Operation evaluation device, operation evaluation method and operation evaluation system
By automatically calculating the action compliance rate and prediction model, the accuracy and real-time problems of operator operation evaluation on the production line are solved, and the balance of the production line and product quality are improved.
Patent Information
- Application Number
- CN201811249117.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2018-10-25
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2038-10-25
AI Technical Summary
The operating evaluation methods of existing production line operators rely on human eye measurement, with poor data accuracy and cannot reflect the operator's skill level in real time, affecting production line balance and product quality.
An operation evaluation device is provided, by storing a reference action sequence and an actual action sequence, calculating a compliance rate and a prediction model, and automatically evaluating the operator's operating performance, including action compliance rate, time compliance rate and process compliance rate, and predicting potential defects.
It realizes efficient evaluation of operator skills, identify missing actions, improve production line balance and product quality, reduce human error, and improve production line optimization efficiency.
Smart Images

Figure CN111105106B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to an operation evaluation device, an operation evaluation method, and an operation evaluation system for evaluating the operation of an operator on a production line. Background Art
[0002] At present, production lines for manufacturing or assembling various products are widely used in various industries. On each production line, each production unit only focuses on processing a certain segment of work to improve work efficiency and output. Usually, multiple workstations can be arranged on the production line according to factory needs, product complexity, and manufacturing or assembly complexity. The operator at each workstation manually repeats operations that are different from other operators. After completing the operation at one workstation, they flow to the next workstation to operate, and finally complete the manufacturing / assembly.
[0003] For example, in the assembly line of the electronics industry, most of the operations are currently performed manually by operators at each station. Therefore, not only are there operational deviations in repeated operations for each operator, but there are also skill differences between operators. Such operational deviations and skill differences may not only cause quality problems such as product defects, but may also affect the line balance of the entire production line. Line balance is a technical means and method to average the operations of each station and adjust the operation load to make the operation time as close as possible. Its quality is directly related to the normal use of the production line. Therefore, it is particularly important to evaluate the operator's operation in a timely and effective manner.
[0004] In terms of operation evaluation, currently, IE engineers who optimize the production line mainly use stopwatches to measure the working hours of each workstation. The reason for measuring working hours is that it can intuitively show whether the operator has completed the operation within the specified time, and it can adjust the bottleneck process time (also known as the highest process time) of the entire production line based on it. However, this method has problems such as poor data accuracy due to the use of human eyes, and the operation performed by the operator in the measurement environment of the IE engineer cannot reflect the usual operation. Summary of the invention
[0005] The present application has been made in view of the above-mentioned situation, and an object of the present application is to provide an operation evaluation device, an operation evaluation method, and an operation evaluation system for automatically evaluating the operation of an operator on a production line.
[0006] In order to achieve the above-mentioned purpose, the present invention provides an operation evaluation device for evaluating the operation of an operator on a production line, and is characterized in that it comprises: a storage unit for storing a pre-set benchmark action sequence, the benchmark action sequence including the action content of each of more than one action that the operator should perform within a process cycle and the benchmark time required for the action; an acquisition unit for acquiring an actual action sequence of the operator who repeatedly performs the more than one action on the production line, the actual action sequence including the content of the executed action and the time consumed; a calculation unit for calculating a compliance rate based on the action content and the time consumed in the actual action sequence acquired by the acquisition unit and the action content and the benchmark time in the benchmark action sequence stored in the storage unit, the compliance rate indicating the ratio of the operation performed by the operator that complies with the benchmark action sequence; and an evaluation unit for evaluating the operation of the operator based on the compliance rate calculated by the calculation unit.
[0007] Furthermore, the calculation unit may calculate a ratio of the number of action contents in the actual action sequence to the total number of action contents that should be included in the plurality of cycles as the first operation matching rate.
[0008] Furthermore, the calculation unit may calculate, among the plurality of cycles of action contents, a ratio of action contents whose consumption time matches the corresponding reference time as the second operation matching ratio.
[0009] Furthermore, it may also be that the calculation unit calculates the ratio of action contents whose consumption time conforms to the corresponding reference time as the third operation compliance rate in each cycle of the multiple cycles, and calculates the variance of the third operation compliance rate within the multiple cycles; and the evaluation unit also evaluates the operation of the operator based on the variance of the third operation compliance rate calculated by the calculation unit.
[0010] Furthermore, the calculation unit may calculate a ratio of the number of times a process conforms to the reference operation sequence in the plurality of cycles as a process conformity rate.
[0011] Furthermore, a process conforming to the benchmark action sequence may mean that: the process consumption time obtained based on the consumption time of each action in the process conforms to the process benchmark time obtained based on the benchmark time of each action; or, the consumption time of each action in the process conforms to the corresponding benchmark time.
[0012] Furthermore, the calculation unit may calculate, for each action, a ratio of the number of times the consumed time of the action matches the reference time of the action in the plurality of cycles as the first action matching ratio.
[0013] Furthermore, it may also be that the benchmark action sequence and the actual action sequence also include at least one information of the tool used, workpiece, operation position, container, and operation direction for each action; the calculation unit also calculates, for each action, the ratio of the number of times the at least one information of the action conforms to the benchmark action sequence in the multiple cycles as the second action compliance rate, and the calculation unit weights the first action compliance rate and the second action compliance rate for each action as the action compliance rate.
[0014] Furthermore, the calculation unit may weight the first action matching rate of each of the one or more actions to obtain the weighted first action matching rate.
[0015] Furthermore, it may be that the storage unit further stores at least one prediction model, each of which represents an action in an action sequence that may cause a defect; the operation evaluation device further includes a prediction unit, which predicts a defect by comparing the actual action sequence with the prediction model stored in the storage unit; the evaluation unit evaluates the operator's operation based on the compliance rate and the prediction result of the prediction unit, and gives a prompt when the prediction unit predicts a defect.
[0016] Furthermore, the prediction unit may predict a defect by comparing the actual action sequence with the prediction model stored in the storage unit only when the coincidence rate calculated by the calculation unit is lower than a preset threshold value.
[0017] In order to achieve the above-mentioned purpose, the present invention provides an operation evaluation method, which is used in an operation evaluation device for evaluating the operation of an operator on a production line, wherein the operation evaluation device has a storage unit for storing a pre-set benchmark action sequence, wherein the benchmark action sequence includes the action content of each of more than one action that the operator should perform within a process cycle and the benchmark time required for the action; in the operation evaluation method, it includes: an acquisition step, acquiring an actual action sequence performed by the operator in multiple cycles of the process, wherein the actual action sequence includes the action content and its consumed time for each action performed; a calculation step, calculating a compliance rate based on the action content and its consumed time in the actual action sequence acquired by the acquisition step and the action content and benchmark time in the benchmark action sequence stored in the storage unit, wherein the compliance rate represents the ratio of the operation performed by the operator that complies with the benchmark action sequence; and an evaluation step, evaluating the operation of the operator based on the compliance rate calculated by the calculation step.
[0018] In order to achieve the above-mentioned purpose, the present invention also provides an operation evaluation system for evaluating the operations of operators at each workstation on a production line, comprising: any one of the operation evaluation devices as described above, arranged at each workstation; a server, communicating with the operation evaluation device via a network, receiving data related to the operator's operations from the operation evaluation device and storing the data; and a client, accessing the server via the network.
[0019] Effects of the Invention
[0020] According to the operation evaluation device, operation evaluation method and operation evaluation system of the present invention as described above, by calculating the ratio of the operations in the actual action sequence to the benchmark action sequence, that is, the compliance rate, the operator's skill level can be evaluated, and by comparing the action content and its consumed time in the actual action sequence with the action content and benchmark time in the benchmark action sequence, it can also be known which actions the operator has missed. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 It is a diagram showing the overall architecture of the operation evaluation system of the present invention.
[0022] Figure 2 This is a block diagram showing the structure of the operation evaluation device according to the first embodiment of the present invention.
[0023] Figure 3 This is a diagram showing an example of a reference operation sequence in the first embodiment of the present invention.
[0024] Figure 4 This is a flowchart showing the operation evaluation method according to the first embodiment of the present invention.
[0025] Figure 5 It is a block diagram showing the structure of the operation evaluation device according to the second embodiment of the present invention.
[0026] Figure 6 This is a diagram showing an example of a prediction model in the second embodiment of the present invention.
[0027] Figure 7 : is a flowchart showing the operation evaluation method according to the second embodiment of the present invention. DETAILED DESCRIPTION
[0028] Hereinafter, the implementation manner of the present invention will be described in detail with reference to the accompanying drawings, taking a notebook computer assembly line as an example.
[0029] (First Embodiment)
[0030] Figure 1 It is a diagram showing the overall architecture of the operation evaluation system of the present invention.
[0031] like Figure 1 As shown, the operation evaluation system 1 includes an operation evaluation device 10 , a server 20 , and a client 30 .
[0032] The operation evaluation device 10 is set at each workstation of the assembly line. Figure 1 There are five stations, S1 to S5, in the system. The operators at each station repeatedly perform different processes from those at other stations. The assembly object is operated in turn at each station, and finally the laptop is assembled through the operation at station S5. Here, the number of stations is an example, and any number of stations can be set according to factory needs, product complexity, and operation complexity.
[0033] The operation evaluation device 10 evaluates the operation of the operator at the station in such an assembly process.
[0034] The server 20 communicates with each operation evaluation device 10 via the IoT network, receives and stores various data related to the operator's operation from the operation evaluation device 10. As an IoT communication protocol of the IoT network, for example, a protocol such as MQTT can be used.
[0035] The client 30 is, for example, a computer, and can access the server 20 through a network to query various data on the server 20 .
[0036] In addition, the description here assumes that the operation evaluation device is provided for each workstation. However, the present invention is not limited to this, and the operation evaluation can also be performed on the server side. Furthermore, the server and the client can also be integrated.
[0037] Hereinafter, the configuration of the operation evaluation device 10 will be described in detail.
[0038] Figure 2 1 is a block diagram showing the structure of the operation evaluation device of the first embodiment of the present invention. Figure 2 As shown, the operation evaluation device 10 includes a storage unit 11 , an acquisition unit 12 , a calculation unit 13 , and an evaluation unit 14 .
[0039] The storage unit 11 stores a preset reference action sequence, which includes the action content of one or more actions that the operator of the current station should perform in a process cycle and the reference time required for the action. Here, the reference action sequence can be set directly on the operation evaluation device 10, or it can be set on the server 10 first and then imported into the operation evaluation device 10.
[0040] As an example of a reference action sequence, Figure 3 In. Figure 3As shown, the benchmark action sequence of this process includes the action contents of 5 actions, namely, "place the jig only against the edge of the assembly line and put on the COVER", "scan", "install the PW PCB", "install the speaker to the COVER and fix it with screws", "lay the speaker HNS" and the benchmark time required for each action. Here, each "action" can be an event representing the interaction between any two of the configured objects (workpieces, containers, tools, widgets, etc.) including hands, or the appearance and disappearance of a certain configured object, or a combination of multiple events, or an action on a larger scale. Moreover, the number of actions included in a process is certainly not limited to 5, and can be determined to be any number of more than one according to the assembly process, the difficulty of the action, the working hours, etc.
[0041] The acquisition unit 12 acquires the actual action sequence executed in multiple cycles of the process of this workstation. The actual action sequence is similar to the reference action sequence, and includes action content and its consumption time for each action executed. However, as the process cycles, the action content contained therein also repeats periodically.
[0042] The actual action sequence can be data input manually or data detected and recognized by any means. In the case where the actual action sequence is data detected and recognized, the following acquisition method can be envisioned: a camera is set up at each workstation, and the camera captures the operator's operation to obtain a video, and the recognition engine uses deep learning, target detection, target tracking, action recognition and other technologies to analyze the video and obtain the action sequence.
[0043] The calculation unit 13 calculates a compliance rate representing the ratio of the operator performing operations that comply with the reference action sequence based on the action contents and the consumed time in the actual action sequence acquired by the acquisition unit 12 and the action contents and the reference time in the reference action sequence stored in the storage unit 11 .
[0044] That is, the calculation unit 13 calculates the compliance rate of the operator's operation by comparing the action content and the time consumed in the actual action sequence with the action content and the reference time in the reference action sequence. Here, the compliance rate can be the compliance rate of the action unit, or the compliance rate of all actions in a process cycle, i.e., the process unit, or the compliance rate of the entire operation. The details of the compliance rate will be described in detail later.
[0045] The evaluation unit 14 evaluates the operation of the operator at the own station based on the conformity calculated by the calculation unit 13 , and outputs the evaluation result together with the conformity.
[0046] For example, when the compliance rate is higher than a predetermined threshold, the evaluation is good, and when it is lower than a predetermined threshold, the evaluation is poor. In addition, two or more different thresholds can be set to divide the evaluation results into more levels. The thresholds here are all empirical values or statistical values, which can be changed appropriately according to the situation.
[0047] In addition, if the compliance rate is relatively high over a long period of time, on the one hand, it can be considered that the operator has a high proficiency, and on the other hand, it can be considered that the process configuration is relatively appropriate. When the compliance rate is always high, it can be suspected that the process configuration is too simple, the reference time is too long, and other inappropriateness. For another example, when the compliance rate is relatively low, it can be considered that the operator has a low proficiency, and when the compliance rate is always low, it can be suspected that the process configuration is too complicated, the reference time is too short, and other inappropriateness. Therefore, the IE engineer can optimize the process configuration of the production line based on the information output by the evaluation unit. In addition, the evaluation unit 14 can also analyze the compliance rate to further evaluate the appropriateness of the process configuration of the workstation.
[0048] As mentioned above, the calculation of the compliance rate not only provides a basis for evaluating the skills and performance of operators, making it easier to analyze operating data, but also provides effective information for IE engineers' optimization work such as improving production line balance, making optimization more efficient.
[0049] The following is a detailed description of the calculation method of the matching rate.
[0050] (Specific example 1)
[0051] In this specific example, the calculation unit 13 calculates the ratio of the number of action contents in the actual action sequence to the total number of action contents that should be included in multiple cycles (i.e., the total number of action contents when the benchmark action sequence is repeated the multiple times) as the first operation compliance rate.
[0052] For example, if the process of this station includes 5 actions and the actual action sequence is repeated 100 times, the total number of action contents in the actual action sequence should be 500. Therefore, if the actual action sequence includes 400 action contents, the first operation compliance rate is 400 / 500=80%, and if the actual action sequence includes 500 action contents, the operation compliance rate is 500 / 500=100%.
[0053] Based on the first operation matching rate, the evaluation unit 14 can evaluate whether there are any actions that are missed in the plurality of cycles. Therefore, this method is particularly effective in the evaluation focusing on whether the corresponding action is executed.
[0054] In addition, since the actual action sequence may include action contents that are not actions in the reference action sequence, it is preferred that the calculation unit 13 compares the action contents of the actual action sequence and the reference action sequence, and calculates the ratio of the number of action contents in the actual action sequence corresponding to the reference action sequence to the total number of action contents when the reference action sequence is repeated for the multiple times as the first operation compliance rate. In this way, it can be more accurately reflected whether an action to be executed is omitted in the operation.
[0055] Furthermore, in this specific example, the calculation unit 13 further calculates the ratio of the action contents whose consumption time matches the corresponding reference time among the action contents in the plurality of cycles as the second operation matching ratio.
[0056] Specifically, the calculation unit 13 compares the consumption time of each action content in the actual action sequence with the corresponding reference time in the reference action sequence, and calculates the ratio of the number of action contents that consume time within the corresponding reference time to the number of action contents included in the actual action sequence.
[0057] For example, when the actual action sequence includes 500 action contents, if the number of action contents that consume time within the reference time is 400, the second operation conformity rate is 400 / 500=80%; if the number of action contents that consume time within the reference time is 500, the second operation conformity rate is 500 / 500=100%.
[0058] Based on such a second operation compliance rate, the evaluation unit 14 can perform an overall evaluation of the entire operation of a plurality of cycles.
[0059] In the above specific example, the case where two operation matching rates are calculated is described, but the present invention is not limited thereto, and only one operation matching rate may be calculated.
[0060] (Variation of Specific Example 1)
[0061] In the above specific example 1, as long as the consumption time of the action content is within the reference time, the action content is considered to be in compliance with the reference action sequence. However, in the production line, the less consumption time the better, but the closer to the reference time the better, and its stability is also very important.
[0062] Therefore, in this variant, focusing on the stability of the operation, the calculation unit 13 calculates the ratio of the action contents whose time consumed in each cycle is consistent with the corresponding reference time as the third operation compliance rate, and calculates the variance of the third operation compliance rate in the multiple cycles.
[0063] For example, if there are 5 actions in each process and 100 cycles, in the first cycle, if it is assumed that the number of action contents whose consumption time conforms to the corresponding reference time is 4, the third operation conformity rate is 80%, in the second cycle, if it is assumed that the number of action contents whose consumption time conforms to the corresponding reference time is 5, the third operation conformity rate is 100%, in the third cycle, if it is assumed that the number of action contents whose consumption time conforms to the corresponding reference time is 3, the third operation conformity rate is 60%, etc. Then, the variance of the conformity rates of the 100 third operations in 100 cycles is calculated.
[0064] Based on the variance of the operation compliance rate, the evaluation unit 14 can evaluate whether the operation in a plurality of cycles is stable.
[0065] (Specific example 2)
[0066] In the specific example 2, the calculation unit 13 calculates the ratio of the number of times a process conforms to the reference action sequence in a plurality of cycles of the process as the process conformity rate.
[0067] Specifically, the calculation unit 13 calculates the process reference time based on the reference time of each action. Figure 3 In the example shown, the reference times of five actions included in one process are 2.0 seconds, 4.0 seconds, 2.0 seconds, 7.0 seconds, and 11.0 seconds, respectively, and the process reference time can be calculated as 26.0 seconds.
[0068] By the same method, the process consumption time of each cycle in the actual action sequence of the calculation unit 13 is calculated. If the process consumption time is within the process reference time, the process is considered to be in compliance with the reference action sequence.
[0069] Next, the calculation unit 13 calculates the ratio of the number of times the process conforms to the reference action sequence to the total number of cycles as the process conformity rate. For example, if the number of times the process consumption time is within the process reference time (i.e., the number of cycles) is 90 times in 100 cycles, the process conformity rate is 90%.
[0070] According to such a process conformity rate, the evaluation unit 14 can evaluate the operation of the operator in process units, and thus can perform a more detailed evaluation compared to the operation conformity rate of the specific example 2.
[0071] In addition, in the above description, when the process consumption time is within the process reference time, the process is considered to be in compliance with the reference action sequence. However, it is not limited to this. It can also be considered that the process is in compliance with the reference action sequence when the consumption time of each action in a process is in compliance with the corresponding reference time, that is, when the consumption time of each action is within the corresponding reference time. In this way, the process that meets the reference action sequence can be found more reliably.
[0072] (Concrete example 3)
[0073] In the specific example 3, the calculation unit 13 calculates, for each action, the ratio of the number of times the time taken for the action matches the reference time in a plurality of cycles as the action matching rate.
[0074] Specifically, since the cycle is a process cycle, each action should be executed the same number of times in multiple cycles. Therefore, the action compliance rate of a certain action can be calculated by the ratio of the number of times the action consumes time within the reference time to the total number of cycles.
[0075] For example, if each process has 5 actions and is cycled 100 times, for the first action, if the time consumed by the action is within the benchmark time ( Figure 3 In the example of 2.0 seconds, the number of times is 90, then the action compliance rate of the first action is 90 / 100=90%. For the second action, if the time consumed by the action is within the reference time ( Figure 3 In the example of , the number of times within 4.0 seconds is 95, and the action conformity rate of the second action is 95%. The action conformity rates of the third to fifth actions are calculated in the same manner.
[0076] Based on such an action conformity rate, the evaluation unit 14 can evaluate the operator's operation in action units, and thus can perform a more detailed evaluation compared to the operation conformity rate and process conformity rate in specific examples 1 and 2.
[0077] (Variation 1 of Specific Example 3)
[0078] In the above specific example 3, whether the action conforms to the reference action sequence is evaluated in terms of time. However, the present invention is not limited thereto, and whether the action conforms to the reference action sequence may also be evaluated in terms of the standardization of the action itself.
[0079] In this case, the reference action sequence and the actual action sequence include at least one piece of information of the tool used, workpiece, operation position, container, and operation direction for each action in addition to the action content and time. Figure 3 The action "Put the fixture close to the edge of the assembly line and put it on the cover" includes the workpiece "body", the operation position "any horizontal position, vertical bottom", etc. For another example, regarding the action "scanning", it includes the tool "scanning gun", the operation position "any horizontal position, any vertical position", the container "in front of the operating table", etc. For another example, regarding the action "install the speaker to the cover and fix it with screws", it includes the tool "screwdriver", the workpiece "body", the operation direction "lower left", etc. In addition, this information is just an example and other information can also be included.
[0080] The calculation unit 13 calculates, for each action, the ratio of the number of times that at least one of the information described above of the action matches the reference action sequence in the plurality of cycles as the action conformity rate. That is, the ratio of the number of times that the information of a certain action matches the information in the reference action sequence to the total number of cycles is calculated as the action conformity rate.
[0081] Based on such an action conformity rate, the evaluation unit 14 can evaluate the operator's operation in action units in terms of standardization.
[0082] In addition, the calculation unit can also calculate the action compliance rate in time in specific example 3 (hereinafter also referred to as the first action compliance rate) and the action compliance rate in standardization in this variant (hereinafter also referred to as the second action compliance rate) for each action, and weight these two action compliance rates to obtain the final action compliance rate. Here, the weight can be set appropriately according to the needs. For example, if more emphasis is placed on the time requirement, the weight of the first action compliance rate is set to be large; if more emphasis is placed on the standardization requirement, the weight of the second action compliance rate is set to be large. In this way, an action compliance rate that takes into account both the time requirement and the standardization requirement can be obtained.
[0083] (Variation 2 of Specific Example 3)
[0084] Since the entire operation in multiple cycles is composed of various actions. Therefore, in this variation, the calculation unit 13 uses any action compliance rate obtained in the above-mentioned specific example 3 and its variation 1 to weight the action compliance rate of each action (for example, 5) included in the process to obtain the operation compliance rate of the entire operation. Here, the weight can be appropriately set according to the importance, complexity, sequence, etc. of the action. In this way, the operation compliance rate can be flexibly obtained according to the needs.
[0085] Below, refer to Figure 4 The operation evaluation method of this embodiment will be described. Figure 4 : is a flowchart showing the operation evaluation method according to the present embodiment.
[0086] like Figure 4 As shown, the acquisition unit 12 acquires the actual action sequence executed in multiple cycles of the process of the current station, and the actual action sequence includes the action content and the time consumed by the action for each action (step S11).
[0087] Next, the calculation unit 13 calculates a compliance rate representing the ratio of the operator's execution of operations that comply with the benchmark action sequence based on the action content and consumed time in the actual action sequence obtained by the acquisition unit 12 and the action content and benchmark time in the pre-stored benchmark action sequence. The benchmark action sequence includes the action content and benchmark time required for each of one or more actions that should be performed by the operator of this workstation within a process cycle (step S12).
[0088] Next, the evaluation unit 14 evaluates the operation of the operator at the own station based on the calculated coincidence rate, outputs the evaluation result together with the coincidence rate (step S13), and ends the processing.
[0089] In the above embodiment, several specific examples and variations of calculating the coincidence rate are described. These specific examples and variations can be combined arbitrarily. That is, any one or more coincidence rates among the multiple coincidence rates described in the above specific examples and variations can be calculated as needed.
[0090] (Second Embodiment)
[0091] Below, refer to Figure 5 An operation evaluation device according to a second embodiment of the present invention will be described.
[0092] The operation evaluation device of this embodiment adds a function of predicting potential quality defects in operation based on the operation evaluation device of the first embodiment. The following focuses on the differences between this embodiment and the first embodiment, and the same or similar contents as the first embodiment are omitted in this embodiment.
[0093] Figure 5 2 is a block diagram showing the structure of the operation evaluation device according to the second embodiment of the present invention. Figure 5 As shown, the operation evaluation device 100 of the present embodiment includes a storage unit 101 , an acquisition unit 102 , a calculation unit 103 , an evaluation unit 104 , and a prediction unit 105 .
[0094] The storage unit 101 not only stores the reference action sequence, but also stores at least one prediction model.
[0095] Each prediction model represents an action in the action sequence that may cause a defect. For example, the action may be that a certain component appears at a time when it should not appear, the number of components that should appear at a time is inaccurate, the order of actions that should not be reversed is reversed, etc. In addition, each prediction model can be obtained by the user marking defects in the action sequence in advance.
[0096] For example, the defect marking Figure 6 middle. Figure 6: is a diagram showing an example of a prediction model according to this embodiment. Figure 6 As shown in the figure, the prediction model is that action 2 appears after action 1 in the action sequence. That is, executing action 2 at this moment may cause defect 1. For example, 4 screws should be screwed in the action after action 1, but if only 3 screws appear, it will cause a defect. Therefore, action 2 can be marked as a defect when only 3 screws appear in the action sequence of actions 1 to 3. Figure 6 Another prediction model shown in FIG. 1 is action 1 in the action sequence shown in the figure. That is, executing action 1 at this moment may cause defect 1. For example, if a fan appears at this moment when it should not appear, it will cause a defect. Therefore, action 1 in which a fan appears in the sequence of actions 1 and 3 can be marked as a defect.
[0097] in addition, Figure 6 It is shown that different prediction models may cause the same defect 1, but it is not limited to this, and other prediction models that cause different defects can also be set.
[0098] In addition, the prediction model is not limited to being obtained by artificial labeling as described above, but can also be obtained through deep learning technology.
[0099] The acquisition unit 102 acquires the actual operation sequence in the same manner as the acquisition unit 12 in the first embodiment, but the description thereof will be omitted here.
[0100] The calculation unit 103 calculates the coincidence rate in the same manner as the calculation unit 13 in the first embodiment, but detailed description is omitted here.
[0101] The prediction unit 105 predicts defects by comparing the actual action sequence with the prediction model stored in the storage unit 101. That is, when the actual action sequence contains an action sequence that is consistent with an action sequence in a prediction model, it is predicted that a corresponding defect will occur.
[0102] The evaluation unit 104 evaluates the operator's operation based on the conformity calculated by the calculation unit 103 and the prediction result of the prediction unit 105. For example, when the prediction unit 105 does not predict a defect, the evaluation is performed in the same manner as the evaluation unit 14 in the first embodiment, and when the prediction unit 105 predicts a defect, the evaluation is directly evaluated as a low level. In addition, when the prediction unit 105 predicts a defect, the evaluation unit 104 gives a prompt. The prompt method may be any one of a display method, a voice method, an alarm method, and the like.
[0103] According to the operation evaluation device of this embodiment, it is possible to provide a more sufficient evaluation basis for the operation evaluation and to predict the potential defects that may occur, which is of great significance for the subsequent product tracing and the like.
[0104] Below, refer to Figure 7 The operation evaluation method of this embodiment will be described. Figure 7 : is a flowchart showing the operation evaluation method according to the present embodiment.
[0105] like Figure 7 As shown, first Figure 4 Similarly to step S11, the acquisition unit 102 acquires an actual action sequence executed in multiple cycles of the process of the current station, and the actual action sequence includes the action content and the time consumed by the action for each action (step S21).
[0106] Then, with Figure 4 Similarly, in step S12, the calculation unit 103 calculates a compliance rate representing the ratio of the operator's execution of operations that comply with the benchmark action sequence based on the action content and consumed time in the actual action sequence acquired by the acquisition unit 12 and the action content and benchmark time in the pre-stored benchmark action sequence. The benchmark action sequence includes the action content and benchmark time required for one or more actions that should be performed by the operator of this workstation within a process cycle (step S22).
[0107] Furthermore, the prediction unit 105 predicts defects by comparing the actual action sequence with at least one prediction model stored in the storage unit 101, each prediction model representing an action in the action sequence that may cause a defect (S23).
[0108] Next, the evaluation unit 14 evaluates the operation of the operator at the own station based on the calculated coincidence rate and the prediction result of the prediction unit (step S24), and ends the process.
[0109] In addition, in the above description, it is assumed that the prediction unit predicts defects independently of the calculation of the compliance rate. However, it can be imagined that these defects are likely to be caused by actions that do not conform to the benchmark action sequence, or it can be said that actions that do not conform to the benchmark action sequence are prone to cause quality problems. Therefore, the above-mentioned defect prediction can also be linked to the compliance rate. Specifically, the prediction unit predicts defects by comparing the actual action sequence with the prediction model only when the compliance rate calculated by the calculation unit is lower than a preset threshold. In this way, the processing volume can be reduced compared to the case where defects are always predicted.
[0110] The embodiments, specific examples and modified examples of the present invention are described above, but these are provided as examples and are not intended to limit the scope of the invention. These embodiments, specific examples and modified examples can be implemented in other various forms, and various omissions, substitutions and changes can be made within the scope of the main purpose of the invention.
[0111] For example, in the above-mentioned embodiment, a laptop computer assembly line is used as an example for explanation. However, the present invention is not limited thereto, and can be applied to any production line. Furthermore, the present invention can also be applied to the training of new operators, etc.
[0112] In addition, in the above-mentioned embodiment, the reference time and the consumed time are expressed in working hours. However, it is not limited thereto, and the start time and the end time may also be used to express them. In this way, when the actual action sequence is data detected and recognized from the video, the position of the action in the video can be found for each action.
Claims
1. An operation evaluation device for evaluating the operation of an operator on a production line, characterized in that: have: A storage unit storing a preset reference action sequence, wherein the reference action sequence includes the action content of each of more than one action that the operator should perform in a process cycle and the reference time required for the action; An acquisition unit, which acquires an actual action sequence performed by the operator in a plurality of cycles of a process, wherein each action in the actual action sequence includes action content and its consumption time in a manner that the action content is repeatedly performed in the plurality of cycles; a calculation unit, which calculates a compliance rate based on the action content and the consumed time in the actual action sequence including the action content repeatedly executed in the plurality of cycles acquired by the acquisition unit and the action content and the reference time in the reference action sequence stored in the storage unit, wherein the compliance rate indicates a ratio of the operations performed by the operator that comply with the reference action sequence in the actual action sequence including the action content repeatedly executed in the plurality of cycles; as well as an evaluation unit that evaluates the operation of the operator in the actual action sequence including the action content repeatedly executed in the plurality of cycles based on the conformity rate of the actual action sequence including the action content repeatedly executed in the plurality of cycles calculated by the calculation unit; The calculation unit calculates a plurality of the coincidence rates at different times; The evaluation unit evaluates the appropriateness of the process arrangement of the workstation corresponding to the operator based on the plurality of conformity rates calculated by the calculation unit; The calculation unit calculates, in each of the multiple cycles, a ratio of action contents whose consumption time conforms to the corresponding reference time as a first operation conformity rate, and calculates a variance of the first operation conformity rate within the multiple cycles; The evaluation unit evaluates the operation of the operator based on the variance of the first operation compliance rate calculated by the calculation unit.
2. The operation evaluation device according to claim 1, wherein: The calculation unit calculates a ratio of the number of action contents in the actual action sequence to the total number of action contents that should be included in the multiple cycles as a second operation compliance rate.
3. The operation evaluation device according to claim 1 or 2, characterized in that: The calculation unit calculates the ratio of the action contents whose consumption time matches the corresponding reference time in the action contents of the plurality of cycles as the third operation matching ratio.
4. The operation evaluation device according to claim 1, wherein: The calculation unit calculates a ratio of the number of times a process conforms to the reference action sequence in the plurality of cycles as a process conformity rate.
5. The operation evaluation device according to claim 4, characterized in that: The process complies with the benchmark action sequence when: The process consumption time obtained according to the consumption time of each action in the process is consistent with the process reference time obtained according to the reference time of each action; or, The consumption time of each action in the process meets the corresponding reference time.
6. The operation evaluation device according to claim 1, wherein: The calculation unit calculates, for each action, a ratio of the number of times the consumed time of the action matches the reference time of the action in the plurality of cycles as a first action matching ratio.
7. The operation evaluation device according to claim 6, characterized in that: The reference action sequence and the actual action sequence further include at least one information of a tool used, a workpiece, an operation position, a container, and an operation direction for each action; The calculation unit further calculates, for each action, a ratio of the number of times the at least one information of the action matches the reference action sequence in the plurality of cycles as a second action matching ratio, The calculation unit weights the first action matching rate and the second action matching rate for each action to obtain an action matching rate.
8. The operation evaluation device according to claim 6, wherein: The calculation unit weights the first action matching rate of each of the one or more actions to obtain an operation matching rate.
9. The operation evaluation device according to claim 1, wherein: The storage unit also stores at least one prediction model, the at least one prediction model respectively representing actions in the action sequence that may cause defects; The operation evaluation device further includes a prediction unit that predicts a defect by comparing the actual action sequence with the prediction model stored in the storage unit; The evaluation unit evaluates the operation of the operator based on the coincidence rate and the prediction result of the prediction unit, and gives a prompt when the prediction unit predicts a defect.
10. The operation evaluation device according to claim 9, characterized in that: The prediction unit predicts a defect by comparing the actual action sequence with the prediction model stored in the storage unit only when the coincidence rate calculated by the calculation unit is lower than a preset threshold value.
11. An operation evaluation method, used in an operation evaluation device for evaluating the operation of an operator on a production line, characterized in that: The operation evaluation device has a storage unit for storing a preset reference action sequence, the reference action sequence including the action content of each of more than one action that the operator should perform in one process cycle and the reference time required for the action; The operation evaluation method includes: An acquisition step of acquiring an actual action sequence performed by the operator in multiple cycles of the process, wherein each action in the actual action sequence includes action content and its consumption time in a manner that the action content is repeatedly performed in the multiple cycles; a calculating step of calculating a compliance rate based on the action contents and the consumed time in the actual action sequence including the action contents repeatedly executed in the plurality of cycles acquired in the acquiring step and the action contents and the reference time in the reference action sequence stored in the storage unit, wherein the compliance rate indicates the ratio of the operations performed by the operator in the actual action sequence including the action contents repeatedly executed in the plurality of cycles that comply with the reference action sequence; and an evaluation step of evaluating the operator's operation in the actual action sequence including the action content repeatedly executed in the plurality of cycles based on the conformity rate of the actual action sequence including the action content repeatedly executed in the plurality of cycles calculated in the calculation step; In the calculation step, a plurality of the coincidence rates at different times are calculated; In the evaluation step, based on the plurality of conformity rates calculated in the calculation step, the appropriateness of the process arrangement of the workstation corresponding to the operator is evaluated; In the calculation step, for each of the multiple cycles, the ratio of the action contents whose consumption time meets the corresponding reference time is calculated as the first operation compliance rate, and the variance of the first operation compliance rate within the multiple cycles is calculated; In the evaluation step, the operation of the operator is evaluated based on the variance of the first operation compliance rate calculated in the calculation step.
12. An operation evaluation system for evaluating the operation of operators at various workstations on a production line, characterized in that: have: The operation evaluation device according to any one of claims 1 to 10, provided for each workstation; A server, communicating with the operation evaluation device via a network, receiving data related to the operator's operation from the operation evaluation device and storing the data; as well as The client accesses the server through the network.
