JP2005134536A - Work training support system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system for self-sustainedly and efficiently carrying out work training. <P>SOLUTION: The operator receiving the guidance of the work to be carried out by display and speech receives the pseudo-work training using an artificial sense for the real in front of a large-sized screen 2a on which a virtual work region and the operator's model are three-dimensionally displayed, and learns the work. Simultaneously whether the work is right or wrong and the skill level of the work are fed back by speech or display to the operator who is evaluated. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a system for supporting work training of various kinds of work in a production line and the like, and more particularly, to a work training support system for training a work in a virtual work area using a computer.

In general, companies are required to shorten the preparation period from the design of the assembly process to mass production in order to cope with time competition and multi-product small-lot production. In order to cope with this problem, a system that supports design of an assembly process using artificial reality has been developed (for example, see Patent Document 1).
JP-A-9-16550

  However, the conventional system mainly focuses on the design support of the assembly process, so that the worker himself can perform the training independently so that the efficient work training and the effect of the work training can be correctly evaluated. It is hoped that this system will be developed.

  The present invention has been made in view of the above-described points, and allows the user to experience the work in a simulated manner and efficiently perform the work training, and to evaluate the effect of the work training. However, the objective is to be able to perform work training independently while confirming the level of work acquisition.

  In order to achieve the above object, the present invention is configured as follows.

  That is, the present invention is a work training support system for supporting work training for an operator, and displays a virtual work area and a work guidance means for guiding work contents in the virtual work area and a virtual work area. And a work training means for detecting the movement of the worker and displaying it in the virtual work area to perform work training, and a work evaluation means for evaluating the result of the work training by the work training means. Evaluates the correctness of work and the degree of work acquisition.

  Here, the virtual work area is not an actual work area, but a virtual work area imitating an actual work area in order to give guidance or training to the worker. This virtual work area is a 3D-CAD. The virtual work area is preferably displayed by stereoscopic display, that is, three-dimensional graphics, at least in the work training means.

  Guidance of work contents means guidance of work contents to be performed in the virtual work area. For example, display corresponding to work contents in the virtual work area, for example, display of characters or animations explaining the work contents, etc. It means performing or giving guidance by outputting work contents by voice.

  The movement of the worker refers to the movement of the worker's body, for example, movements of the left and right hands, the back, the waist, the head, the elbow, and the like.

  According to the present invention, a worker who receives training receives guidance on the work content in the virtual work area displayed by the work guidance means, and the work training means displays a pseudo-work in the virtual work area in which the worker's own movement is displayed. Since the work can be mastered through proper work training, the correctness of the work and the level of mastery of the work (proficiency level) are evaluated. Can be performed.

  In one embodiment of the present invention, the work guidance means displays a position corresponding to the work content in the virtual work area, and the work training means includes the virtual work area, the worker's model. The position corresponding to the work content is displayed so as to be stereoscopically viewed, and the work evaluation means evaluates the correctness of the work and the degree of work acquisition based on the detected movement of the worker, and evaluates The result is notified to the operator by at least one of display and voice.

  Here, the position corresponding to the work content refers to, for example, the position where the part is taken, the position where the work is performed, and the part according to the content of various work such as assembly, soldering, and board division. This refers to the position where you place.

  The position corresponding to the work content may be displayed with characters or marks, or may be displayed with a specific object such as a red sphere or a blue sphere, and the movement locus from the start position of the work may be displayed. You may display together.

  According to this embodiment, the worker who receives training receives guidance on the work position in the virtual work area by the work guidance means, and while viewing the model of the worker in the virtual work area displayed in three dimensions by the work training means. It is possible to learn a work position and the like by performing a pseudo work.

  In a preferred embodiment of the present invention, the work instruction means outputs the work content by voice, and the work training means recognizes a worker's voice corresponding to the work content, The work evaluation means evaluates the correctness of the work based on the recognized voice.

  According to this embodiment, a worker who receives training receives guidance of work contents by display and voice by the work guidance means, and when performing pseudo work by the work training means, the work contents instructed by voice. , The work evaluation means can determine the correctness of the work based on the recognized voice of the worker.

  In another embodiment of the present invention, the work instruction means includes a display unit that displays the virtual work area, and a voice output unit that outputs work contents by voice, and the work training means includes the virtual work area. A stereoscopic image display unit that displays the region and the worker's model in a stereoscopic manner; a motion detection unit that detects the movement of the worker; and a voice recognition unit that inputs and recognizes the worker's voice. Yes.

  Here, for example, a motion capture is preferable as the motion detection unit that detects the motion of the worker.

  According to this embodiment, the worker who receives training receives guidance of work contents by voice from the voice output unit while looking at the virtual work area displayed on the display unit of the work guidance unit, and the work training unit Training can be carried out while performing pseudo work while observing the worker's model corresponding to the movement of the worker displayed on the video display unit, and speaking and confirming the work content.

  In still another embodiment of the present invention, the work evaluation means evaluates the correctness of the work and the degree of work acquisition based on the movement of the operator's hand to the position corresponding to the work content. .

  According to this embodiment, based on the movement of the operator's hand to the work position, for example, whether or not the hand is carried to the correct work position, whether or not the hand is carried smoothly, etc. The degree of correctness and mastery of work can be evaluated.

  In one embodiment of the present invention, the work evaluation means evaluates the degree of work acquisition based on a predetermined standard operation time and the detected operation time of the worker.

  According to this embodiment, the degree of work acquisition can be evaluated based on a standard work time that is predetermined according to the work content and the operation time of the worker.

  In another embodiment of the present invention, the work evaluation means determines the degree of work acquisition based on at least one of the operator's ambiguity and the smoothness of the operation to the position corresponding to the work content. It is something to evaluate.

  According to this embodiment, the worker's movement at the work position or the smoothness of the movement is grasped as, for example, hand movement shaking (meandering) or movement locus, and based on this, the degree of work acquisition is evaluated. can do.

  As described above, according to the present invention, a worker who receives training receives guidance on the work content in the virtual work area displayed by the work guidance means, and the virtual work in which the movement of the worker is displayed by the work training means. It is possible to acquire work by receiving pseudo work training in the area, and the correctness of the work and the degree of work acquisition in this work training are evaluated. Can be performed.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a schematic configuration diagram of a work training support system according to an embodiment of the present invention.

  The work training support system according to this embodiment displays a virtual work area to instruct (educate) work contents to the worker, and then causes the worker to experience the work in the virtual work area in a pseudo manner. A virtual work training is performed, and the result of the work training is evaluated as a degree of work acquisition, a so-called work proficiency level, etc., and enables the worker's own work training.

  For this reason, the work training support system according to the present embodiment stereoscopically displays a virtual work area on a large screen of a polarized stereoscopic three-dimensional VR (Virtual Reality) display device 2 based on 3D data in the database 1 or the like. The VR control personal computer 3 that can be displayed and the voice control personal computer that synthesizes the voice based on the data in the voice database 4 and outputs the synthesized voice through the speaker 5 and recognizes the voice of the worker inputted through the microphone 6. 7 and a motion control personal computer that transfers data corresponding to the worker's movement captured by the motion detection device 8 to the VR control personal computer 3 based on the data in the database 24 and displays the movement of the worker in the virtual work area. 9 and. The motion detection device 8 detects a worker's motion at a three-dimensional spatial coordinate position, and includes, for example, motion capture (registered trademark).

  The database 1 stores 3D data and work content data of a work area in a production line designed by 3D-CAD, and the VR control personal computer 3 stores the data on the large screen of the VR display device 2 based on such data. 3D space image display is performed to display a virtual work area and a worker's model, in this embodiment, a worker's hand model. Further, this VR control personal computer 3 has a function as a work evaluation means, and evaluates the result of training of the worker. As described later, the correctness of work and the proficiency level of work (work mastery) Grade).

  As shown in FIG. 2, the motion detection device 8 of this embodiment is a virtual work area that is displayed in three dimensions on the screen 2a in front of the large screen 2a of the VR display device 2 as a three-dimensional image display device. The magnetic sensor 11 attached to the left and right hands of the worker 10 who performs work training, the magnetic field generator 12, and the control box 13 are magnetic devices that digitize human movements.

  The motion control personal computer 9 converts the data from the motion detection device 8 into 3D coordinate value information and angle information based on the human motion data in the database 24, and gives it to the VR control personal computer 3.

  Hereinafter, work training support by the system of this embodiment will be described.

  In the system of this embodiment, work guidance for instructing an operator about work contents and virtual work training for making an operator perform the instructed work contents in a virtual work area are performed. In this virtual work training, In addition, it is possible to evaluate the correctness of the work whether or not the work is correctly performed and the proficiency level of the work (degree of work acquisition) and to provide feedback to the worker.

  First, when starting up the system, for example, when an operator sits in front of the VR control personal computer 3 and starts the system, the screen shown in FIG. 3 is displayed on the display of the VR control personal computer 3 to perform work training. Enter the name code and name of the worker to receive, and click the work start button. The same screen may be displayed on the large screen 2a of the VR display device 2.

  When the work start button is clicked, the work content setting screen shown in FIG. 4 is displayed and the work content is set. In the setting of the work content, the selection of the product type to be subjected to the work training, the selection of the target production line, and the selection of the number of processes in the production line are set, and voice guidance is provided. In addition, labeling of work attention items and the like, display of red spheres and blue spheres indicating positions to be described later, setting of the speed (reproduction speed) of animation movement of red spheres and the like are performed.

  In this work content setting screen, a work guidance for instructing the work content to the worker, a virtual work training for performing virtual work training, and a button for returning to the work selection screen are displayed. To start.

  FIG. 5 is a diagram for explaining a work guidance screen. In this screen, a virtual work area where the parts box 16 and the like are placed is displayed on the work desk 15. The parts box 16 displays an image 25 of the housed parts. Further, in this virtual work area, for example, the position (place) where the part is picked up is displayed as a red sphere 17, and the position (place) where the part is taken is displayed as a blue sphere 18. The red sphere 17 is animated toward the blue sphere 18 as shown in FIG. In addition, the work content, work caution items, and confirmation items are displayed in characters using yellow or green labels 19. Further, a photograph 26 of the part to be worked is displayed in an inset manner.

  Along with the display of the work guidance screen, work contents, work caution items and confirmation items are output from the speaker of the VR control personal computer 3 by voice.

  There are various work contents, caution items, and confirmation items depending on the work. For example, in the case of the work of removing the board, the work contents include “breaking the board” and cautions for the work. The item is, for example, “careful about the dividing direction of the board”, and the work confirmation item is, for example, “whether the dividing direction is correct”. In this case, the position where the substrate is taken is displayed as a red sphere, and the position where the substrate is divided is displayed as a blue sphere.

  FIG. 6 is a diagram showing the development of the display screen and the flow of processing in the work guidance for instructing work contents, and (a) to (c) correspond to the first process to the third process.

  In this work guidance, the worker uses the screen displayed on the display of the VR control personal computer 3 having a function as work guidance means and the sound output from the speaker of the VR control personal computer 3 to obtain the work content, work position, Notes and confirmation items are repeatedly memorized at their own pace by key operation.

  In this work guidance, on the display of the VR control personal computer 3, the virtual work area and the work position corresponding to the work content are indicated by the red sphere and the blue sphere as described above for each process as in FIG. They are displayed in order, and the work contents, caution items, confirmation items, etc. are displayed in labels. Further, work contents, caution items, and confirmation items are output by voice.

  That is, as shown in FIG. 6A, when the start key 20 of the keyboard of the VR control personal computer 3 is operated in a state where the virtual work area is displayed, for example, The animation moves toward the blue sphere indicating the work position, and the work contents, caution items, and confirmation items are displayed in a label, while being output by voice, and the first step ends.

  Next, when the operator operates the key 21, the screen is switched to the virtual work area screen in the second step, and the red sphere moves toward the blue sphere in accordance with the work content. While the caution item and the confirmation item are displayed as labels, they are output by voice, and the second step ends.

  Further, when the operator operates the key 21, the screen is switched to the virtual work area screen in the third step, and the red sphere moves in an animation toward the blue sphere corresponding to the work content. The caution item and the confirmation item are displayed as labels, while being output by voice, and the third step ends.

  Thereafter, work guidance for a series of steps is completed in the same manner. In addition, it is possible to return to the previous process by operating the key 22, and to return to the first process by operating the reset key 23. The work content can be stored reliably.

  Further, by performing a fully automatic setting operation, it is possible to eliminate the need to switch the screen by the key 21 to each process.

  The worker who has received the work guidance and stored the work content as described above acquires the work by performing the virtual work training in the virtual work area.

  FIG. 7 is a diagram showing the screen development and processing flow of virtual work training.

  In this virtual work training, as shown in FIG. 2 described above, the worker wears the magnetic sensor 11 of the motion detection device 8 on the left and right hands, and also the microphone 6 and the speaker 5 of the voice control personal computer 7 (the microphone with headphones). ) In addition, wearing polarized glasses 27 for stereoscopic viewing, standing in front of the large screen 2a of the VR display device 2 and working in a full-scale virtual work area displayed stereoscopically on the large screen 2a This is a pseudo training.

  Further, the VR control personal computer 3 determines whether the work is correct or not based on the voice of the work content by the worker recognized by the voice control personal computer 7 and the movement of the worker's hand from the motion control personal computer 9 as described later. Determine the proficiency level of the work.

  First, on the large screen 2a of the VR display device 2, all red spheres and blue spheres of a series of work guidance processes are displayed in a three-dimensional manner together with a full-size virtual work area.

  Next, when the worker performs a start operation on the VR control personal computer 3, a hand model corresponding to the worker's hand is displayed in the virtual work area and linked to the movement of the worker's hand. The worker reaches for the red sphere corresponding to the first step of the three-dimensional virtual work area, takes the red sphere, moves it to the blue sphere, and corresponds to the first step. The work is performed in a pseudo manner while speaking the contents of work to be performed, caution items, and confirmation items. An operator learns hand movement and position while looking at a hand model in a virtual work area linked to the worker's own hand movement.

  On the system side, the operator's voice is recognized, and the correctness of the work in the first step and the proficiency level of the work are evaluated based on the movement of the hand.

  Note that the left and right hands in the work are divided into the left hand when the red sphere is on the left side for the worker, the right hand when the red sphere is on the right side, and the left and right hand for the two red spheres. Although it is basic to use, you may specify the hand to use according to the work contents.

  In this embodiment, the correctness / incorrectness of the work is determined based on whether the work contents, caution items, and confirmation items corresponding to the work are correctly uttered, and the red and blue spheres at the correct positions corresponding to the work. It is determined based on whether or not the person's hand has been touched.

  When the work is not correct, for example, a comment such as “Wrong” is output by voice to prompt the user to work again. When the work is performed correctly, the red and blue spheres in the first step are erased and Encourage work.

  Next, the worker takes the red sphere at the position corresponding to the second step and moves it to the position of the blue sphere to perform the work corresponding to the second step in a pseudo manner, and also the work contents, caution items and confirmation items. The system side determines whether the work in the second process is correct or not, and then performs the work in each process in the same manner.

  When the operator operates the reset key 23 on the VR control personal computer 3, the initial screen before the first process is displayed and the operation can be started again from the beginning.

  In this virtual work training, work proficiency is also evaluated as described later.

  In this virtual work training, work training is performed using virtual reality, so that the training is performed in a state close to the actual work, and the accuracy of the work is improved, and the time required for the training is less than before. Shortened. Moreover, it is possible to perform training independently without a work instructor.

  Next, the general operation of the system in the above work guidance and virtual work training will be described based on a flowchart.

  FIG. 8 is a flowchart of work guidance. First, a virtual work area is displayed on the display of the VR control personal computer 3 (step n1), and red and blue spheres are displayed in this work area (step n2). ), The red sphere moves to the position of the blue sphere (step n3), “work contents”, “caution items”, and “confirmation items” are displayed as labels and output from the speaker of the VR control personal computer 3 by voice. The process proceeds to step n5 (step n4).

  In step n5, it is determined whether or not the work is a final process. When the work is not the final process, the process returns to step n2, and when the work is the final process, the work guidance is terminated.

  Next, in the virtual work training shown in FIG. 9, the virtual work area is stereoscopically displayed on the large screen 2a of the VR display device 2 (step n11), and the model of the operator's hand is displayed in the virtual work area. (Step n12), the hand model and the worker's hand detected by the motion detection device 8 are linked (step n13), and the red and blue spheres corresponding to the work of all processes in the virtual work area All are displayed (step n14).

  Next, when the start of the virtual work training is input (step n15), the work data of the worker is acquired as described later (step n16), whether the work is correct or not is determined. If it is not correct, an NG comment such as “I am wrong” is output by voice and the process returns to step n16 (step n27). The correctness / incorrectness determination in step n17 is performed based on whether or not the operator has touched the correct red sphere.

  If it is determined in step n17 that it is correct, the proficiency level of the work is determined as described later (step n18), the work data of the worker is acquired (step n19), and the correctness of the work is determined (step n20). If it is not correct, an NG comment such as “I am wrong” is output by voice and the process returns to step n19 (step n28). The correctness / incorrectness determination in step n20 is performed based on whether or not the operator has touched the correct blue sphere.

  When it is determined in step n20 that it is correct, the proficiency level of the work is determined as described later (step n21), and the contents of the work by the worker, the caution items and the utterances of the confirmation items are recognized by voice (step n22), The correctness / incorrectness of the work as to whether or not a correct utterance has been made is determined (step n23). If it is not correct, an NG comment such as “I am wrong. Please say again” is output by voice and the process returns to step n22 (step n29). ).

  If it is determined in step n23 that it is correct, for example, an OK comment such as “It is correct” is output by voice (step n24), and it is determined whether or not it is work of the final process (step n25). When it is not a work, the process returns to step n16, and when it is a work of the final process, the comprehensive judgment result including the skill level of the work is displayed on the screen 2a and the process is ended (step n26).

  As shown in FIG. 10, the current time T1 is acquired (step n30), and the current right hand (model) position Ar [X, Y, Z] is acquired in the above-described step n16 and step n19. (Step n31), the current left hand (model) position Al [X, Y, Z] is obtained (Step n32), and the current red sphere (or blue sphere) position B [X, Y. Z] (step n33), the position Cri [X, Y, Z] of the operating right hand (model) is acquired (step n34), and the position Cli [X, Y, of the operating left hand (model) is acquired. Z] (step n35), it is determined whether or not the right hand (or left hand) has touched the red sphere (or blue sphere) (step n36). If not touched, the process returns to step n34. The current time T2 is acquired and the process ends (step n37).

  As described above, the position data until the left hand or the right hand touches the red sphere or the blue sphere and the data required for the position are acquired.

  Next, the process of determining the proficiency level of work (the degree of work acquisition) in step n18 and step n21 in FIG. 9 described above will be described.

  In this embodiment, the evaluation is performed using three types of indices: work time, hand movement blur (meandering), and hand movement secondary regression curve contribution rate.

  That is, as shown in FIG. 11, in the evaluation based on the work time (operation time), is the work time (T2-T1) calculated from the work data acquired in FIG. 10 described above equal to or less than the standard work time? If it is not less than the standard work time, the master comprehensive judgment is displayed as NG on the screen 2a (step n41).

  Further, for example, in the evaluation by hand movement shaking (meandering), the normal vectors of A, B, and Ci of the work data acquired in FIG. 10 described above are obtained (step n42), and one of them is obtained. Is used as a reference to obtain an angle with another vector (step n43).

  It is determined whether or not the average value of the angles is equal to or less than the reference value (step n44). If the average value is not equal to or less than the reference value, the learning comprehensive determination is displayed on the screen 2a as NG because the movement of the hand is meandering. (Step n41).

  In step n45, it is determined whether or not the maximum value of the angle is equal to or less than the reference value. If the maximum value of the angle is not equal to or less than the reference value, it is determined that the learning comprehensive determination is NG on the screen 2a. Display (step n41).

  In step n46, it is determined whether or not the ratio of points exceeding the reference value is equal to or less than the reference (%). Is displayed on the screen 2a (step n41).

  Further, in the evaluation based on the contribution rate of the quadratic regression curve of the hand movement, A, B, and C of the work data acquired in FIG. 10 are projected onto the XY plane and the YZ plane (step n47), and the secondary is obtained. Approximate with a curve to determine the contribution rate (step n48), determine whether the contribution rate is greater than or equal to the reference value (step n49), and if not greater than the reference value, display the master comprehensive judgment as NG on the screen 2a. (Step n41).

  When it is determined that the above three types of evaluations are OK, that is, when it is determined YES in the above-described steps n40, 44, 45, 46, and 49, the skill level comprehensive determination OK is displayed on the screen 2a. The virtual work training is finished (step n50).

  Thus, simultaneously with the virtual work training, the correctness of the work and the proficiency level of the work are evaluated in real time and fed back to the worker.

  Note that if the proficiency level comprehensive evaluation is evaluated as OK, the work training is ended and the work is performed on the actual production line.

  Here, the evaluation of the proficiency level of the above work will be described in more detail.

The proficiency level of the work is evaluated based on the wonder during the work, the trajectory of the hand, and the like. As described above, the presence / absence of shaking (meandering) of the hand trajectory and the working time are used as the index for the stray behavior.
(1) Lost movement (meandering)
As one of the methods for discriminating the movement, the meandering (blurring) of the hand during the holding operation is measured to determine the presence / absence of “stuck”.

  The position of the hand at the operation start point is A (Xa, Ya, Za), the part position (end point) is B (Xb, Yb, Zb), and the position of the hand during the holding operation is Ci = C (Xi, Yi, Zi). ) (I = 1, 2,..., Ni). A normal vector αCi of a plane passing through the start point A, the end point B, and Ci is obtained. Then, an angle θi with another vector is calculated based on one (αCw) of the obtained vectors.

  The normal vector αCi is obtained by the equation (1).

同様に、αＣｗを求める。αＣｉとαＣｗの２つのベクトルのなす角度θｉは、式（２）により求まる。

Similarly, αCw is obtained. The angle θi formed by the two vectors αCi and αCw can be obtained from Equation (2).

そして、θｉが０に近ければ、蛇行がないと判定するのである。

If θi is close to 0, it is determined that there is no meandering.

FIGS. 12A and 12B show examples of plane normal vectors when there is no meandering and when there is meandering.
(2) Loss of operation (time)
In addition to the trajectory, it is conceivable that the movement ambiguity occurs before the movement starts (stagnation time). Therefore, the presence / absence of “being lost” is determined based on the time required for the reserve operation.

  The standard operation time is set from the distance between the start point and the end point of the movement operation using standard data (MTM method). If the operator's operation time is equal to or less than the standard operation time, it is determined that no hesitation has occurred.

（３）動作軌跡の滑らかさ
作業に慣れるほど、手の軌跡は、滑らかになると考えられる。そこで、手の動きの滑らかさに着目して、到達度を評価する。手の軌跡を、ＸＹ平面、ＹＺ平面に投影し、２次曲線で近似する。

(3) Smoothness of motion trajectory It is considered that the trajectory of the hand becomes smoother as it gets used to work. Therefore, focusing on the smoothness of hand movement, the degree of achievement is evaluated. The locus of the hand is projected on the XY plane and the YZ plane and approximated by a quadratic curve.

The contribution ratio R ² l (X, Y) of the approximate expression on the XY plane is obtained by the following expression.

そして、寄与率が、１に近ければ、動作軌跡が２次曲線で表され、滑らかな動きになったと判定できる。そこで、ＸＹ平面における寄与率Ｒ^２ｌ（ＹＺ）を横軸に、ＹＺ平面における寄与率Ｒ^２ｌを縦軸にプロットとし、それぞれの点Ｐｌと座標（１，１）との距離の和εを求め、動作の滑らかさの指標とする。

If the contribution ratio is close to 1, the motion trajectory is represented by a quadratic curve, and it can be determined that the movement is smooth. Therefore, the contribution rate R ² l (YZ) in the XY plane is plotted on the horizontal axis, and the contribution rate R ² l in the YZ plane is plotted on the vertical axis, and the sum ε of the distances between the respective points Pl and coordinates (1, 1) As an index of smoothness of movement.

なお、上述の実施の形態では、採用していなかったが、次のような身体部位の動きを用いて評価してもよい。
（４）身体部位の移動範囲
体全体の動きから、作業習熟度の到達度を評価する。作業に慣れる程、動作は規則的となり、各部位の動作範囲も小さくなると仮定する。そこで、作業中の身体の各部位の位置データを求める。各部位、例えば、肘２８の動作範囲は、図１３に示される２次元信頼楕円により近似する。信頼楕円の長径ｒ１および短径ｒ２は、次式により与えられる。

Although not employed in the above-described embodiment, the following movement of the body part may be used for evaluation.
(4) Movement range of body part The achievement level of work proficiency is evaluated from the movement of the whole body. It is assumed that the operation becomes regular and the operation range of each part becomes small as it gets used to work. Therefore, the position data of each part of the body during work is obtained. The movement range of each part, for example, the elbow 28 is approximated by a two-dimensional confidence ellipse shown in FIG. The major axis r1 and minor axis r2 of the confidence ellipse are given by the following equations.

以上の本発明の実施の形態の効果を明らかにするために、以下のような条件でバーチャル作業訓練を行い、上述の（１）〜（４）の４種類の指標によって、作業習熟度を評価した。

In order to clarify the effects of the embodiment of the present invention described above, virtual work training is performed under the following conditions, and work proficiency is evaluated by the four types of indicators (1) to (4) described above. did.

  As shown in FIG. 14, the target assembly process is an operation of bringing different types of parts placed in 10 parts storage places of the parts shelf 29 to the assembly position in the order shown. Virtual work training was conducted 5 times for the subjects.

  FIG. 15 is a diagram showing the evaluation result of the stray operation (meandering). FIG. 15 shows the result of analyzing the action of extending the right hand to the tenth part. However, a preliminary experiment was performed, and the criterion for meandering was set to 10 °.

  In the first work training, meandering was confirmed twice, whereas in the fifth time, meandering did not appear. Similarly, by conducting 5 work trainings on all 10 parts, hand meandering disappeared.

  FIG. 16 is a diagram showing the evaluation result of the ambiguity (time) of the operation. In this drawing, the time required for the operation of reaching the tenth component is shown.

  After the fourth work training, the operation time WT became the standard operation time SWT or less. As a result, after the fourth time, it was determined that no hesitation occurred. Similarly, as a result of measuring the time for all the ten parts, there was no longer any inconvenience from the working time by the fifth time.

  FIG. 17 is a diagram showing the evaluation results of the smoothness of the operation, and shows the result of analyzing the trajectory of the operation reaching for the tenth part. In FIG. 17, the quadratic curve of the motion trajectory and the contribution rate of the regression equation are shown. (A) shows the first time and (b) shows the fifth time.

  The contribution rate is closer to 1 in the fifth time, and the operation is smooth. Similar results were obtained in the YZ plane, and the total distance ε was reduced by 20% in the fifth time compared to the first time.

  FIG. 18 is a diagram showing the evaluation result of the movement range of the body part, and shows the movement range of the right elbow in the movement of extending the right hand to the tenth part. In the first time shown in (a), the operation range bends irregularly and exceeds the range obtained by the preliminary experiment. However, at the fifth time shown in (b), the operation is smooth and the operation range is within the range obtained by the preliminary experiment.

  The work training support system of the present invention may evaluate the proficiency level of the work using at least one of the above-mentioned indices, but it is preferable to comprehensively evaluate by combining all the indices. .

(Other embodiments)
The present invention is not limited to the analysis method of the above-described embodiment using a normal vector, a quadratic regression curve, or the like, but detects the movement of each part of the body as a three-dimensional spatial coordinate position and angle, and performs other analysis. You may make it determine the grade of work acquisition using a method.

  In the above-described embodiment, the virtual work area is displayed on the large-size polarizing stereoscopic screen and the virtual work training is performed. However, as another embodiment of the present invention, as a VR display device that displays the virtual work area. A head mounted display may be used.

  The VR display device 2 is preferably capable of stereoscopic display, but may be a two-dimensional display device.

  In the above-described embodiment, the magnetic sensors of the motion detection device are mounted on the left and right hands, but the number and mounting positions of the magnetic sensors may be arbitrarily selected, and accordingly, The model of the worker is not limited to the hand, but other parts or the model of the entire worker may be displayed.

  The motion detection device is not limited to a magnetic type, and may be an optical type or a mechanical type.

  The present invention is useful for work training of various work.

1 is a schematic configuration diagram of a system according to an embodiment of the present invention. It is a figure for demonstrating the motion detection apparatus of FIG. It is a figure which shows the display screen at the time of starting of a system. It is a figure which shows a setting screen. It is a figure which shows the screen of work guidance. It is a figure which shows the screen expansion of a work guidance, and the flow of a process. It is a figure which shows the flow of virtual work training and a process. It is a flowchart with which it uses for operation | movement description of work guidance. It is a flowchart with which it uses for operation | movement description of virtual work training. It is a flowchart with which operation | movement description of operation | work data acquisition is provided. It is a flowchart with which it uses for operation | movement description of work training evaluation. It is a figure for demonstrating a normal vector. It is a figure for demonstrating a trust ellipse. It is a figure which shows the components shelf of an assembly process. It is a figure which shows the evaluation result of the ambiguity (meandering) of operation | movement. It is a figure which shows the evaluation result of the hesitation of operation (time). It is a figure which shows the evaluation result of the smoothness of operation | movement. It is a figure which shows the evaluation result of the movement range of a body part.

Explanation of symbols

2 VR control device 3 VR control personal computer 7 Voice control personal computer 8 Motion detection device 9 Motion control personal computer 17 Red ball 18 Blue ball

Claims (7)

A work training support system for supporting work training for workers,
A work guidance means for displaying a virtual work area and guiding work contents in the virtual work area;
A work training means for displaying a virtual work area, detecting a movement of an operator and displaying the virtual work area to perform work training;
Work evaluation means for evaluating the result of work training by the work training means,
The work evaluation support system characterized in that the work evaluation means evaluates the correctness of work and the degree of work acquisition. The work guidance means displays a position corresponding to the work content in the virtual work area,
The work training means displays the position corresponding to the virtual work area, the worker's model and work contents so as to be stereoscopically viewed,
The work evaluation means evaluates the degree of work accuracy and work acquisition based on the detected movement of the worker, and notifies the worker of the evaluation result by at least one of display and sound. The work training support system according to 1. The work guidance means outputs the work content by voice,
The work training means recognizes a worker's voice corresponding to the work content,
The work training support system according to claim 1, wherein the work evaluation unit evaluates the correctness of the work based on the recognized voice. The work guidance means includes a display unit that displays the virtual work area, and a voice output unit that outputs work contents by voice.
The work training means inputs a stereoscopic video display unit that displays the virtual work area and the worker's model in a stereoscopic view, a motion detection unit that detects the movement of the worker, and voice of the worker. The work training support system according to claim 3, further comprising a voice recognition unit for recognizing.   The work training support according to any one of claims 2 to 4, wherein the work evaluation means evaluates the correctness of the work and the degree of work acquisition based on the movement of the worker's hand to the position corresponding to the work content. system.   6. The work training support system according to claim 5, wherein the work evaluation means evaluates the degree of work acquisition based on a predetermined standard operation time and the detected operation time of the worker.   The work according to claim 5 or 6, wherein the work evaluation means evaluates the degree of work acquisition based on at least one of a worker's ambiguity and smoothness of movement to the position corresponding to the work content. Training support system.

Images