JP7645168B2 - Worker combination calculation device, work team management system, and worker combination calculation method - Google Patents

Description

The present invention relates to a worker combination calculation device, a work team management system, and a worker combination calculation method.

In recent years, there has been growing expectation for technology that uses IT (Information Technology) to efficiently and appropriately manage the work performance and physical condition (health status) of employees, thereby enabling the implementation of appropriate improvement measures to improve the work performance of employees.

Patent document 1 describes an invention that proposes appropriate measures to improve employees' work performance by managing the work performance status and physical condition of employees with high accuracy.

Patent No. 6724948

For example, in nuclear power plants where radiation levels are high, workers operate robots from remote locations to perform specific tasks. However, remote work in unknown environments generally places a high mental burden on the workers who operate the robots. Furthermore, when multiple workers operate a robot remotely, the work efficiency varies greatly depending on the compatibility between the workers and the condition of each worker.
To achieve efficient remote work, it is necessary to manage work teams based on the compatibility between workers and their status at the time.

The present invention aims to take into account the compatibility of workers when allocating work teams.

In order to solve the above-mentioned problems, the worker combination calculation device of the present invention is characterized by comprising: a work efficiency calculation unit that calculates the work efficiency of past combinations of workers from statistics of work analysis data for the combination of workers engaged in the work; a worker combination calculation unit that ranks the work efficiency levels of the work teams engaged in the work and stores them in a work status database ; and a work efficiency estimation unit that refers to the past work efficiency levels of each work team and their work results accumulated in the work status database , and estimates the work efficiency of the work team formed by combining workers using a statistical method .

The work team management system of the present invention is characterized by comprising: a work efficiency calculation unit that calculates the work efficiency of past combinations of workers from statistics of work analysis data for the combination of workers engaged in the work; a worker combination calculation unit that ranks the work efficiency levels of the work teams engaged in the work and stores them in a work status database; and a work efficiency estimation unit that references the past work efficiency levels of each work team and their work results accumulated in the work status database and estimates the work efficiency of the work team formed by combining workers using statistical methods .

The worker combination calculation method of the present invention is characterized in that it executes the steps of: a work efficiency calculation unit calculating the work efficiency of a past combination of workers from statistics of work analysis data for the combination of workers engaged in the work; a work efficiency calculation unit ranking the work efficiency levels of the work teams engaged in the work and storing them in a work status database ; and a work efficiency estimation unit referencing the past work efficiency levels of each work team and their work results accumulated in the work status database , and estimating the work efficiency of the work team formed by combining workers using a statistical method .
Other means will be described in the description of the embodiment of the invention.

The present invention makes it possible to take into account the compatibility of workers when allocating work teams.

1 is a configuration diagram of a work team management system according to a first embodiment. FIG. 2 is a configuration diagram of a work status processing device. This is data output from the worker vital signs measurement unit. This is an image of vital analysis data output. FIG. 2 is a diagram showing the configuration of a work status database. This is a data management method for a work status database using a time stamp method. This is a data management method for a work status database using a work name method. FIG. 2 is a configuration diagram of a worker combination calculation device. FIG. 2 is a configuration diagram of a work management device. 13 is a flowchart of a work team management process. 13 is a flowchart of a work team management process. This is an example of data calculation. 1 is a graph showing the fatigue and tension levels of workers. This is a graph showing the degree of compatibility calculated from the fatigue and tension levels of workers. 1 is an example of histogram data calculation of frequency characteristics. 1 is an example of a histogram data calculation according to a selected statistical indicator. 13 is a detailed flowchart of a worker combination ranking calculation process. 13 is a detailed flowchart of a work efficiency calculation process. This is a deep learning model for work efficiency calculation processing. 13 is a detailed flowchart of a modified example of a worker combination ranking calculation process. 13 is a detailed flowchart of a working efficiency calculation process according to a modified example. FIG. 11 is a configuration diagram of a work team management system according to a second embodiment. 13 is a flowchart of a work team management process. 13 is a flowchart of a work team management process. FIG. 2 is a configuration diagram of a worker combination calculation device. 13 is a flowchart of a process for setting an objective function and constraint conditions. FIG. 13 is a diagram showing change data of work efficiency for each work time period of worker A. FIG. 13 is a diagram showing change data of work efficiency for each work time period of worker B.

The following describes in detail how to implement the present invention with reference to the figures.

First Embodiment
FIG. 1 is a configuration diagram of a work team management system 1 according to the first embodiment.
The work team management system 1 includes a data interface unit 11, a work status processing device 12, a work status database 13, a worker combination calculation device 14, and a work management device 15. This work team management system 1 allows workers 2a to 2c to update a work plan using, for example, a work robot system (not shown).

Here, workers 2a to 2c are robot operators, assistant workers, work supervisors, etc. Workers 2a to 2c have individual differences in ability, and there is variation in work processing ability. Even when the same worker performs the same work, there is variation in ability depending on their physical condition at the time. A work team is a team of multiple people who perform the same work together. It is advisable to organize a work team taking into account the differences in the initial abilities of individuals.

The data interface unit 11 is an interface for exchanging data between the workers 2a-2c or the working robot system and the work team management system 1. The data interface unit 11 is composed of, for example, a group of multiple types of sensors, a wireless router that transmits instructions to the worker's terminal and the robot, etc. The data interface unit 11 issues work instructions to the workers 2a-2c and acquires vital data (described below) from the workers 2a-2c at pre-specified time intervals. The working robot system (not shown) is remotely controlled by the workers 2a-2c, etc., and includes the robot, its control device, an operation log recording device, an operating device, a monitoring device, and associated equipment for these.
This data interface unit 11 acquires work-related data from the workers 2a to 2c at pre-specified time intervals.

The work-related information 4 refers to information such as the layout of the work environment, temperature, humidity, hydrogen concentration, oxygen concentration, radiation dose, as well as the date, time, etc. The work-related information 4 is on-site data that measures the status of the workers 2a to 2c and the work robot system. The data interface unit 11 acquires this on-site data at pre-specified time intervals.

The work attribute information 5 is a history indicated by work results such as the work name, work purpose, work difficulty, work organization, worker proficiency, worker history, work process, work time period, work equipment specifications, work equipment history, work completion time, and work time for each work. The data interface unit 11 updates this history at pre-specified time intervals. The work results include whether the work was successful, the work time (including the degree of delay), radiation exposure amount, cumulative dose, whether a risk event occurred, and whether the plan was updated.

The work status processing device 12 analyzes the work status and work attributes based on the worker status and on-site data that measure the status of the workers 2a to 2c, and stores this analysis data and on-site data in the work status database 13. The work status processing device 12 further outputs the analysis data to the worker combination calculation device 14. In other words, the work status processing device 12 functions as a work status processing unit that analyzes the data acquired by the data interface unit 11.

The worker status analyzed by the work status processing device 12 refers to the vital data and movement data of the workers 2a to 2c. Vital data includes heart rate, brain waves, sweat rate, body temperature, blood pressure, physical condition, etc. Movement data includes acceleration, position, movement path/history, line of sight, number of steps, etc.

The work conditions analyzed by the work status processing device 12 include the date of work, the time period of work, whether it was on a weekday, Saturday, Sunday, or holiday, the work hours per day, the total number of work days and the cumulative number of work days, the temperature and humidity in the work environment, the layout of structures, images and point clouds of the work environment, etc.
The work attributes analyzed by the work status processing device 12 include the work purpose, work difficulty (risk level), work urgency, work team structure, worker proficiency, and work system.

The worker combination calculation device 14 calculates a combination of workers based on the analysis data and history data. When it detects a state in which the data acquired by the work status processing device 12 exceeds a pre-specified threshold, the worker combination calculation device 14 calculates a combination of workers who will be engaged in this work. The worker combination calculation device 14 then stores the calculation result of the worker combination in the work status database 13, and outputs it to the work management device 15.
In other words, the worker combination calculation device 14 functions as a worker combination calculation unit that calculates the combination of workers engaged in the work.

The work management device 15 outputs operation instruction data to the workers 2a to 2c based on the work resource calculation results.

FIG. 2 is a configuration diagram of the work status processing device.
The work status processing device 12 includes a worker vital sign measuring unit 121 , a work state analysis and evaluation unit 122 , and a work attribute analysis and evaluation unit 123 .

The work status processing device 12 acquires data from equipment element testing, equipment combination testing, mock-up testing, training, actual equipment construction, etc.

The worker vitals measuring unit 121 measures the condition of the workers 2a to 2c and outputs vital data and movement data. Here, vital data refers to the worker's heart rate, brain waves, sweat rate, body temperature, blood pressure, physical condition, etc. Movement data refers to acceleration, position, movement path/history, line of sight, number of steps, etc.

3 is a graph showing an example of data output from the worker vital sign measuring unit 121. The vertical axis of the graph indicates the measured value of the worker's vital signs, and the horizontal axis indicates time.
The solid line in the graph indicates acceleration, the dashed line in the graph indicates heart rate, and the dashed line in the graph indicates sweat rate.

Returning to Figure 2 for further explanation, the work status analysis and evaluation unit 122 analyzes vital data and motion data, which are the status of each worker 2a to 2c, to evaluate their work status.

4 is a graph showing an example of output data of vital analysis data of each worker, in which the vertical axis of the graph indicates the measured value of each worker, and the horizontal axis indicates time.
The vital analysis data refers to fatigue level, concentration level, happiness level, relaxation level, alertness level, and agitation level. The range of each analysis data indicates the range of each task time. Variations due to each individual's physical condition are set as deviations from the standard value based on each individual's long-term data.

There is also variation in the initial values and deterioration curves (differential values) in Figure 4. The work status analysis and evaluation unit 122 sets vital stability and initial value tolerance ranges for standard work.

FIG. 5 is a diagram showing the configuration of the work status database 13.
The work status database 13 includes a worker compatibility record management unit 131, a work result record management unit 132, a work status record management unit 133, and a work attribute record management unit 134. The work status database 13 stores the on-site data and analysis data shown in Fig. 3 and Fig. 4, as well as associated data such as processing conditions thereof.

The worker compatibility record management unit 131 stores the vital data and movement data of workers 2a to 2c measured by the worker vitals measurement unit 121.

The work result record management unit 132 records and manages the status data of the work evaluated by the work status analysis evaluation unit 122 .
The work status record management unit 133 records and manages the attribute status of a work. The time stamp method and the work name method are just examples of the management method, and any method may be used as long as no data is lost during data management.

FIG. 6 shows a data management method for the work status database 13 using a time stamp system.
6 includes a date column, a time column, a worker status column, a work attribute column, a work status column, and a work incidental information column. The worker status column includes a heart rate column, an acceleration column, a tension level column, a happiness level column, etc.
The work status processing device 12 stores data in the work status database 13 using the time stamp as a key.

FIG. 7 shows a data management method for the work status database 13 using the work name method.
The work status database 13 in Fig. 7 includes a work attribute column, a date column, a time column, a worker status column, a work attribute column, a work status column, and a work incidental information column. The work attribute column stores the name of the work. The worker status column includes a heart rate column, an acceleration column, a tension level column, a happiness level column, etc.
The work status processing device 12 stores data in the work status database 13 using the work name as a key.

FIG. 8 is a configuration diagram of the worker combination calculation device 14.
The worker combination calculation device 14 includes a worker compatibility estimation unit 141, a worker combination calculation unit 142, and a work efficiency estimation unit 143. The worker combination calculation device 14 calculates compatibility between combinations of workers based on the analysis data and history data.

The worker compatibility estimation unit 141 quantifies or levels the compatibility of the workers engaged in the work, and estimates the compatibility of the workers. The worker combination calculation unit 142 ranks the compatibility levels of the work teams engaged in the work based on the compatibility of the workers calculated by the worker compatibility estimation unit 141, and calculates the combination of workers. The worker combination calculation unit 142 ranks the compatibility levels of the work teams using a mathematical model that implements global optimization or local optimization. Here, implementing global optimization refers to, for example, optimizing the overall work related to the plant. Also, local optimization refers to, for example, optimizing the work of a building that constitutes part of the plant.
The work efficiency estimation unit 143 refers to the past compatibility levels and work results of each work team stored in the work status database 13, and estimates the work efficiency of a work team made up of a combination of workers.

FIG. 9 is a configuration diagram of the work management device 15.
The work management device 15 includes a work process management unit 151 and a worker allocation management unit 152 .

The work process management unit 151 manages the work process. The worker allocation management unit 152 manages the allocation of workers to work teams.

10A and 10B are a flow chart of the work team management process.
First, the workers 2a to 2c operate the robot to perform a task (step S10).

The work status processing device 12 measures the work status and analyzes and evaluates it (step S11). Specifically, the worker vitals measurement unit 121 measures the worker's vitals and regards them as the work status. The work status analysis and evaluation unit 122 analyzes and evaluates the work status. And the work attribute analysis and evaluation unit 123 analyzes and evaluates the work attributes.

Next, the work status processing device 12 records and manages the work status in the work status database 13 (step S12). Information recorded in the work status database 13 is, for example, vital data and motion data of the worker, as well as data obtained by analyzing and evaluating the same.
Then, the worker combination calculation device 14 judges whether or not a deployment calculation is necessary (step S13). Here, the worker combination calculation device 14 judges that a deployment calculation is necessary when the data measured by the work status processing device 12 or the data obtained by analyzing the measured data exceeds a pre-specified threshold value.

If the worker combination calculation device 14 determines that a placement calculation is necessary (Yes), it proceeds to step S20 in FIG. 10B. If the worker combination calculation is not necessary (No), it proceeds to processing in step S14.

In step S14, the work status processing device 12 determines whether the work has been completed. Here, the completion of the work is determined, for example, by whether the work supervisor has input that the work has been completed. If the work has not been completed (No), the work status processing device 12 returns to the processing of step S10, and if the work has been completed (Yes), the processing of FIG. 10A ends.

In step S20, the worker combination calculation device 14 sets an objective function and constraint conditions. The objective function is a function that calculates the compatibility between workers and work efficiency. Constraint conditions are, for example, conditions related to work attributes and work conditions. Work attributes refer to the work shift (work team) composition, work type, number of workers, the presence/absence of core workers (individuals essential for work to be completed), etc. And work conditions refer to the work time period, temperature, humidity, etc.

Next, the worker combination calculation device 14 selects input items for calculating the compatibility of workers (step S21). Specifically, the worker combination calculation device 14 selects from the analysis data indices such as fatigue level, concentration level, happiness level, relaxation level, alertness level, and agitation level. The worker combination calculation device 14 may also select one or more indices using an analysis method. Here, work conditions such as work time zones and work attributes such as work shift configurations may be attached to the analysis data.

Then, the worker combination calculation device 14 calculates the compatibility of the workers (step S22). The compatibility of the workers here means the work efficiency in the past combination of workers. The worker combination calculation device 14 calculates the work efficiency from the statistics of the analysis data in a certain time period. The statistics of the analysis data in a certain time period are a time trend, a frequency trend, or a data histogram in a statistical index. The statistical index means the mean, median, variance, standard deviation, minimum value, maximum value, first quartile, third quartile, total number (sum), first derivative, second derivative, skewness, kurtosis, etc. When calculating the statistics of the analysis data, the worker combination calculation device 14 may filter the analysis data in any time range or frequency range (low pass filter, high pass filter, band pass filter). An example of filtering the analysis data in a frequency range will be described later with reference to FIG. 14 and FIG. 15.

Next, the worker combination calculation device 14 classifies the compatibility of the workers (step S23). Specifically, the worker combination calculation device 14 quantifies the compatibility of the workers and classifies them into levels, and sets a numerical range and a compatibility range (e.g., good/average/bad) in the worker compatibility calculation. The numerical range and the compatibility range are stored in the work status database 13, and can be changed as appropriate according to the history data. Note that the change history of the numerical range and the compatibility range may be saved, and the compatibility range may be automatically updated by using a machine learning algorithm.
Then, the worker combination calculation device 14 calculates the ranking of the combinations of workers (step S24). The worker combination calculation device 14 calculates the ranking by arranging the combinations of workers in the order of the worker compatibility results with reference to the work attributes, work states, and worker compatibility results.

Next, the worker combination calculation device 14 determines whether the calculation of the corresponding worker combination has been completed (step S25).

In step S25, if the calculation of the worker combination has not been completed (No), the worker combination calculation device 14 returns to step S20, and if the calculation of the worker combination has been completed (Yes), the worker combination calculation device 14 proceeds to step S26.

In step S26, the work supervisor decides to update the work team based on the combination of workers calculated by the worker combination calculation device 14. In response to this decision, the work management device 15 updates and manages the worker allocation (step S27). As a result, the worker and robot resume work (step S28), and the process returns to step S10 in FIG. 10A.

11 is a graph for explaining an example of data calculation. The vertical axis of the graph is the analysis value. The horizontal axis of the graph is the time. The dashed line is the heart rate of worker A. The solid line is the work efficiency of the work team.
In each of the time ranges a and b, the worker combination calculation device 14 calculates the statistics of each analysis data.

《Calculating the compatibility between workers》
Kazuo Yano, author of "Measuring Happiness with Wearable Technology," and his colleagues found that there is a certain correlation between physical exercise and happiness. When there is 1/T fluctuation in the relationship between the duration and frequency of physical exercise, happiness is high, and when there is unnatural fluctuation, happiness is low.
・Kazuo Yano et al., "Measuring Happiness with Wearable Technology," Hitachi Hyoron, June 17, 2015, Internet <URL:https://www.hitachihyoron.com/jp/pdf/2015/06_07/2015_06_07_13.pdf>
Therefore, by analyzing the vital data of each worker to measure the duration of physical movement and calculate whether the relationship between the duration of physical movement and its frequency is close to 1/T fluctuation, it is possible to calculate the happiness level of each worker.

The worker compatibility estimation unit 141 measures the duration of fatigue and tension of each worker during each shift period, and calculates whether the relationship between the duration and frequency is close to 1/T fluctuation, thereby calculating the happiness degree x _n of each worker during each shift. Then, the compatibility degree f of a combination of workers in the same shift who have similar happiness degrees x _n calculated is calculated using the following formula (1). The compatibility degree threshold and weighting coefficient α may be set by a work supervisor or a work planner, etc., with reference to past cases and interviews and questionnaires with workers.

"Positive Intelligence" by S. Achor, Harvard Business Review (2012.1), reports that a person's sense of happiness has a significant impact on performance. Happy people are 37% more productive in sales and 300% more creative than those who are not.
Therefore, the worker compatibility estimation unit 141 of the present embodiment estimates a combination of workers with high compatibility so that each worker feels a greater sense of happiness. By forming a work team with a combination of multiple workers with high compatibility, the performance of the work team can be improved.

In addition, the compatibility degree f of a combination of workers may be calculated using the following formula (2).

Figure 12 is a graph showing the fatigue and tension of workers. The vertical axis of the graph shows the fatigue and tension of workers. The horizontal axis of the graph shows time, which is divided into shift periods A to E. The fatigue and tension of workers are manifestations of physical exercise.

Figure 13 is a graph showing the degree of compatibility between workers A and B, calculated from the fatigue and tension levels of the workers. The vertical axis of the graph shows the degree of compatibility. The horizontal axis of the graph shows the period of each shift.

Figure 14 is a graph illustrating an example of a histogram data calculation of frequency characteristics. The vertical axis of the graph is the analysis value. The horizontal axis of the graph is the frequency. The solid line is the histogram of the worker's acceleration.

In the frequency range of the graph in FIG. 14, the histogram of the worker acceleration exceeds a predetermined value. Therefore, the worker combination calculation device 14 can know that periodic operations in this range are occurring frequently.

Figure 15 is a graph illustrating an example of histogram data calculation using a selected statistical index. The vertical axis of the graph is the analysis value. The horizontal axis of the graph is the statistical index. The dashed line shows the correspondence between the index in the analysis range and the analysis value.

16 is a detailed flowchart of the worker combination ranking calculation process, which is a detailed explanation of the process in step S24 of FIG. 10B.
The worker combination calculation unit 142 refers to the task attributes, task status, and worker compatibility results (step S60). Here, the task attributes include any of the number of workers engaged in the task, the difficulty of the task, the number of days for the task, the working hours per day for the task, the work schedule of the workers engaged in the task, the area environment for the task, and the radiation exposure dose of the workers engaged in the task.
Then, the worker combination calculation unit 142 arranges the workers in order of their compatibility results (step S61), and ends the processing in FIG.

17 is a detailed flowchart of the work efficiency calculation process, which is a detailed example of step S24 in FIG. 10B and is executed after the process in FIG.
The worker combination calculation unit 142 refers to the task attributes, task status, and worker compatibility results (step S70). Then, the task efficiency estimation unit 143 extracts the task efficiency in this worker combination example (step S71), and ends the process in FIG.

The worker combination calculation device 14 sets a numerical range and a compatibility range when calculating the compatibility of workers. The numerical range and the compatibility range are types such as good, normal, and poor, and are stored in the work status database 13 and can be changed as appropriate according to the historical data.

FIG. 18 shows a deep learning model 1431 that realizes the work efficiency calculation process.
The deep learning model 1431 takes a combination of task attributes, numerical ranges, and compatibility ranges for a given task team as input, and performs regression to minimize the error between the output of the deep learning model 1431 and task efficiency. Task attributes, numerical ranges, and compatibility ranges are input to the 0th layer of the deep learning model 1431. The nth layer of the deep learning model 1431 is the output, and the error with the task efficiency, which is the teacher data, is calculated, and regression calculation is performed to minimize this error. This makes it possible to calculate task efficiency for a combination of unknown task attributes, numerical ranges, and compatibility ranges.

The worker combination calculation device 14 may store a history of changes to the numerical range and compatibility range, and automatically update the compatibility range by using a machine learning algorithm using the deep learning model 1431 or the like.

When considering individual and short-term issues, the worker combination calculation device 14 should accumulate and manage data showing the relationship between the decline in work efficiency due to lack of concentration and vital data such as brain waves and pulse rate. The worker combination calculation device 14 should then probabilistically calculate work efficiency from the vital data.

When considered on an individual basis and over the long term, the learning effect reduces the time it takes to solve routine problems. The relationship between these vital data and work efficiency varies from person to person, and can be plotted as a curve over time.

When considering a work team, in the short term it is often better to collect individual data from members with similar curves. This makes it easier to manage shift times, as each member of the work team will experience similar levels of fatigue and deterioration.

On the other hand, if individual data is collected for members with similar curves, the entire work team will deteriorate over time. This is dangerous when it comes to the ease of managing the work team's shift times alone, and it may be better to pair workers with long-term focus with workers with short-term focus. In other words, for tasks where failure is not tolerated, it is a good idea to change the worker selection policy and pair workers with long-term focus with workers with short-term focus.

In addition, it is thought that the work efficiency data for a work team may tend to differ from that inferred from the work efficiency data of each individual worker. This is because workers have different chemistry with each other. Therefore, this system reflects the chemistry between workers based on stress data such as the workers' brain waves in the composition of the work team.

Furthermore, the system improves the accuracy of updates by the work team by making it possible to reflect history using machine learning and other techniques.
When considering a work team over the long term, we can expect a learning effect that shortens the time it takes to complete prescribed standard practice problems, just as we can with data obtained on an individual basis. Also, it is possible that a work team that is initially compatible may be formed, but that compatibility may deteriorate over time. In other words, the efficiency curve may necessitate a review of the composition of the work team. Conversely, there are also cases where a work team's work efficiency is poor at the beginning, but a growth curve can be drawn that shows improvement over time.

The work team management system 1 of this embodiment may be able to handle the above-mentioned cases by setting the update cycle of the team management.
For a core worker in a work team, the entire work team depends on that particular person, so work planning and personnel allocation are centered on that person. This is also an example of reflecting personal data in the work team management plan.

FIG. 19 is a detailed flowchart of the worker combination ranking calculation process of the modified example.
The worker combination calculation unit 142 refers to the task attributes, task status, and worker compatibility results, including the history information in the task status database (step S80).

Then, the worker combination calculation unit 14 determines whether or not the task is one in which failure is permitted (step S81). If the task is not one in which failure is permitted (No), the worker combination calculation unit 14 changes the worker selection policy, defines a combination of workers whose vital data changes over time are dispersed as a result of good compatibility (step S82), and proceeds to step S84. In other words, the worker combination calculation unit 14 defines a combination of a worker who has poor concentration sustained and a worker who has good concentration sustained as a result of good worker compatibility.

If in step S81 the worker combination calculation unit 14 determines that the task is not one for which failure is permitted (No), it proceeds to step S83, where it defines a combination of workers with similar changes in vital data over time as a result of good compatibility, and proceeds to step S84. In other words, the worker combination calculation unit 14 defines a combination of workers with similar levels of concentration as a result of good compatibility.

In step S84, the worker combination calculation unit 142 performs optimization calculations using a mathematical model and sorts the worker compatibility results in order (step S85).

In other words, in the worker combination calculation step of ranking the compatibility levels of work teams, the worker combination calculation unit 142 performs calculations using a calculation algorithm based on a mathematical model that performs global optimization or local optimization.

Then, the worker combination calculation unit 14 uses a machine learning algorithm or a mathematical model to extract a combination of similar tasks and work efficiency with a work efficiency higher than a predetermined value (step S86), and ends the processing of FIG. 19.

FIG. 20 is a detailed flowchart of the working efficiency calculation process of the modified example.
The worker combination calculation unit 142 refers to the work attributes, work status, and worker compatibility results, including the history information in the work status database (step S90). The work efficiency estimation unit 143 then performs calculations using statistical methods (step S91), calculates a work efficiency example for this worker combination (step S92), and ends the processing in FIG.

The worker combination calculation device 14 estimates the work efficiency of a combination of workers by referring to the compatibility levels and work results of past work teams stored in the work situation database 13. In doing so, the worker combination calculation device 14 takes work attributes such as the number of workers, task difficulty, number of work days, work hours per day, work system, work area environment, and radiation exposure dose as input values, and weights each input value with a coefficient according to the work situation.

Second Embodiment
FIG. 21 is a configuration diagram of a work team management system 1A according to the second embodiment.
A work team management system 1A of the second embodiment further includes an overall work management device 16 in addition to the work team management system 1 of FIG.

The overall work management device 16 manages the overall work. When the overall work management device 16 receives work management data from the work management device 15, it creates update data for each task and stores it in the work status database 13. The overall work management device 16 further outputs the created update data for each task to the work management device 15.

22A and 22B are a flow chart of the work team management process.
First, the workers 2a to 2c operate the robot to perform a task (step S30).

The work status processing device 12 measures the work status and analyzes and evaluates it (step S31). Next, the work status processing device 12 records and manages the work status in the work status database 13 (step S32). Then, the worker combination calculation device 14 determines whether or not a placement calculation is necessary (step S33). Here, the worker combination calculation device 14 determines that a placement calculation is necessary when the data acquired by the work status processing device 12 exceeds a pre-specified threshold value.

If the worker combination calculation device 14 determines that a placement calculation is necessary (Yes), it proceeds to step S40 in FIG. 22B. If the worker combination calculation is not necessary (No), it proceeds to processing in step S34.

In step S34, the work status processing device 12 determines whether the work is complete. If the work is not complete (No), the work status processing device 12 returns to the process of step S30, and if the work is complete (Yes), the work status processing device 12 ends the process of FIG. 22A.

In step S40, the worker combination calculation device 14 sets the objective function and constraint conditions. The worker combination calculation device 14 selects input items for calculating the compatibility of workers (step S41). The worker combination calculation device 14 then calculates the compatibility of workers (step S42). Here, the compatibility of workers refers to the work efficiency of past combinations of workers.

Next, the worker combination calculation device 14 classifies the compatibility of the workers (step S43). Specifically, the worker combination calculation device 14 converts the compatibility of the workers into numerical values and classifies them into levels (step S44). Then, the worker combination calculation device 14 calculates a ranking of the combinations of workers.

Next, the worker combination calculation device 14 determines whether the calculation of the corresponding worker combination has been completed (step S45).

In step S45, if the calculation of the worker combination has not been completed (No), the worker combination calculation device 14 returns to step S40, and if the calculation of the worker combination has been completed (Yes), the worker combination calculation device 14 proceeds to step S46.

In step S46, the work supervisor decides to update the work team based on the combination of workers calculated by the worker combination calculation device 14. In response to this decision, the work management device 15 updates and manages the worker allocation (step S47).

The overall work management device 16 manages and updates the allocation of the overall work (step S48). In step S48, the overall work management device 16 calculates worker shifts using a calculation algorithm based on a mathematical model that performs overall optimization or local optimization.
After the process of step S48, the worker and the robot resume the work (step S49) and the process returns to step S30 in FIG. 22A.

<Changes in worker efficiency depending on the time of day>
When measuring worker data during work, there are cases where the data varies depending on the time period. For example, when a 24-hour system is operated in four shifts, the work efficiency of each worker varies depending on the work time period. Therefore, the worker combination calculation device of this modified example has a function of determining a work pattern from the change data of the work efficiency of each worker, and a function of clustering workers based on the change data of the work efficiency of each worker.

FIG. 23 is a configuration diagram of the worker combination calculation device 14A.
The worker combination calculation device 14A has the same configuration as the worker combination calculation device 14 of the first embodiment, and further includes a work pattern determination unit 144 and a worker clustering unit 145.
The work pattern determination unit 144 determines a work pattern from the data on changes in the work efficiency of each worker. The worker clustering unit 145 clusters each worker based on the data on changes in the work efficiency of each worker. The subsequent processing is premised on the formation of a work team by combining the workers clustered by the worker clustering unit 145.

The worker compatibility estimation unit 141 quantifies or levels the compatibility of the workers engaged in the work, and estimates the compatibility of the workers. The worker combination calculation unit 142 provisionally organizes work teams with combinations of workers clustered by the worker clustering unit 145 based on the compatibility of the workers calculated by the worker compatibility estimation unit 141, ranks the compatibility levels of the work teams, and calculates the combinations of workers. The worker combination calculation unit 142 ranks the compatibility levels of the work teams using a mathematical model that performs global optimization or local optimization. At this time, the worker combination calculation unit 142 calculates the work efficiency of the entire work team based on the work time period and changes in work efficiency of each worker's work efficiency data.

The work efficiency estimation unit 143 estimates the work efficiency of a work team made up of workers by referring to the past compatibility levels of each work team and their work results stored in the work status database 13.

FIG. 24 is a flowchart of the process of setting the objective function and constraint conditions shown in step S20 of FIG. 10B.
The work pattern determination unit 144 sets the work efficiency of each worker as the objective function (step S100), and sets the minimum value of the work efficiency as the constraint condition (step S101). Next, the work pattern determination unit 144 selects a pattern of work hours and rest times that is optimal for the constraint condition based on the change in the work efficiency of the objective function and the minimum value of the work efficiency that is the constraint condition (step S102). This allows the worker combination calculation device 14A to select a pattern of work hours and rest times that is suitable for the current worker, thereby suppressing a decrease in work efficiency.

Then, the worker clustering unit 145 clusters the workers according to the degree of change in work efficiency (step S103), and ends the processing in FIG. 24. The worker combination calculation device 14A organizes a work team from among the clustered workers with similar degrees of change in work efficiency, so that it is possible to organize an ideal work team with little variation in changes in work efficiency.

FIG. 25 is a diagram showing data on changes in the work efficiency of worker A for each work time period.
The first column shows the start time of the work. The second column shows the work efficiency of worker A at the start of the work. The third column shows the work efficiency of worker A after one hour. The fourth column shows the work efficiency of worker A after two hours. The fifth column shows the work efficiency of worker A after three hours.

FIG. 26 is a diagram showing data on changes in the work efficiency of worker B for each work time period.
The first column shows the start time of the work. The second column shows the work efficiency of worker B at the start of the work. The third column shows the work efficiency of worker B after one hour. The fourth column shows the work efficiency of worker B after two hours. The fifth column shows the work efficiency of worker B after three hours.

For example, a work pattern may be such that a worker works for three hours from the start of work, is then allowed to rest for three hours, and then is made to work for another three hours. However, by referring to such data on changes in work efficiency, if there is a certain percentage or more of workers whose work efficiency has significantly decreased after three hours, the work pattern may be changed. For example, a work pattern may be such that a worker works for two hours from the start of work, is allowed to rest for two hours, and then is made to work for another two hours. This makes it possible to select a work pattern that is suitable for the worker.

<Changes in task difficulty and work efficiency>
Work efficiency also changes depending on the difficulty of the work. For example, the work efficiency drops significantly after a worker performs a difficult task. However, the work efficiency does not drop so much after a worker performs an easy task. It can be estimated that there is a certain trend in the data on changes in work efficiency according to the difficulty of the task. Therefore, the work status processing device 12 may estimate the work efficiency of each worker for each level of task difficulty.

The worker combination calculation device 14A may set a break after a worker performs a difficult task for a predetermined period of time, so that a worker whose work efficiency has decreased can be removed from the task.
Furthermore, the worker combination calculation device 14A may take into consideration the recovery status of a worker and set the next easiest task after the worker has completed a difficult task. This allows the worker to be assigned a task that the worker can perform even if his/her work efficiency has decreased.

(Modification)
The present invention is not limited to the above-described embodiment, and includes various modified examples. For example, the above-described embodiment has been described in detail to clearly explain the present invention, and is not necessarily limited to those having all of the configurations described. It is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. In addition, it is also possible to add, delete, or replace a part of the configuration of each embodiment with another configuration.

The above configurations, functions, processing units, processing means, etc. may be realized in part or in whole by hardware such as an integrated circuit. The above configurations, functions, etc. may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as the programs, tables, and files that realize each function can be stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or on a recording medium such as a flash memory card or a DVD (Digital Versatile Disk).

In each embodiment, the control lines and information lines are those that are considered necessary for explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are interconnected.

1, 1A Work team management system 11 Data interface unit 12 Work status processing device 121 Worker vital sign measurement unit 122 Work status analysis evaluation unit 123 Work attribute analysis evaluation unit 13 Work status database 131 Worker compatibility record management unit 132 Work result record management unit 133 Work status record management unit 134 Work attribute record management unit 14 Worker combination calculation device (worker combination calculation unit)
141 Worker compatibility estimation unit 142 Worker combination calculation unit 143 Work efficiency estimation unit 144 Work pattern determination unit 145 Worker clustering unit 15 Work management device 151 Work process management unit 152 Worker allocation management unit 16 Overall work management device 2a to 2c Worker 4 Work incidental information 5 Work attribute information

Claims (14)

a work efficiency calculation unit that calculates work efficiency for a past combination of workers from statistics of work analysis data for the combination of workers engaged in the work;
a worker combination calculation unit that ranks the work efficiency levels of the work teams engaged in the work and stores the rankings in a work status database ;
a work efficiency estimation unit that estimates the work efficiency of a work team by combining workers using a statistical method by referring to the past work efficiency levels and work results of each work team stored in the work status database;
A worker combination calculation device comprising: the worker combination calculation unit further includes a worker clustering unit that clusters the workers having similar degrees of change in work efficiency based on the past work efficiency of the workers,
the worker combination calculation unit organizes a work team to be engaged in the work from the workers clustered by the worker clustering unit, and ranks the work efficiency levels of the work teams.
2. The worker combination calculation device according to claim 1 . The worker combination calculation unit further includes a work pattern determination unit that, when efficiency decline in a work pattern of a predetermined time period of the worker is significant, determines a work pattern of a period shorter than the predetermined time period based on data on changes in past work efficiency of the worker .
2. The worker combination calculation device according to claim 1 . the work efficiency calculation unit calculates the duration of the worker's physical movement in the past work, calculates a happiness level based on whether the relationship between the duration and its frequency is close to 1/T fluctuation, and quantifies the work efficiency of the work team engaged in the work based on the calculated happiness level of the worker;
2. The worker combination calculation device according to claim 1 . The worker combination calculation unit defines a combination of workers that results in a dispersion of changes in vital data over time as a result of good work efficiency if the work is one in which failure is not permitted.
2. The worker combination calculation device according to claim 1 . The worker combination calculation unit defines a combination of workers having similar changes in vital data over time as a result of good work efficiency if the work is one in which failure is permitted.
2. The worker combination calculation device according to claim 1 . The worker combination calculation unit ranks the work efficiency levels of the work teams using a mathematical model that performs overall optimization.
2. The worker combination calculation device according to claim 1 . The worker combination calculation unit ranks the work efficiency levels of the work teams using a mathematical model that performs local optimization.
2. The worker combination calculation device according to claim 1 . 2. The worker combination calculation device according to claim 1, wherein the worker combination calculation unit inputs a combination of task attributes, numerical value ranges, and compatibility ranges, calculates an error with respect to task efficiency, which is training data, and ranks the compatibility levels of the work teams using a deep learning model that performs regression calculation so as to minimize this error . the worker combination calculation unit uses attributes of the tasks as input values, weights each input value with a coefficient according to the status of the task, and uses a machine learning algorithm to rank similar tasks from the past, thereby extracting combinations of similar tasks with task efficiency higher than a predetermined value and the task efficiency;
2. The worker combination calculation device according to claim 1 . the worker combination calculation unit uses attributes of the tasks as input values, weights each input value with a coefficient according to the status of the task, and uses a mathematical model to rank similar tasks from the past, thereby extracting combinations of similar tasks with task efficiency higher than a predetermined value and the task efficiency;
2. The worker combination calculation device according to claim 1 . The attributes of the work include any one of the following: the number of workers engaged in the work, the difficulty of the work, the number of days for the work, the working hours per day for the work, the work schedule of the workers engaged in the work, the area environment for the work, and the radiation exposure dose of the workers engaged in the work.
2. The worker combination calculation device according to claim 1 . a work efficiency calculation unit that calculates work efficiency for a past combination of workers from statistics of work analysis data for the combination of workers engaged in the work;
a worker combination calculation unit that ranks the work efficiency levels of the work teams engaged in the work and stores the rankings in a work status database ;
a work efficiency estimation unit that estimates the work efficiency of a work team by combining workers using a statistical method by referring to the past work efficiency levels and work results of each work team stored in the work status database;
A work team management system comprising: a work efficiency calculation unit calculating work efficiency of a past combination of workers from statistics of analysis data of the work for the combination of workers engaged in the work;
a worker combination calculation unit ranking the work efficiency levels of the work teams engaged in the work and storing the rankings in a work status database ;
a work efficiency estimation unit estimating a work efficiency of a work team made up of a combination of workers by a statistical method , by referring to the past work efficiency levels and work results of each work team stored in the work status database;
A worker combination calculation method comprising the steps of:

Images