JP6765449B2 - Work analyzer and work analysis method - Google Patents

Description

The present invention analyzes the presence or absence of work (hereinafter referred to as non-standard work) that deviates from the standard work set as work-efficient work content for the worker, and performs appropriate standard work. It relates to a work analysis device and a work analysis method that can be instructed in.

At production sites such as factories, the contents of standard work are stipulated so that there will be no adverse effects on work productivity maintenance, quality assurance, or worker occupational health. In this case, in order to confirm whether or not non-standard work has occurred, the analyst who checks the work content observes the movement of the worker and determines the trajectory of the worker's movement or the time required for the movement, such as video equipment. There is known a classical IE method that encourages standard work to be performed by measuring with (see, for example, Non-Patent Document 1 below).

In such a conventional technique, a two-handed work analysis that analyzes a worker's two-handed movement, or a method that divides a human movement into basic elements and performs motion research such as fine motion analysis based on the information is introduced. There is. However, in such a conventional motion analysis method, there is a problem that the analyst has to analyze and organize the observation results, which requires a lot of labor for the analyst and takes time to feed back the analysis results.

Therefore, in recent years, a technique for automatically identifying a person's work motion and using it for work analysis by using a motion capture technique using various sensors such as Patent Document 1 and Patent Document 2 below. Is disclosed.

Patent No. 5159263 JP-A-2015-125578

Akihisa Fujita, "New Edition of IE Basics" Kenjosha (published in January 2007)

Here, in the above-mentioned Patent Document 1, the contents of the standard work are registered in advance as a database called "motion dictionary", human movement is detected by a sensor, and the detected value is compared with the above-mentioned "motion dictionary". By doing so, a technique for specifying the operation corresponding to the detected value is disclosed.

However, it is necessary for the analyst to judge whether or not the operation specified in this way deviates from the standard work, and it is good whether or not the actual work deviates from the predetermined standard work. Badness cannot be judged automatically. For this reason, it imposes an excessive burden on the analyst and makes it impossible to make an objective judgment.

Further, Patent Document 2 discloses a technique in which a worker engaged in manufacturing a product inputs a posture to be analyzed and objectively evaluates the quality of the posture by using various sensors. There is.

However, in the configuration described in Cited Document 2, the operator must intentionally select the content of the standard work to be evaluated by himself / herself, and the series of work includes non-standard work. Even in such cases, such behavior may be overlooked and erroneously determined to be performing standard work.

The present invention has been made to solve the above problems, and automatically, objectively, and quickly analyzes the presence or absence of non-standard work without imposing an excessive burden on the analyst who checks the work contents. It is an object of the present invention to provide a work analysis device and a work analysis method that can be instructed to always perform appropriate standard work.

The work analyzer according to the present invention includes a coordinate information acquisition unit that detects the site coordinates of the joint portion of the worker and the existing position coordinates of the worker.
Based on the part coordinates and the existing position coordinates acquired by the coordinate information acquisition unit, the work procedure and movement route associated with the series of work of the worker, and the time required for them are extracted as work content information. Content extraction section and
A work content analysis unit that analyzes the work content performed by the worker based on the work content information obtained by the work content extraction unit and determines whether or not there is non-standard work that deviates from the standard work.
When the work content analysis unit determines that the worker's work is non-standard work, the feedback unit gives an instruction to the worker to return to the standard work.
Criteria for analyzing the work content of the person with better work efficiency by comparing the previous work content information extracted by the work content extraction unit with the current work content information. The learning department that learns to become
A learning content recording unit that records the work content information learned in the learning unit retroactively, and a learning content recording unit.
It is provided with a process selection unit for selecting whether or not to operate the learning unit at the same time as the work content analysis unit.
If it can be determined that the standard work has been established as a result of the worker repeatedly performing the work according to the series of operation procedures in the learning unit, the process selection unit of the learning unit is instructed by the analyst . The learning operation is stopped, and the optimum work content information registered in the learning content recording unit is used as the standard work information as a reference.

Further, the work analysis method according to the present invention includes a coordinate information acquisition step for detecting the site coordinates of the joint portion of the worker and the existing position coordinates of the worker.
Based on the above-mentioned part coordinates and the above-mentioned existing position coordinates acquired in the above-mentioned coordinate information acquisition step, the work of extracting the procedure and movement route of the work operation associated with the series of work of the worker and the time required for them as work content information. Content extraction steps and
A work content analysis step that analyzes the work content performed by the worker based on the work content information obtained in the work content extraction step and determines whether or not there is a non-standard work that deviates from the standard work.
When the work content analysis step determines that the worker's work is non-standard work, a feedback step that gives an instruction to the worker to return to the standard work, and a feedback step.
Criteria for analyzing the work content of the one with better work efficiency by comparing the previous work content information extracted in the above work content extraction step with the current work content information in the above work content analysis step. Learning steps to learn to be
A learning content recording step that records the work content information learned in the learning step retroactively, and
It is provided with a process selection step for selecting whether or not to operate the above learning steps at the same time as the above work content analysis step.
In the learning step, if it can be determined that the standard work has been established as a result of the worker repeatedly performing the work according to the series of operation procedures, the process selection step of the learning step is performed according to the instruction of the analyst . The learning operation is stopped, and the optimum work content information registered in the learning content recording step is used as the standard work information as a reference.

According to the present invention, the presence or absence of non-standard work is automatically, objectively, and quickly analyzed without imposing an excessive burden on the analyst who checks the work contents performed by the worker, and always appropriate standard work. Can be instructed to do. Therefore, it is possible to ensure work safety, improve work efficiency, and guarantee product quality.

It is a block diagram which shows the whole of the work analyzer according to Embodiment 1 of this invention. It is a block diagram which shows the configuration example of the hardware of the personal computer (PC) as the analysis apparatus of Embodiment 1 of this invention. It is explanatory drawing of the coordinate system for setting a coordinate position with the installation position of the motion capture of the work analysis apparatus of Embodiment 1 of this invention as an origin. It is a functional block diagram of the work analyzer of Embodiment 1 of this invention. It is explanatory drawing which shows the joint part of the worker which is the acquisition target in the part coordinate information acquisition part of the work analysis apparatus of Embodiment 1 of this invention. It is explanatory drawing which shows an example of the part coordinate information acquired about the joint part of a worker by the part coordinate information acquisition part of the work analysis apparatus of Embodiment 1 of this invention. It is a graph which shows an example of the movement path of the worker extracted by the movement information extraction part of the work analysis apparatus of Embodiment 1 of this invention. It is a flowchart for demonstrating the processing content which extracts the work operation procedure of a worker in the operation procedure information extraction part of the work analysis apparatus of Embodiment 1 of this invention. It is explanatory drawing for extracting the "walking" movement as the work movement of an operator in the operation procedure information extraction unit of Embodiment 1 of this invention. It is explanatory drawing for extracting the operation of "turning around" as the work operation of an operator in the operation procedure information extraction unit of Embodiment 1 of this invention. It is explanatory drawing for extracting the "up and down" operation as the work operation of an operator in the operation procedure information extraction unit of Embodiment 1 of this invention. It is explanatory drawing for extracting the "crouching" motion as the work motion of an operator in the operation procedure information extraction unit of Embodiment 1 of this invention. It is explanatory drawing for extracting the "stretching" movement as the work movement of an operator in the operation procedure information extraction unit of Embodiment 1 of this invention. It is explanatory drawing for extracting the "bending" movement as the work movement of an operator in the operation procedure information extraction unit of Embodiment 1 of this invention. It is a flowchart explaining the learning processing procedure by the learning part of Embodiment 1 of this invention. It is a flowchart which shows the processing procedure for analyzing the work content by the work content analysis part of Embodiment 1 of this invention and determining whether it is a nonstandard work. It is explanatory drawing of the feedback processing content by the feedback part of Embodiment 1 of this invention. It is a block diagram which shows an example of the feedback apparatus which urges an operator to return to a standard work by an instruction from the feedback part of Embodiment 1 of this invention. It is explanatory drawing which provides the operation explanation in the feedback device at the time of urging the return to the standard work by the feedback device shown in FIG. It is a block diagram which shows another example of the feedback apparatus which urges an operator to return to a standard work by an instruction from the feedback part of Embodiment 1 of this invention. It is a functional block diagram of the work analyzer of Embodiment 2 of this invention. It is a block diagram which shows the modification of the work analyzer of Embodiment 1 or 2 of this invention.

Embodiment 1.
FIG. 1 is a configuration diagram showing the entire work analyzer according to the first embodiment of the present invention.

The work analysis device according to the first embodiment is a motion capture 1 that acquires position information on the three-dimensional space (3Dspace) of the worker M, and a work of the worker M based on the position information obtained by the motion capture 1. A personal computer (hereinafter referred to as a PC) 2 as an example of an analysis device for analyzing the contents is provided, and the motion capture 1 and the PC 2 are connected to each other via a USB cable 3. A power supply 1A is connected to the motion capture 1.

FIG. 2 is a block diagram showing a configuration example of the hardware of PC2 as an analytical instrument. The PC 2 includes a processor 200, a storage device 201, and an input / output unit 202. Although the storage device 201 is not shown, it includes a volatile storage device such as a random access memory and a non-volatile auxiliary storage device such as a flash memory. A hard disk auxiliary storage device may be provided instead of the flash memory or the like. The processor 200 executes the program input from the storage device 201. In this case, a program is input from the auxiliary storage device to the processor 200 via the volatile storage device. Further, the processor 200 may output data such as an arithmetic processing result to the volatile storage device of the storage device 201, or may store the data in the auxiliary storage device via the volatile storage device. Data from the motion capture 1 is input to the input / output unit 202 via the USB cable 3.

As the motion capture 1 in the first embodiment, for example, by using Kinect (registered trademark) of Microsoft Corporation, the position information on the three-dimensional space (3D space) is acquired without attaching a sensor marker or the like to the worker M. be able to. FIG. 3 is a diagram showing each direction of x, y, and z with the installation position of the motion capture 1 as the observation origin, where x means the horizontal direction, y means the vertical direction, and z means the position in the depth direction. ..

The motion capture 1 is not limited to such a configuration, and various detection methods such as a configuration in which the operator M is used with a sensor marker or a configuration in which heat is detected by an infrared camera are used. It is possible.

FIG. 4 is a functional block diagram of the work analyzer according to the first embodiment.
This work analyzer is composed of a part coordinate information acquisition unit 5 for acquiring the part coordinate information of the joint part of the worker, an existence position coordinate information acquisition unit 6 for acquiring the existence position coordinate information of the worker, and a part coordinate information acquisition unit 5. Recorded in the part coordinate information recording unit 7 that records the acquired part coordinate information, the existence position coordinate information recording unit 8 that records the existence position coordinate information acquired by the existence position coordinate information acquisition unit 6, and the part coordinate information recording unit 7. The existence position recorded in the operation procedure information extraction unit 9 and the existence position coordinate information recording unit 8 for extracting a series of procedures of the worker's work operation and the time required for it as the operation procedure information based on the part coordinate information. The movement information extraction unit 10 that extracts the worker's movement route and the time required for it as movement information based on the coordinate information, and the worker's operation procedure information and movement information extraction unit 10 that are extracted by the operation procedure information extraction unit 9 Learning unit 11 that learns so as to obtain the optimum work content without waste based on the movement information to be performed, learning content recording unit 15 that records the learning result of the work content obtained by the learning unit 11, and the above operation procedure. Based on the operation procedure information extracted by the information extraction unit 9, the movement information extracted by the movement information extraction unit 10, and the information recorded by the learning content recording unit 15, the work content performed by the operator is analyzed and standardized. Work other than standard work is performed in the work content analysis unit 12 that determines the presence or absence of external work, the process selection unit 13 that selects whether to operate the learning unit 11 at the same time as the work content analysis unit 12, and the work content analysis unit 12. It is provided with a feedback unit 14 that gives an instruction to urge the operator to return to the standard work when it is determined that the work has been performed.

Here, the above-mentioned part coordinate information acquisition unit 5 and the existence position coordinate information acquisition unit 6 correspond to the motion capture 1. Further, the above-mentioned operation procedure information extraction unit 9, movement information extraction unit 10, learning unit 11, work content analysis unit 12, process selection unit 13, and feedback unit 14 install a predetermined program in the storage device 201 of the PC2, and install a predetermined program. It operates by executing the program by the processor 200. Further, the site coordinate information recording unit 7, the existence position coordinate information recording unit 8, and the learning content recording unit 15 correspond to the storage device 201 included in the PC 2.

The above-mentioned part coordinate information acquisition unit 5 and the existence position coordinate information acquisition unit 6 correspond to the coordinate information acquisition unit in the claim range (claim 1), and the above-mentioned operation procedure information extraction unit 9 and movement information. The extraction unit 10 corresponds to the work content extraction unit in the claims (claim 1), respectively.

In this way, by configuring the part coordinate information acquisition unit 5 and the existence position coordinate information acquisition unit 6 by applying the commercially available motion capture 1, it is possible to make a flexible device configuration. Further, although PC2 is used for each other part, the present invention is not limited to this, and a tablet terminal or a smart phone having the same function as PC2 may be used, and the present invention is not limited to the personal computer.

Further, in the first embodiment, the part coordinate information acquisition unit 5 and the existence position coordinate information acquisition unit 6 are composed of one motion capture 1, but in general, the part coordinate information is of the human body. Since it is a range of motion, it is necessary to record with an accuracy of 50 cm or less, whereas the existing position coordinate information is a work area (several meters square), so it may be necessary to record on a scale of 1 m or more. .. Therefore, motion capture 1 that can record the shape of the body in detail is used to acquire the site coordinate information, and a GPS (Global Positioning System) device, a beacon, a camera, an image recognition device, etc. are used to acquire the existence position coordinate information. , It is also possible to use different devices.

Further, when the moving range of the worker M is wide and exceeds the detection limit of the motion capture 1, the detection limit is exceeded by having the worker M carry the beacon and continue the detection of the existing position coordinates. It is possible to estimate what was being done in the event of a disturbance.

Next, the specific processing operation of each part of the work analyzer shown in FIG. 4 will be described.

The part coordinate information acquisition unit 5 composed of the motion capture 1 acquires the part coordinate information of each part of the human body of the worker. FIG. 5 shows the joint parts of the worker to be acquired by the part coordinate information acquisition unit 5, from the head (HEAD, hereinafter referred to as H) to the right toe (FOOT_RIGHT, hereinafter referred to as FR) and the left toe (hereinafter referred to as FR). Coordinate information of a total of 20 joint parts up to FOOT_LEFT (hereinafter referred to as FL) is to be acquired.

FIG. 6 is an example of the part coordinate information of the worker recorded by the part coordinate information acquisition unit 5. The site coordinate information includes the date and time when the data was acquired, the name of the joint site, and the coordinates (x, y, z) of the joint site. Then, this part coordinate information is recorded in the part coordinate information recording unit 7.

In addition, the existence position coordinate information acquisition unit 6 acquires the existence position coordinate information of the worker. In this case, for example, the position of the head H of the human body in the three-dimensional space is used as the existence position coordinate information. This existence position coordinate information is recorded in the existence position coordinate information recording unit 8 together with the date and time information synchronized with the part coordinate information.

The movement information extraction unit 10 extracts the movement route of the worker and the time required for the movement information as movement information based on the existence position coordinate information recorded in the existence position coordinate information recording unit 8. FIG. 7 shows an example of the movement information of the worker extracted by the movement information extraction unit 10.

Further, the operation procedure information extraction unit 9 extracts a series of procedures of the worker's work operation and the time required for the procedure as the operation procedure information based on the part coordinate information recorded in the part coordinate information recording unit 7. The processing content of the operation procedure information extraction unit 9 will be described with reference to the flowchart shown in FIG. In the following, reference numeral S means each processing step.

For the operation procedure information extraction unit 9, the analyst who checks the work content performs the worker's operation from among the various movements of the worker (for example, various movements such as walking, turning, bending, climbing and descending). Then, the expected operation (hereinafter referred to as the evaluation target operation) is set in advance (step S11). For example, when working on a flat floor, the walking motion is an evaluation target motion, but the work motion such as going up and down stairs is excluded from the evaluation target motion.

Then, the operation procedure information extraction unit 9 takes in the part coordinate information recorded in the part coordinate information recording unit 7 (step S12), and collates with the rule for extracting the work operation described later based on the part coordinate information. Then, it is determined what kind of operation the worker has performed (step S13). This is performed for all the evaluation target operations preset in step S11, and if the operation extraction for all the evaluation target operations is completed (YES in step S14), the information indicating a series of procedures of those evaluation target operations (YES). Time-series information of the operation to be evaluated) and time information required for this series of operation procedures are extracted, and the information is output as operation procedure information (step S15).

Next, in the rule collation step (step S13) shown in FIG. 8, the details of the processing contents when the evaluation target operation is extracted based on the part coordinate information will be described.

Here, from among various work operations, six typical operations will be described as evaluation target operations, but the operation is not necessarily limited to these six operations. Both operations are collated using the amount of change between the part coordinate information at the immediately preceding time.

(1) Walking FIG. 9 shows an example of a matching rule for a “walking” motion.
In order to collate whether or not it is a "walking" motion, the movement distance in the horizontal direction (x-z coordinates) of the central part of the shoulder (Shoulder_Center, hereinafter referred to as SC) is used from the input part coordinate information. If the x-coordinate of the shoulder center SC before movement is SCx and the z-coordinate is SCz, and the x-coordinate of the shoulder center SC'after movement is SC'x and the z-coordinate is SC'z, then the shoulder center SC When the moving distance is larger than the worker's stride length of 0.5 m, it can be considered that the "walking" operation has been performed. In the following equation (1), the left side means the moving distance of the shoulder central portion SC, and the right side means the stride (judgment threshold) of the worker.

In the example of the formula (1), the stride as the judgment threshold of the "walking" motion is set to 0.5 m, but this value is not limited to 0.5 m, and the judgment of the analyst who checks the work content. Can be set freely by.

(2) Turning around FIG. 10 shows an example of a collation rule for a “turning around” operation. In order to collate whether or not the operation is "turning around", from the input part coordinate information, SLx which is the x coordinate of the left shoulder (Shoulder_Left, hereinafter referred to as SL) and SLz which is the z coordinate, or the right shoulder (Shoulder_Right). , Hereinafter referred to as SR), the angle θ formed by the straight line connecting SRx, which is the x-coordinate, and SRz, which is the z-coordinate, and the observation origins x0 and z0 is calculated. When the angle difference of θ is 90 ° or more, it can be considered that the “turning around” operation has been performed.

In the example of equation (2), the angle difference of the vector connecting the center of the shoulder and the left shoulder before and after the movement of the "turning around" motion is set to 90 ° (π / 2) as the judgment threshold value, but this value is 90. It is not limited to ° (π / 2), but can be set freely at the discretion of the analyst who checks the work content.

(3) Ascending / descending FIG. 11 shows an example of a collation rule for the “ascending / descending” operation.
To check whether or not it is a "climbing" operation, the vertical movement distance of both the left foot tip (hereinafter referred to as FL) and the right foot tip (hereinafter referred to as FR) is calculated from the input part coordinate information. Use. The smaller (lower) value of FLy, which is the vertical y-coordinate of the left foot FL before movement, and FRy, which is the vertical y-coordinate of the right foot FR, is defined as FL / Ry. , The smaller (lower) value of FL'y, which is the vertical y-coordinate of the left foot FL after movement, and FR'y, which is the vertical y-coordinate of the right foot FR, is FL'/ R'. If y, it can be considered that the "up and down" operation is performed when the difference between the two before and after the movement (that is, the step) is large. In the following equation (3), the left side means the amount of change in the position of the right foot tip FR or the left foot tip FL, whichever is lower, that is, the step, and the right side means the ascending / descending judgment threshold. ing.

In the example of equation (3), the step before and after the movement of the "up and down" operation is 0.1 m, but this value is not limited to 0.1 m, and the analyst's judgment to check the work content. Can be set freely by.

(4) Crouching FIG. 12 shows an example of a collation rule for a “crouching” operation.
In order to collate whether or not it is a "crouching" motion, the HCy, which is the vertical y-coordinate of the buttocks (Hip_Center, hereinafter referred to as HC), and the "climbing" motion are determined from the input part coordinate information. The degree of proximity to the above FL / Ry used at times is used. In the "squatting" motion, the buttock HC approaches the floor surface, and as a result, the vertical distance difference between the left and right toes is reduced. Therefore, when the difference between the buttock HC and FL / Ry is small, the "squatting" motion is performed. It can be considered that it was done. In the equation (4), the left side means the distance between the buttocks HC and the left and right toes, and the right side means the judgment threshold value of the "squatting" motion.

In the example of equation (4), the judgment threshold value for the "squatting" motion is set to 0.3 m, but this value is not limited to 0.3 m, and can be freely determined by the analyst who checks the work content. Can be set.

(5) Stretching Back FIG. 13 shows an example of a matching rule for the “stretching back” operation.
To collate whether or not it is a "stretching" motion, from the input part coordinate information, ALY and the right heel (Ankle_Right, hereinafter AR), which are the y-coordinates of the left heel (Ankle_Left, hereinafter referred to as AL) in the vertical direction. The smaller (lower) value of ARY, which is the y-coordinate in the vertical direction (notation), is AL / Ry, and the distance between the left and right toes and FL / Ry is used. In the stretched posture, the heels are separated from the floor surface, and as a result, the vertical distance between the left and right toes and the heels increases. Therefore, when the difference between AL / Ry and FL / Ry is large, the “stretched” motion is performed. It can be regarded as being broken. In the following equation (5), the left side means the distance difference in the vertical direction between the left and right toes and the left and right heels, and the right side means the determination threshold value of the back extension posture.

In the example of equation (5), the judgment threshold value for the "stretching" motion is set to 0.05 m, but this value is not limited to 0.05 m, and can be freely determined by the analyst who checks the work content. Can be set.

(6) Bending FIG. 14 shows an example of a matching rule for a “bending” operation.
To collate whether or not it is a "bending" motion, the distance between Hy, which is the vertical y-coordinate of the head H, and HCy, which is the vertical y-coordinate of the buttock HC, is calculated from the input part coordinate information. Use. In the crouching posture, the head H approaches the floor surface, and as a result, the vertical distance from the buttock HC approaches. Therefore, when the difference between Hy and HCy is small, it is considered that the "bending" motion is performed. be able to. In the following equation (6), the left side means the distance difference between the head H and the buttocks HC in the vertical direction, and the right side means the judgment threshold value of the bending posture.

In the example of equation (6), the threshold value for determining the "bending" motion is 0.4 m, but this value is not limited to 0.4 m and can be freely determined by the analyst who checks the work content. Can be set.

In order to establish the optimum work content as standard work for the worker to proceed with the work efficiently without waste or waste in the work operation of the worker, the worker follows a series of operation procedures by trial and error. It is preferable to repeat the work and learn so as to obtain the optimum work content. Then, as a result of performing the work according to the series of operation procedures, if the time required for it is the shortest and there is no waste or waste in the work operation, the work according to the series of optimum operation procedures is set as the standard work. Set.

Therefore, in the first embodiment, there is provided a learning unit 11 for establishing a work that has no waste or waste in the work content according to a series of operation procedures and has the shortest time required for the work as a standard work. ..

In the initial stage, which has not yet been established as a standard work, the process selection unit 13 selects the learning unit 11 to operate together with the work content analysis unit 12. The learning operation by the learning unit 11 will be described with reference to the flowchart shown in FIG.

As described above, when the operator performs a series of operations, the operation procedure information extraction unit 9 performs operation procedure information (that is, each evaluation target operation of the operator and a series of procedures of each of these evaluation target operations, and Since the movement information (that is, the movement route of the worker and the time information required for it) is extracted by the movement information extraction unit 10 and the time information required for this series of operation procedures), the learning unit 11 These operation procedure information and movement information (hereinafter, both are collectively referred to as work content information) are taken in (step S21, step S22).

Next, the learning unit 11 compares the work content of this time (nth time) with the work content of the previous time (n-1th time) based on these work content information (operation procedure information and movement information) (step). S23), when the work content of this time is less wasteful or wasteful than the previous work and the time required for the work is short, the work content of this time is adopted as the optimum work (step S24). Then, the optimum work content information in step S24 is recorded in the learning content recording unit 15 as a reference when the work content analysis unit 12 analyzes the work content (step S26). On the other hand, when the previous work content is less wasteful or wasteful than this time and the time required for the work is short, the previous work content is adopted as the optimum work (step S25). Then, the optimum work content information in step S25 is recorded in the learning content recording unit 15 as a reference when the work content analysis unit 12 analyzes the work content (step S26). As a result, even if the work content is changed due to, for example, a change in the production line or an improvement in the work site, the evaluation can be performed based on the work that always follows the optimum operation procedure.

In addition, in the above-mentioned learning unit 11, when it can be judged that it has been established as a standard work as a result of the worker repeating the work according to a series of operation procedures by trial and error, according to the instruction of the analyst. The learning operation of the learning unit 11 is stopped by the processing selection unit 13, and the optimum work content information (operation procedure information and movement information) registered in the learning content recording unit 15 is used as the standard work information as the standard work content described below. The analysis unit 12 is used as a comparison target when determining the work content of the worker.

Refer to the flowchart shown in FIG. 16 for the processing operation when the work content analysis unit 12 analyzes the work content performed by the worker and determines whether or not there is non-standard work, that is, whether or not it matches the standard work. I will explain.

First, the work content analysis unit 12 first obtains operation procedure information regarding the worker obtained by the operation procedure information extraction unit 9 (that is, a series of procedures for each evaluation target operation of the worker, and time information required for this series of operation procedures. ) And the worker's movement information (worker's movement route and time information required for the movement information) obtained by the movement information extraction unit 10 are taken in as work content information (steps S31 and S32).

Next, the work content analysis unit 12 compares the current work content obtained by the work of the worker with the optimum work content recorded in the learning content recording unit 15 (step S33), and both of them. It is determined whether or not the work contents match (step S34). If the current work content and the optimum work content match in step S34, it is further determined whether or not the time required for a series of work (operation procedure or movement) matches (step S35). If the times required for a series of operations match, it is determined that the operations are standard operations (step S36).

On the other hand, in step S34, if the current work content obtained by the worker's work and the optimum work content recorded in the learning content recording unit 15 do not match, it does not match the standard work, that is, the standard. It is determined that the work is outside work (step S37). Further, if the times required for a series of operations (operation procedure and movement) do not match in step S35, it is determined that the work pace is disturbed (step S38). The determination results are sent to the feedback unit 14.

The feedback unit 14 gives an instruction to the worker to comply with the standard work according to the determination result by the work content analysis unit 12. That is, as shown in FIG. 17, when it is determined that the work matches the standard work, no instruction is given to the worker. If it is determined that the pace is disturbed, an instruction is given to encourage the person to return to the standard work pace. If it is determined that the work does not match the standard work (non-standard work), a warning to stop the work is generated.

Next, a specific example of feedback processing for the worker by the feedback unit 14 will be described.

FIG. 18 is an example of a feedback device that prompts the worker M to return to the standard work by an instruction from the feedback unit 14, and shows a feedback device using sound.
A speaker 16 having directivity toward the worker M is installed at the work site, and the speaker 16 is connected to the feedback unit 14 via the scale generator 17. Then, when the standard work is normally performed, the speaker 16 generates a scale sound of a fixed cycle synchronized with one cycle of the standard work like a metronome so that the worker M is aware of the work pace. To.

If it matches the standard work, there is no special feedback to the worker M, but if it is determined that the pace is disturbed, as shown in FIG. 19, according to the instruction from the feedback unit 14. Change the scale speed to encourage a return to the standard work pace. That is, by gradually increasing the scale speed in the case of lagging the pace and gradually decreasing the scale speed in the case of advancing the pace, the worker M is naturally encouraged to return to the standard work pace without impacting. .. If it is determined that the work does not match the standard work (non-standard work), a warning to stop the work is generated.

FIG. 20 is another example of a feedback device that urges the worker M to return to the standard work by an instruction from the feedback unit 14, and uses a wearable terminal 18 equipped with a vibration generator and an alarm. It shows.

When the moving range of the worker M is wide, it becomes difficult to fix and install the equipment such as the directional speaker in one place. Therefore, the worker M is made to wear the wearable terminal 18, and the feedback unit 14 sends an instruction to the wearable terminal 18 by wireless communication RC via the work instruction generator 19. Then, when the standard work is normally performed, the vibration generator of the wearable terminal 18 generates vibration synchronized with one cycle of the standard work to make the worker M aware of the work pace.

If it matches the standard work, there is no special feedback to the worker M, but if it is determined that the pace is disturbed, the wearable terminal 18 is provided with vibration according to the instruction from the feedback unit 14. Change the vibration speed of the device to encourage the return to the standard work pace. That is, by gradually increasing the vibration speed in the case of a pace delay and gradually decreasing the vibration speed in the case of a pace advance, the worker M is naturally encouraged to return to the standard work pace without giving an impact. .. Further, when it is determined that the work does not match the standard work (non-standard work), the alarm device provided in the wearable terminal 18 is activated to generate a warning sound for stopping the work.

Embodiment 2.
In the second embodiment, in addition to the configuration of the first embodiment described above, the connection portions 51 and 52 shown in FIG. 4 are replaced with the connection portions 51A and 52A shown in FIG. 21 as network connections. To. Then, by arranging the learning unit 11 and the learning content recording unit 15 on the cloud, it is possible to perform work analysis in a remote location. Other configurations are the same as in the case of the first embodiment.

With such a configuration, the learning unit 11 and the learning content recording unit 15 are stored as big data related to work information. As a result, it becomes possible to immediately apply the learned standard work to a new factory or production line.

The following modifications and applications can be considered for the above embodiments 1 and 2.
For example, in the above-described first and second embodiments, the site coordinate information acquisition unit 5 and the existence position coordinate information acquisition unit 6 are configured by one motion capture 1, but the present invention is not limited to this, and is shown in FIG. As described above, it is also possible to install a plurality of motion captures 1a and 1b corresponding to the plurality of workers M1 and M2, and to configure the site coordinate information acquisition unit 5 and the existence position coordinate information acquisition unit 6 by them. In this case, a plurality of motion captures 1a and 1b are connected to one PC2 via the LAN 20. The power supplies 1aA and 1bA are connected to the motion captures 1a and 1b. In this way, by using the plurality of motion captures 1a and 1b, it is possible to efficiently perform the work analysis of the workers M1 and M2 who are working at remote locations.
The number of motion captures 1a and 1b is not limited to such two, and three or more may be provided.

Further, in the first and second embodiments, the case where the analysis target is whether or not the work performed mainly by the worker engaged in the manufacture of the product in the factory deviates from the standard work has been described. Not limited to such fields, it can be effectively applied even when it is judged whether or not the work content is appropriate for improving the work environment of workers and occupational health by analyzing routine work in the non-manufacturing industry. ..

Further, the present invention is not limited to the configurations of the above-described first and second embodiments, and a part of the configurations of the first and second embodiments is modified without departing from the spirit of the present invention. Alternatively, the configuration can be omitted.

1,1a, 1b Motion capture, 2 Personal computer (PC), M, M1, M2 Worker, 5 Part coordinate information acquisition unit, 6 Existence position coordinate information acquisition unit, 7 Part coordinate information recording unit, 8 Existence position coordinate information Recording unit, 9 operation procedure information extraction unit, 10 movement information extraction unit, 11 learning unit, 12 work content analysis unit, 13 processing selection unit, 14 feedback unit, 15 learning content recording unit.

Claims (6)

A coordinate information acquisition unit that detects the part coordinates of the joint part of the worker and the existing position coordinates of the worker,
Based on the part coordinates and the existing position coordinates acquired by the coordinate information acquisition unit, the work procedure and movement route associated with the series of work of the worker, and the time required for them are extracted as work content information. Content extraction section and
A work content analysis unit that analyzes the work content performed by the worker based on the work content information obtained by the work content extraction unit and determines whether or not there is non-standard work that deviates from the standard work.
When the work content analysis unit determines that the worker's work is non-standard work, the feedback unit gives an instruction to the worker to return to the standard work.
Criteria for analyzing the work content of the person with better work efficiency by comparing the previous work content information extracted by the work content extraction unit with the current work content information. The learning department that learns to become
A learning content recording unit that records the work content information learned in the learning unit retroactively, and a learning content recording unit.
It is provided with a process selection unit for selecting whether or not to operate the learning unit at the same time as the work content analysis unit.
If it can be determined that the standard work has been established as a result of the worker repeatedly performing the work according to the series of operation procedures in the learning unit, the process selection unit of the learning unit is instructed by the analyst . A work analyzer in which the learning operation is stopped and the optimum work content information registered in the learning content recording unit is used as standard work information as a reference. The work analyzer according to claim 1, wherein the learning unit and the learning content recording unit are connected to a network, and the learning unit and the learning content recording unit are arranged on a cloud. Claim 1 or claim 2 that the coordinate information acquisition unit includes a site coordinate information acquisition unit that detects the site coordinates of the joint portion of the worker and an existence position coordinate information acquisition unit that detects the existence position coordinates of the worker. The work analyzer described in. The work content extraction unit includes an operation procedure information extraction unit that converts the part coordinates of the worker detected by the part coordinate information acquisition unit into operation procedure information, and the existence position coordinate information acquisition unit that detects the existence. The work analyzer according to claim 3, further comprising a movement information extraction unit that converts position coordinate information into movement information of an operator. A coordinate information acquisition step for detecting the part coordinates of the joint part of the worker and the existing position coordinates of the worker,
Based on the above-mentioned part coordinates and the above-mentioned existing position coordinates acquired in the above-mentioned coordinate information acquisition step, the work of extracting the procedure and movement route of the work operation associated with the series of work of the worker and the time required for them as work content information. Content extraction steps and
A work content analysis step that analyzes the work content performed by the worker based on the work content information obtained in the work content extraction step and determines whether or not there is a non-standard work that deviates from the standard work.
When the work content analysis step determines that the worker's work is non-standard work, a feedback step that gives an instruction to the worker to return to the standard work, and a feedback step.
Criteria for analyzing the work content of the one with better work efficiency by comparing the previous work content information extracted in the above work content extraction step with the current work content information in the above work content analysis step. Learning steps to learn to be
A learning content recording step that records the work content information learned in the learning step retroactively, and
It is provided with a process selection step for selecting whether or not to operate the above learning steps at the same time as the above work content analysis step.
In the learning step, if it can be determined that the standard work has been established as a result of the worker repeatedly performing the work according to the series of operation procedures, the process selection step of the learning step is performed according to the instruction of the analyst . A work analysis method in which the learning operation is stopped and the optimum work content information registered in the above learning content recording step is used as the standard work information as a reference. The work analysis method according to claim 5, wherein the worker is in possession of a beacon, and the detection of the worker's existing position coordinates is continued based on the information from the beacon in the coordinate information acquisition step.

Images