correlation analysis unit 13, a formulating unit 14, and a layout determination unit 15.

[0019] The pre-factor analysis unit 11 performs a pre-factor analysis on the results of a questionnaire regarding the working environment. The structural equation modeling unit 12 applies structural equation modeling to the results of the questionnaire regarding the work environment to extract multiple factors including privacy and proximity, derives an overall effect regarding privacy and an overall effect regarding proximity for each of the multiple factors other than privacy and proximity, and derives a factor score regarding privacy and a factor score regarding proximity for each of the respondents to the questionnaire. The correlation analysis unit 13 performs a correlation analysis between the privacy factor score for each of the multiple respondents and the survey results, and between the proximity factor score for each of the multiple respondents and the survey results, based on the results of a survey of the isovist area, which is the area of ​​the area visible to the respondent within a circle extending in a horizontal plane centered on the viewpoint of the respondent when seated at their own seat, for each of multiple radii and multiple eye heights, and derives a privacy correlation coefficient, which is the correlation coefficient between the isovist area and privacy, and a proximity correlation coefficient, which is the correlation coefficient between the isovist area and proximity, for each of the multiple radii and multiple eye heights.

[0020] The formulating unit 14 formulates an evaluation function for privacy for all workers and an evaluation function for proximity to surrounding workers for all workers, using layout variables including both the placement of each of the multiple desks D at which each of the multiple workers is seated at multiple candidate positions and the height of the partitions to be provided for each of the multiple desks as variables. The formulating unit 14 extracts combinations of radius and eye level height for which the absolute value of the privacy correlation coefficient is large, and formulates an evaluation function for privacy using the isovist area at the radius and eye level height of the combination as a variable, and also extracts combinations of radius and eye level height for which the absolute value of the proximity correlation coefficient is large, and formulates an evaluation function for proximity using the isovist area at the radius and eye level height of the combination as a variable.

[0021] The layout determination unit 15 determines the values ​​of each of the layout variables for each of the multiple desks D by applying an evolutionary computation method to an objective function formulated based on the privacy evaluation function and the proximity evaluation function. The above objective function includes a first objective function based on a privacy evaluation function and a second objective function based on a proximity evaluation function, and the layout determination unit 15 determines the values ​​of each layout variable by applying an evolutionary computation method to the first objective function and the second objective function to solve a multi-objective optimization problem. The layout determination unit 15 also constructs an evaluation function by adding a privacy term obtained by multiplying the first coefficient and the privacy evaluation function and a proximity term obtained by multiplying the second coefficient and the proximity evaluation function, with the first coefficient being the total effect on privacy calculated for one factor selected from the multiple factors, and the second coefficient being the total effect on proximity calculated for the one factor, and proposes the solution with the largest evaluation function as the layout result from among multiple solutions obtained by solving the multi-objective optimization problem.

[0022] FIG. 4 is a flowchart of an office layout design support method using the office layout design support device. Hereinafter, the operation of each component of the office layout design support device 1 shown in FIG. 3 will be described by explaining the office layout design support method in detail.

[0023] The office layout design support device 1 and the office layout design support method in this embodiment determine the layout by taking into consideration privacy, which is an index indicating whether each worker is free from interference or intrusion from surrounding workers at their own desk, and proximity, which indicates how close the relationship between the worker and surrounding workers feels at their own desk. To do this, it is necessary to understand what other factors the indicators (factors) indicated as privacy and proximity affect. For this purpose, before the office layout design support method shown in FIG. 4 is executed, a questionnaire regarding the workplace environment is conducted.

[0024] Here, other factors that may be considered besides privacy and proximity include, for example, group cohesion, which indicates the strength of attraction of the group to which an employee belongs, family culture, which is the culture within the group, productivity at the employee's desk, satisfaction at the employee's desk, and job satisfaction, which is satisfaction with the job. It is desirable for the above questionnaire to have questions prepared for each factor that may be considered at the time of conducting the questionnaire, including privacy and proximity. For example, regarding privacy, possible questions include, "I often feel the gaze of others at my desk," and "It is almost impossible to do anything at my desk without being noticed by others." Regarding proximity, possible questions include, "I can see my colleagues well from my desk," and "I can see my boss well from my desk." In this way, multiple factors that may be related to office layout, including privacy and proximity, are assumed, and questionnaire questions are prepared for each of these possible factors. The results of the questionnaire, that is, the answers given by the respondents, are stored in a database or the like (not shown) of the office layout design support device 1.

[0025] In addition, the isovist area of ​​each of the respondents to the questionnaire will be investigated. FIG. 5 is an explanatory diagram regarding the isovist area. The isovist area is the area of ​​the region visible to the respondent within a circle C extending in a horizontal plane centered on the viewpoint SP when the respondent is seated at his / her seat. For example, in Figure 5, within the circle C with a radius of R mm centered on the respondent's viewpoint SP, the range that the respondent can see is displayed with a pattern. The area of ​​this part is the isovist area.

[0026] Figure 6 is an explanatory diagram of the isovist area for each of seven types of radii, and Figure 7 is an explanatory diagram of the isovist area for each of three types of eye heights. In this embodiment, the isovist area was surveyed for each respondent for all 21 combinations of the seven types of radii shown in Figure 6 and the three types of eye level heights corresponding to the three types of partition PA heights. The results of the isovist area survey for each respondent are stored in a database or the like (not shown) of the office layout design support device 1, similar to the results of the questionnaire.

[0027] The above-mentioned questionnaire and isovist area survey may be conducted on each worker for whom the office layout is actually being carried out, but they may also be conducted on other personnel. In other words, the person responding to the questionnaire may be different from the worker who sits at the desk D that is being arranged. In this case, it is desirable to select respondents so that the working environment of the worker and the working environment of the respondent are as similar as possible, and so that the results of the questionnaire and isovist area survey conducted on the respondent are close to the results that would be expected if the questionnaire and isovist area survey were conducted on workers.

[0028] In the office layout design support method, once the above data has been collected, the pre-factor analysis unit 11 performs a factor analysis on the questionnaire results (pre-factor analysis step S1). In the questionnaire exemplified in this embodiment, questions related to equality are also included in the evaluation of the seating environment, based on the idea that equality with other workers can also be a factor in addition to the privacy and proximity mentioned above. Figure 8 shows an example of the results of a factor analysis of the results of the questions on privacy, proximity, and equality in this questionnaire. Figure 8 shows three factors, Factor 1, Factor 2, and Factor 3, which are extracted and are shown to be related to the previously assumed factors of proximity, privacy, and equality, respectively. In this way, the pre-factor analysis unit 11 performs a pre-factor analysis on the results of the questionnaire, thereby confirming whether or not the factors assumed when the questionnaire was created can actually be factors.

[0029] Next, the structural equation modeling unit 12 applies structural equation modeling to the results of the questionnaire regarding the working environment (structural equation modeling step S2). When structural equation modeling is performed, each of the factors confirmed to be potential factors in the preliminary factor analysis unit 11 is set as a factor by the operator. In addition, relationships that are thought to influence each other among these factors are set as arrows between the factors. After these settings are completed, structural equation modeling is performed by the structural equation modeling unit 12. FIG. 9 is a diagram showing an example of the results of applying structural equation modeling to the results of a questionnaire about the working environment. Figure 9 shows the two factors of privacy and proximity, as well as the factors of "group cohesion / family culture," "productivity at one's desk," "satisfaction at one's desk," and "job satisfaction." In this figure, of the paths (arrows) connecting factors, paths that are not significant at the 5% level have been deleted. Each path is accompanied by an unstandardized estimate of the path coefficient. In this example, the factor related to "equality" was deleted because no influence on other factors was observed.

[0030] Figure 10 is a diagram showing an example of the total effect calculated as a result of structural equation modeling. In Figure 10, four factors other than privacy and proximity of the six factors depicted in Figure 9 are listed as "dependent variables." In addition, the direct effect, indirect effect, and total effect of privacy and proximity for each of these four factors are listed. The direct effect is the degree to which privacy and proximity directly affect each factor listed as the dependent variable (path coefficient). The indirect effect is the degree to which privacy and proximity indirectly affect each factor listed as the dependent variable through other factors (multiplication of path coefficients). The total effect is the sum of the direct and indirect effects. In this way, the total effect of privacy calculated for each of the multiple factors other than privacy and proximity represents the strength of the influence of privacy on that factor. Also, the total effect of proximity calculated for each of the multiple factors other than privacy and proximity represents the strength of the influence of proximity on that factor.

[0031] In this way, the structural equation modeling unit 12 applies structural equation modeling to the results of the questionnaire regarding the work environment to extract multiple factors including privacy and proximity, and derives the overall effect regarding privacy and the overall effect regarding proximity for each of the multiple factors other than privacy and proximity.

[0032] When the structural equation modeling unit 12 executes the structural equation modeling in the above manner, a factor score related to privacy and a factor score related to proximity are derived for each of the respondents to the questionnaire. FIG. 11 is a diagram showing examples of 21 types of isovist area, factor scores related to privacy, and factor scores related to proximity for each of the respondents to the questionnaire. In Fig. 11, the identification number of each respondent is written as "id." Corresponding to each respondent, the factor score for privacy and the factor score for proximity are written as "pri_2" and "pro_2."

[0033] Figure 11 also shows the isovist area surveyed for each respondent for all 21 combinations of the seven radii shown in Figure 6 and the three eye level heights corresponding to the three partition PA heights shown in Figure 7. In this embodiment, strictly speaking, the isovist area is calculated by dividing the actual isovist area by the area of ​​circle C shown in FIG. 5, and the ratio of the isovist area to the area of ​​circle C is used.

[0034] Next, the correlation analysis unit 13 performs a correlation analysis between the factor scores and the isovist area as shown in FIG. 11 (correlation analysis step S3). The correlation analysis unit 13 first performs a correlation analysis between the isovist area and the factor score related to privacy for each combination of multiple radii and multiple eye level heights. Based on the results of a survey of multiple respondents on the isovist area for each of multiple radii and multiple eye level heights, as shown as 21 types of values ​​in Fig. 11, the correlation analysis unit 13 performs a correlation analysis between the factor score related to privacy for each of the multiple respondents and the survey results, and derives a privacy correlation coefficient, which is the correlation coefficient between the isovist area and privacy for each of the multiple radii and multiple eye level heights.

[0035] FIG. 12 is a diagram showing an example of a privacy correlation coefficient derived as a result of a correlation analysis between the isovist area and the factor score related to privacy for each of a plurality of radii and a plurality of eye-level heights. In this embodiment, the questionnaire included five questions regarding privacy. Figure 12 shows the results of using factor scores for all five questions as a five-variable model. Furthermore, the results of using only the three questions with the highest factor loadings in structural equation modeling are shown as a three-variable model. In Figure 12, the highest negative correlation can be seen at a radius of 3000 mm for eye heights of 720 mm and 1150 mm in both the five-variable model and the three-variable model. In comparison, the correlation is relatively low at an eye height of 1550 mm.

[0036] Furthermore, the correlation analysis unit 13 performs a correlation analysis between the isovist area and the factor score related to proximity for each combination of multiple radii and multiple eye heights. Based on the results of a survey of multiple respondents on the isovist area for each of multiple radii and multiple eye heights, as shown as 21 types of values ​​in Fig. 11, the correlation analysis unit 13 performs a correlation analysis between the factor score related to proximity for each of the multiple respondents and the survey results, and derives a proximity correlation coefficient, which is the correlation coefficient between the isovist area and proximity for each of the multiple radii and multiple eye heights.

[0037] Fig. 13 shows an example of a proximity correlation coefficient derived as a result of a correlation analysis between the isovist area and the factor scores related to proximity for each of a plurality of radii and a plurality of eye level positions. In Fig. 13, a six-variable model and a three-variable model are shown, as in Fig. 12. In the example of this embodiment, as shown in FIG. 13, the highest positive correlation was confirmed for both the combination of an eye-level height of 720 mm and a radius of 5000 mm, and the combination of an eye-level height of 1150 mm and a radius of 6000 mm.

[0038] Based on the above results, the formulating unit 14 formulates an evaluation function for privacy for all workers and an evaluation function for proximity to surrounding workers for all workers, using layout variables including either or both of the placement of each of the multiple desks D at which each of the multiple workers is seated at multiple candidate positions CP and the height of the partitions PA to be provided for each of the multiple desks D as variables (formulation step S4). First, the layout variables will be described. The layout variables in this embodiment include a first layout variable and a second layout variable. The number of candidate positions CP is n (15 in the example of FIG. 1), the set of shape and position types of desks D is X (a set of 6 elements in the example of FIG. 2), and the set of heights of partitions PA is Y (for example, a set of 3 elements, such as 720 mm, 1150 mm, and 1550 mm). Also, the shape and position type of desk D to be installed at the i-th candidate position CP as an element of X is x i , Y, the height of the partition PA to be provided at the i-th candidate position CP is y i In this case, the layout state is expressed as the following equation (1) using a first layout variable φ.

number

[0039] From the privacy correlation coefficient shown in Figure 12, it can be confirmed that the factor score related to privacy has a negative correlation with the isovist area. Also, from the proximity correlation coefficient shown in Figure 13, it can be confirmed that the factor score related to proximity has a positive correlation with the isovist area. From this, if we assume a linear relationship between each of privacy and proximity and some variable related to the isovist area, the privacy E at the desk D installed at the i-th candidate position CP can be calculated as follows: prv i and proximity E prx i can be expressed as the following equations (2) and (3), respectively.

number

number

[0040] In this case, the average value of privacy and proximity per seat is privacy E prv i and proximity E prx i Using these, it can be expressed as the following equations (4) and (5).

number

number

[0041] As already explained, there is a correlation between the isovist area and the factor score related to privacy. In addition, there is a correlation between the isovist area and the factor score related to proximity. By utilizing these relationships, the privacy E of the desk D installed at the i-th candidate position CP can be calculated by multiple regression analysis. prv i and proximity E prx i However, since the variables related to the isovist area are expressed as a combination of, for example, eye level height and radius, there are many of them, and even if multiple regression analysis is applied, it is difficult to formulate an equation appropriately and efficiently. Therefore, prior to multiple regression analysis, the variables related to the isovist area to be used are selected.

[0042] For example, with regard to privacy, first, from Figure 12, select the isovist area with the radius that has the highest correlation for each eye height, i.e., the largest absolute value of the privacy correlation coefficient. For example, the isovist area I with a radius of 3000 mm at an eye height of 720 mm is selected. 720、3000 i and an isovist area I with a radius of 3000 mm at an eye level of 1150 mm. 1150、3000 i and an isovist area I with a radius of 6000 mm at an eye level of 1550 mm. 1550、6000 i and can be selected. Then, a correlation analysis is performed between these to create a composite variable between highly correlated variables. This process makes it possible to avoid inappropriate analysis such as multicollinearity. In the case of using the example of this embodiment, the composite variable related to privacy can be expressed by the following formula (6). In formula (6), each variable I 720、3000 i , I 1150、3000 i In order to align the scale, the isovist area I is averaged by subtracting the mean and dividing by the standard deviation. 1550、6000 i was not used as a variable due to its low correlation with other variables.

number

number

[0043] The composite variable and isovist area I as expressed in the above equation (6) 1550、6000 i By performing a stepwise multiple regression analysis using the above formula, the privacy E of the desk D at the i-th candidate position CP, defined as formula (2), is calculated. prv iIn the example of this embodiment, it is expressed as the following equation (8). As a result of variable selection by the stepwise method, the isovist area I 1550、6000 i was excluded from the variables.

number

number

[0044] In this way, in the formulating step S4, the formulating unit 14 extracts a combination of radius and eye level having a large absolute value of the privacy correlation coefficient, and formulates the privacy evaluation function Privacy(φ, ω) expressed as equation (4) using the isovist area at the radius and eye level of the combination as a variable. In the example of this embodiment, the privacy E prv i In practice, equation (8) is used as In addition, in the formulating step S4, the formulating unit 14 extracts a combination of radius and eye level having a large absolute value of the proximity correlation coefficient, and formulates the proximity evaluation function Proximity(φ, ω) expressed as equation (5) using the isovist area at the radius and eye level of the combination as a variable. In the example of this embodiment, the proximity E in equation (5) prx i In practice, equation (9) is used as

[0045] The formulating unit 14 further formulates an objective function based on the privacy evaluation function Privacy(φ,ω) and the proximity evaluation function Proximity(φ,ω). In this embodiment, in a layout determination step S5 described later, the layout determination unit 15 determines the values ​​of the layout variables φ and ω by solving an optimization problem targeting this objective function. In this embodiment, the objective function includes a first objective function F1, which is an objective function for privacy regarding all workers, and a second objective function F2, which is an objective function for proximity regarding all workers. The first objective function F1 and the second objective function F2 are expressed as the following equations (10) and (11), respectively.

number

number

[0046] In equations (10) and (11), the second term subtracted from each of the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω) is a penalty term. ij is set to 1 when the desk D installed at the i-th candidate position CP and the desk D installed at the j-th candidate position CP overlap, and is set to 0 when they do not overlap. γ is a coefficient and can be set to a value such as 5. When solving the optimization problem to derive layout variables φ and ω that optimize the values ​​of the first objective function F1 and the second objective function F2, the derived solution requires that when desks D are placed at adjacent candidate positions CP, these desks D must not overlap in a plan view. This constraint is incorporated as a penalty into the first objective function F1 and the second objective function F2, so that the values ​​of the objective functions become smaller when the desks D overlap. This makes it less likely that a solution in which the desks D overlap will ultimately remain. Thus, in this embodiment, the objective function comprises a first objective function F1 based on the privacy evaluation function Privacy(φ, ω) for all workers, and a second objective function F2 based on the proximity evaluation function Proximity(φ, ω) for all workers.

[0047] Then, the layout determination unit 15 applies an evolutionary computation method to objective functions F1 and F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω), thereby determining the values ​​of the layout variables φ and ω for each of the multiple desks D (layout determination step S5). In the layout determination step S5, the values ​​of the layout variables φ and ω are determined by applying an evolutionary computation method to the first objective function F1 and the second objective function F2 to solve a multi-objective optimization problem. In this embodiment, a genetic algorithm is used as the evolutionary computation method, but other methods, such as differential evolution, may also be used as the evolutionary computation method.

[0048] In this embodiment, the values ​​of the layout variables φ and ω are determined by solving a multi-objective problem to optimize the multiple objective functions F1 and F2 as described above. Because there are multiple objective functions F1 and F2, the derived solutions are multiple Pareto solutions. Fig. 14 is a diagram showing an example of Pareto solutions obtained by solving a multi-objective optimization problem for a first objective function based on a privacy evaluation function and a second objective function based on a proximity evaluation function. Fig. 15 is a diagram showing an example of layouts corresponding to each of the Pareto solutions shown in Fig. 14. Fig. 15 shows layout results for a solution that places the most importance on proximity, labeled 1 in Fig. 14, a solution that places the most importance on privacy, labeled 7, and solutions between them, labeled 2 to 6. Figure 15 shows, for each layout result, the shapes of desk D and partition PA on the top, and the shape of isovist area, which is an explanatory variable for privacy and proximity, rendered and superimposed on the bottom. Figure 15 shows that as we move from solutions that prioritize proximity to solutions that prioritize privacy, there are more seats with high partition PA, and that in solutions that prioritize proximity, desk D tends to face outward from the center, while in solutions that prioritize privacy, desk D points in the same direction. Furthermore, because there is a positive correlation between isovist area and proximity and a negative correlation between isovist area and privacy, the superimposed rendering of the isovist area becomes lighter as we move from a layout that prioritizes proximity to a layout that prioritizes privacy.

[0049] Next, the layout determination unit 15 extracts appropriate layout results for each purpose from the Pareto solutions. In this embodiment, the layout determination unit 15 uses an overall evaluation function EF shown as the following equation (12) to extract an appropriate layout result.

number

[0050] For example, Figure 10 shows the factors other than privacy and proximity: "group cohesion / family culture," "productivity at one's desk," "satisfaction at one's desk," and "job satisfaction." If a worker selects a factor from these factors that he or she particularly values, the total effect of privacy calculated for the selected factor is the first coefficient α, and the total effect of proximity calculated for the selected factor is the second coefficient β. For example, if a worker considers "job satisfaction" to be a factor that he or she particularly values ​​in the layout results, the first coefficient α is 0.313, and the second coefficient β is 0.172. For a given factor, the total effect of privacy represents the strength of the influence of privacy on that factor, and the total effect of proximity represents the strength of the influence of proximity on that factor. Therefore, the overall effect on privacy and the overall evaluation on proximity are multiplied by the privacy evaluation function Privacy(φ,ω) and the proximity evaluation function Proximity(φ,ω) as the first coefficient α and the second coefficient β, respectively, and act as weights for the privacy evaluation function Privacy(φ,ω) and the proximity evaluation function Proximity(φ,ω).

[0051] Here, by transforming equation (12), the following equation (13) is obtained.

number

[0052] In this way, in the layout determination step S5, the overall effect on privacy calculated for one factor selected from the multiple factors is defined as the first coefficient α, the overall effect on proximity calculated for one factor is defined as the second coefficient β, and an overall evaluation function EF is constructed by adding a privacy term (α × Privacy(φ,ω)) obtained by multiplying the first coefficient α by the privacy evaluation function Privacy(φ,ω) and a proximity term (β × Proximity(φ,ω)) obtained by multiplying the second coefficient β by the proximity evaluation function Proximity(φ,ω), and the solution with the largest overall evaluation function EF among multiple solutions obtained by solving the multi-objective optimization problem is proposed as the layout result. This allows factors to be selected according to the purpose, thereby making it possible to obtain an appropriate layout result according to the purpose.

[0053] However, in the example used in this embodiment, as shown in FIG. 16, regardless of which factor is selected as the objective, "group cohesion / family culture," "productivity at own desk," "satisfaction at own desk," or "job satisfaction," the layout result that prioritizes privacy, which is located at the far right, is adopted. This is because, in this example, the Pareto solutions are aligned in a straight line on the graph. From the viewpoint of supporting layout design, it is preferable to display a wider variety of layout results. Therefore, in such a case, for example, the layout with the largest overall evaluation function EF, EF, is adopted. max In this case, the layout result in which the value of Equation (12) is 0.95×EF may be proposed together with a layout result in which the value is, for example, 5% lower. max That's all, EF max All of the following solutions may be proposed as layout results. 17 shows an example of multiple solutions obtained by widening the range of solution selection for each of multiple factors. In FIG. 17, the value of Eq. (12) is EF max The line L1 is the line where the value of equation (12) is 0.95 × EF maxIn such a case, the solution located between the line L1 and the line L2 can be a candidate to be proposed as a layout result.

[0054] The office layout design support method as described above is an office layout design support method that supports layout design within an office, and includes: a formulation step S4 in which a privacy evaluation function Privacy(φ, ω) for all workers and an evaluation function Proximity(φ, ω) for all workers' proximity to surrounding workers are formulated using layout variables φ and ω that include both the placement of each of a plurality of desks D, at a plurality of candidate positions CP, at which each of a plurality of workers is seated, and the height of the partition PA to be provided for each of the plurality of desks D; and a layout determination step S5 in which an evolutionary computation method is applied to objective functions F1 and F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω), thereby determining the values ​​of the layout variables φ and ω for each of the plurality of desks D. According to the above configuration, in formulation step S4, a privacy evaluation function Privacy(φ,ω) for all workers and an evaluation function Proximity(φ,ω) for proximity to surrounding workers are formulated using layout variables φ and ω, which include both the arrangement of each of a plurality of desks D, where each of a plurality of workers is seated, at a plurality of candidate positions CP, and the height of the partitions PA to be provided for each of the plurality of desks. Then, in layout determination step S5, an evolutionary computation method is applied to objective functions F1 and F2 formulated based on the thus formulated privacy evaluation function Privacy(φ,ω) and proximity evaluation function Proximity(φ,ω). As described above, the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω) that constitute the objective functions F1 and F2 use as variables the layout variables φ and ω that include both the arrangement of each of the multiple desks D, where each of the multiple workers is seated, at the multiple candidate positions CP, and the height of the partitions PA to be provided for each of the multiple desks D. Therefore, by applying an evolutionary computation method to the objective functions F1 and F2 to find a solution that optimizes the objective functions F1 and F2, it is possible to determine the values ​​of the layout variables φ and ω, i.e., both the arrangement of the desks D and the height of the partitions PA, so that both privacy and proximity are appropriate for all workers. In this way, it becomes possible to design an office layout that takes into account the seating environments of each of a plurality of workers.

[0055] Furthermore, the layout variables φ and ω include both the placement of each of the multiple desks D at multiple candidate positions CP and the height of the partition PA, and the objective functions F1 and F2 comprise a first objective function F1 based on a privacy evaluation function Privacy(φ, ω) and a second objective function F2 based on a proximity evaluation function Proximity(φ, ω), and in the layout determination step S5, the values ​​of each of the layout variables φ and ω are determined by applying an evolutionary computation method to the first objective function F1 and the second objective function F2 to solve a multi-objective optimization problem. According to the above configuration, objective functions (first objective function F1 and second objective function F2) are provided for each of the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω), and the values ​​of the layout variables φ and ω are determined by applying an evolutionary computation method to each of these to solve a multi-objective optimization problem. In this way, since there are multiple objective functions F1 and F2 to which the evolutionary computation method is applied, multiple solutions can be obtained as Pareto solutions. Since an appropriate solution can be selected as needed from the multiple solutions obtained in this way, the degree of freedom in layout design can be improved.

[0056] In addition, structural equation modeling is applied to the results of the questionnaire regarding the working environment to extract multiple factors including privacy and proximity, and an overall effect regarding privacy and an overall effect regarding proximity are derived for each of the multiple factors other than privacy and proximity, and a factor score regarding privacy and a factor score regarding proximity are derived for each of the respondents to the questionnaire. A structural equation modeling step S2 is also provided in which, based on the results of a survey of each of the multiple respondents regarding the isovist area, which is the area of ​​the area visible to the respondent within a circle C extending in a horizontal plane centered on the viewpoint of the respondent when seated at his / her own desk, for each of multiple radii and multiple eye heights, a correlation is established between the factor score regarding privacy and the survey results for each of the multiple respondents, and between the factor score regarding proximity and the survey results for each of the multiple respondents. and a correlation analysis step S3 in which a correlation analysis is performed for each of the multiple radii and multiple eye level heights to derive a privacy correlation coefficient, which is the correlation coefficient between the isovist area and privacy, and a proximity correlation coefficient, which is the correlation coefficient between the isovist area and proximity, for each of the multiple radii and multiple eye level heights. In the equation formulation step S4, a combination of radius and eye level height with a large absolute value of the privacy correlation coefficient is extracted, and an evaluation function for privacy, Privacy(φ, ω), is formulated using the isovist area at the radius and eye level of the combination as a variable, and a combination of radius and eye level with a large absolute value of the proximity correlation coefficient is extracted, and an evaluation function for proximity, Proximity(φ, ω), is formulated using the isovist area at the radius and eye level of the combination as a variable. According to the above configuration, structural equation modeling is applied to the results of a survey on the workplace environment to derive a privacy factor score and a proximity factor score for each of the survey respondents. Furthermore, for each of the multiple respondents, the isovist area, which is the area of ​​the area visible to the respondent within a circle C extending in a horizontal plane centered on the respondent's viewpoint when seated at their own desk, is surveyed for each of multiple radii and multiple eye heights. Based on the survey results, a correlation analysis is performed between the privacy factor score for each of the multiple respondents and the survey results to derive a privacy correlation coefficient, which is the correlation coefficient between the isovist area and privacy for each of the multiple radii and multiple eye heights. The absolute value of the privacy correlation coefficient thus derived increases as the correlation between the radius and eye height at which the isovist area was obtained and privacy increases. Therefore, by extracting a combination of radius and eye level height with a large absolute value of the privacy correlation coefficient and formulating the privacy evaluation function Privacy(φ,ω) using the isovist area at the radius and eye level height of that combination as a variable, it is possible to realize the privacy evaluation function Privacy(φ,ω) while suppressing a decrease in accuracy while narrowing down the number of variables used. This makes it possible to solve multi-objective optimization problems efficiently. Similarly, a correlation analysis is performed between the proximity factor scores for each of the multiple respondents and the survey results to derive a proximity correlation coefficient, which is the correlation coefficient between the isovist area and proximity for each of the multiple radii and multiple eye heights. The absolute value of the proximity correlation coefficient derived in this manner increases as the correlation between the radius and eye height at which the isovist area is obtained and the proximity is more strongly correlated. Therefore, by extracting a combination of radius and eye height with a large absolute value of the proximity correlation coefficient and formulating the proximity evaluation function Proximity(φ,ω) using the isovist area at the radius and eye height of that combination as a variable, the proximity evaluation function Proximity(φ,ω) can be realized while reducing the number of variables used and suppressing a decrease in accuracy. This allows for efficient solution of multi-objective optimization problems.

[0057] In addition, in the layout determination step S5, the overall effect on privacy calculated for one factor selected from the multiple factors is defined as a first coefficient α, the overall effect on proximity calculated for one factor is defined as a second coefficient β, and an overall evaluation function EF is constructed by adding a privacy term obtained by multiplying the first coefficient α by the privacy evaluation function Privacy(φ,ω) and a proximity term obtained by multiplying the second coefficient β by the proximity evaluation function Proximity(φ,ω), and the solution with the largest overall evaluation function EF among multiple solutions obtained by solving the multi-objective optimization problem is proposed as the layout result. According to the above configuration, in the structural equation modeling step S2, structural equation modeling is applied to extract multiple factors including privacy and proximity, and an overall effect on privacy and an overall effect on proximity are derived for each of the multiple factors other than privacy and proximity. The overall effect on privacy for each of the multiple factors derived in this way represents the strength of the influence that privacy has on that factor. Similarly, the overall effect on proximity for each of the multiple factors represents the strength of the influence that proximity has on that factor. Here, the overall evaluation function EF is constructed by adding a privacy term obtained by multiplying the first coefficient α by the privacy evaluation function Privacy(φ,ω) and a proximity term obtained by multiplying the second coefficient β by the proximity evaluation function Proximity(φ,ω), where α is the total privacy effect calculated for one factor selected from multiple factors, and β is the total proximity effect calculated for the one factor. That is, the first coefficient α and the second coefficient β represent the strength of the influence of privacy and proximity on one factor, respectively. Therefore, by multiplying the first coefficient α and the second coefficient β by the privacy evaluation function Privacy(φ,ω) and the proximity evaluation function Proximity(φ,ω), respectively, the first coefficient α and the second coefficient β act as weights for the privacy evaluation function Privacy(φ,ω) and the proximity evaluation function Proximity(φ,ω). In this way, the overall evaluation function EF can be appropriately expressed.

[0058] Furthermore, the office layout design support device 1 as described above is an office layout design support device 1 that supports layout design within an office, and includes: a formulation unit 14 that formulates a privacy evaluation function Privacy(φ, ω) for all workers and an evaluation function Proximity(φ, ω) for all workers' proximity to surrounding workers, using layout variables φ and ω that include both the placement of each of a plurality of desks D, at a plurality of candidate positions CP, at which each of a plurality of workers is seated, and the height of the partition PA to be provided for each of the plurality of desks D; and a layout determination unit 15 that determines the values ​​of each of the layout variables φ and ω for each of the plurality of desks D by applying an evolutionary computation method to objective functions F1 and F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω). According to the above-described configuration, similar to the explanation of the office layout design support method, it becomes possible to design a layout in an office taking into consideration the seating environments of each of a plurality of workers.

[0059] (First Modification of the Embodiment) In the above embodiment, as shown in equations (10) and (11), the objective functions F1 and F2 comprise a first objective function F1 based on a privacy evaluation function Privacy(φ, ω) and a second objective function F2 based on a proximity evaluation function Proximity(φ, ω), and in the layout determination step S5, the values ​​of the layout variables φ and ω are determined by applying an evolutionary computation method to the first objective function F1 and the second objective function F2 to solve a multi-objective optimization problem, but this is not limited to this. As the objective function to be used to solve the optimization problem, for example, only one function expressed as the following equation (14) may be used.

number

[0060] As described above, in this modified example, the objective function F is constructed by adding a privacy term obtained by multiplying the first coefficient α and the privacy evaluation function Privacy(φ,ω) and a proximity term obtained by multiplying the second coefficient β and the proximity evaluation function Proximity(φ,ω), and in the layout determination step S5, an evolutionary computation method is applied to the objective function F to solve a single-objective optimization problem, thereby determining the values ​​of each of the layout variables. According to the above configuration, the office layout design support method can be appropriately implemented.

[0061] (Second Modification of the Embodiment) Furthermore, in the above embodiment, the layout variables φ and ω were configured to include both the placement of each of the multiple desks D, where each of the multiple workers is seated, at multiple candidate positions CP, and the height of the partitions PA to be provided for each of the multiple desks D, but this is not limited to this. For example, the layout variables φ and ω may include the placement of each of a plurality of desks D, where each of a plurality of workers is seated, at a plurality of candidate positions CP, but may not include the height of the partitions PA to be provided for each of the plurality of desks D. This type of configuration can be used when partitions PA are not used in the office layout, or when only the placement of desks D needs to be considered with the height of partitions PA fixed.

[0062] More specifically, the office layout design support method of this modified example is an office layout design support method that supports layout design within an office, and includes: a formulation step S4 in which a privacy evaluation function Privacy(φ, ω) for all workers and an evaluation function Proximity(φ, ω) for all workers of proximity to surrounding workers are formulated using layout variables φ and ω that do not include the height of the partition PA to be provided for each of the desks D, and a layout determination step S5 in which an evolutionary computation method is applied to objective functions F1 and F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω), thereby determining the values ​​of the layout variables φ and ω for each of the desks D. Moreover, the office layout design support device of this modified example is an office layout design support device that supports layout design within an office, and includes: a formulation unit 14 that formulates a privacy evaluation function Privacy(φ, ω) for all workers and an evaluation function Proximity(φ, ω) for all workers' proximity to surrounding workers, using layout variables φ, ω that include the arrangement of each of a plurality of desks D, at a plurality of candidate positions CP, and that do not include the height of the partitions PA to be provided for each of the plurality of desks D; and a layout determination unit 15 that determines the values ​​of each of the layout variables φ, ω for each of the plurality of desks D by applying an evolutionary computation method to objective functions F1, F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω). With this configuration, even if the layout variables φ and ω do not include the height of the partitions PA provided for each of the multiple desks D, it is possible to design a layout in an office that takes into account the seating environment of each of the multiple workers.

[0063] (Third Modification of the Embodiment) Similarly, the layout variables φ and ω may include the height of the partitions PA to be provided for each of the desks D, without including the arrangement of each of the desks D at the multiple candidate positions CP where each of the multiple workers is seated. This configuration can be used in cases where it is desired to consider only the height of the partitions PA while keeping the arrangement of the desks D fixed.

[0064] More specifically, the office layout design support method of this modified example is an office layout design support method that supports layout design within an office, and includes: a formulation step S4 in which a privacy evaluation function Privacy(φ, ω) for all workers and an evaluation function Proximity(φ, ω) for all workers' proximity to surrounding workers are formulated using layout variables φ and ω that do not include the placement of each of a plurality of desks D, at a plurality of candidate positions CP, but include the height of the partition PA to be provided for each of the plurality of desks D; and a layout determination step S5 in which an evolutionary computation method is applied to objective functions F1 and F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω), thereby determining the values ​​of the layout variables φ and ω for each of the plurality of desks D. Moreover, the office layout design support device of this modified example is an office layout design support device that supports layout design within an office, and includes: a formulation unit 14 that formulates a privacy evaluation function Privacy(φ, ω) for all workers and an evaluation function Proximity(φ, ω) for all workers' proximity to surrounding workers, using layout variables φ, ω that do not include the placement of each of a plurality of desks D at a plurality of candidate positions CP, but include the height of a partition PA to be provided for each of the plurality of desks D; and a layout determination unit 15 that determines the values ​​of each of the layout variables φ, ω for each of the plurality of desks D by applying an evolutionary computation method to objective functions F1, F2 formulated based on the privacy evaluation function Privacy(φ, ω) and the proximity evaluation function Proximity(φ, ω). With this configuration, even if the layout variables φ and ω do not include the placement of each of the multiple desks D at multiple candidate positions CP, it is possible to design a layout in an office that takes into account the seating environments of each of the multiple workers.

[0065] The office layout design support method and office layout design support device of the present invention are not limited to the above-described embodiments and modifications explained with reference to the drawings, and various other modifications are possible within the technical scope.

[0066] For example, in the above embodiment, the isovist area is calculated for each of a plurality of radii and a plurality of eye heights, but this is not limited to this. For example, the isovist area may be calculated within a certain angle range centered on the viewpoint SP, and this angle range may include multiple types, such as 180° in front of the respondent and 180° behind the respondent.

[0067] Furthermore, in the above embodiment, a questionnaire and an isovist area survey are conducted before executing the office layout design support method, but this is not limited to this. For example, if a questionnaire and an isovist area survey have already been conducted in the past for a work environment similar to the work environment of the employees of the office for which the layout design is to be performed, the layout design may be performed using the results of the past questionnaire and isovist area survey without conducting a new questionnaire and isovist area survey for the employees of the office for which the layout design is to be performed.

[0068] In addition to this, it is possible to select and discard the configurations given in the above embodiments and modifications, or to change them to other configurations as appropriate. [Explanation of symbols]

[0069] 1 Office layout design support device D desk 11 Pre-factor analysis part PA partition 12 Structural Equation Modeling Section SP Perspective 13 Correlation analysis section S1 Preliminary factor analysis step 14. Structural Equation Modeling Step S2 15 Layout decision part S3 Correlation analysis step C Circle S4 Vertical Step CP candidate position S5 Layout decision step

Claims

1. An office layout design support method for supporting layout design in an office, comprising: a formulating step of formulating an evaluation function of privacy for all of the workers and an evaluation function of proximity for all of the workers to surrounding workers, using layout variables including one or both of the arrangement of each of a plurality of desks at which each of a plurality of workers is seated at a plurality of candidate positions and the height of a partition to be provided for each of the plurality of desks as variables; a layout determination step of determining values ​​of the layout variables for each of the plurality of desks by applying an evolutionary computation method to an objective function formulated based on the privacy evaluation function and the proximity evaluation function; An office layout design support method comprising:

2. the layout variables include both arrangement of each of the plurality of desks at the plurality of candidate positions and height of the partition; the objective function comprises a first objective function based on the evaluation function of privacy and a second objective function based on the evaluation function of proximity, In the layout determination step, a value of each of the layout variables is determined by applying an evolutionary computation method to the first objective function and the second objective function to solve a multi-objective optimization problem.

2. The office layout design support method according to claim 1.

3. a structural equation modeling step of applying structural equation modeling to the results of the questionnaire regarding the working environment to extract a plurality of factors including privacy and proximity, deriving an overall effect regarding privacy and an overall effect regarding proximity for each of the plurality of factors other than privacy and proximity, and deriving a factor score regarding privacy and a factor score regarding proximity for each respondent of the questionnaire; a correlation analysis step in which, based on the results of a survey of each of the plurality of respondents for each of a plurality of radii and a plurality of eye-level heights, an isovist area, which is the area of ​​an area visible to the respondent within a circle extending in a horizontal plane centered on the viewpoint of the respondent when seated at his / her own seat, is performed to perform a correlation analysis between the factor score related to privacy for each of the plurality of respondents and the survey results, and between the factor score related to proximity for each of the plurality of respondents, to derive a privacy correlation coefficient, which is the correlation coefficient between the isovist area and privacy, and a proximity correlation coefficient, which is the correlation coefficient between the isovist area and proximity, for each of the plurality of radii and the plurality of eye-level heights; Further provided with In the formulating step, a combination of the radius and the eye level for which the absolute value of the privacy correlation coefficient is large is extracted, and the evaluation function for privacy is formulated using the radius and the isovist area at the eye level for the combination as a variable, and a combination of the radius and the eye level for which the absolute value of the proximity correlation coefficient is large is extracted, and the evaluation function for proximity is formulated using the radius and the isovist area at the eye level for the combination as a variable.

3. The office layout design support method according to claim 2.

4. In the layout determination step, the overall effect on privacy calculated for one factor selected from the plurality of factors is set as a first coefficient, the overall effect on proximity calculated for the one factor is set as a second coefficient, and a privacy term obtained by multiplying the first coefficient by an evaluation function for privacy and a proximity term obtained by multiplying the second coefficient by an evaluation function for proximity are added to construct an overall evaluation function, and the solution with the largest overall evaluation function is proposed as a layout result from among a plurality of solutions obtained by solving a multi-objective optimization problem.

4. The office layout design support method according to claim 3.

5. the objective function is constructed by adding a privacy term obtained by multiplying a first coefficient and the privacy evaluation function and a proximity term obtained by multiplying a second coefficient and the proximity evaluation function; In the layout determination step, an evolutionary computation method is applied to the objective function to solve a single-objective optimization problem, thereby determining the values ​​of each of the layout variables.

2. The office layout design support method according to claim 1.

6. An office layout design support device that supports layout design in an office, a formulating unit that formulates an evaluation function of privacy for all of the workers and an evaluation function of proximity for all of the workers to surrounding workers, using layout variables including one or both of the arrangement of each of a plurality of desks at which each of a plurality of workers is seated at a plurality of candidate positions and the height of a partition to be provided for each of the plurality of desks as variables; a layout determination unit that determines values ​​of the layout variables for each of the desks by applying an evolutionary computation method to an objective function formulated based on the privacy evaluation function and the proximity evaluation function; An office layout design support device comprising:

Images