Skill level evaluation system and work machine control system

Abstract

Provided is a skill level evaluation system that makes it possible to provide appropriate instruction to an operator. The skill level evaluation system comprises: an operation data acquisition unit that acquires operation data of a work machine; a storage unit that stores standard data for comparison with the operation data; and an evaluation unit that evaluates the skill level of an operator of the work machine. The evaluation unit evaluates the skill level on the basis of the difference between the operation data acquired by the operation data acquisition unit and the standard data stored in the storage unit.

Description

Skill evaluation system and work machine control system

[0001] The present disclosure relates to a skill evaluation system and a work machine control system.

[0002] The cited document discloses a construction machine operation system that compares a time-series distribution of predetermined state values related to the operation status of a construction machine with a set target, and outputs a message according to the comparison result (Patent Document 1).

[0003] However, in the above system, since a message is displayed during work, it is difficult for an operator who is concentrating on operation during work to check the message. In addition, since a plurality of messages are displayed simultaneously when operation is interrupted or the like, there is a possibility that an excess of information will occur and appropriate instruction cannot be given to the operator.

[0004] Japanese Unexamined Patent Application Publication No. 2005-098076

[0005] An object of the present disclosure is to provide a skill evaluation system capable of providing appropriate instruction to an operator.

[0006] A skill evaluation system according to one aspect of the present disclosure includes: an operation data acquisition unit that acquires operation data of a work machine; a storage unit that stores reference data for comparison with the operation data; and an evaluation unit that evaluates the skill of an operator of the work machine, wherein the evaluation unit evaluates the skill based on a difference between the operation data acquired by the operation data acquisition unit and the reference data stored in the storage unit.

[0007] A work machine control system according to another aspect of the present disclosure includes: the skill evaluation system described above; and a setting unit that sets control parameters for controlling the work machine, wherein the storage unit stores a plurality of pieces of the reference data, and the setting unit extracts specific data, which is the reference data having characteristics approximating the operation data acquired by the operation data acquisition unit, from among the plurality of pieces of reference data stored in the storage unit, and sets the control parameters based on the extracted specific data.

[0008] Figure 1 is a diagram showing the configuration of a skill evaluation system according to the first embodiment of this disclosure. Figure 2 is a flowchart illustrating the processing in the skill evaluation system of Figure 1. Figure 3A is a diagram showing the case where the score takes the lowest value in interval A. Figure 3B is a diagram showing the case where the score takes the lowest value in interval B. Figure 3C is a diagram showing the case where the score takes the lowest value in interval C. Figure 4 is a diagram showing a method for evaluating the operator's skill. Figure 5A is a diagram illustrating a table for storing messages to be output. Figure 5B is a diagram illustrating the content of a teaching comment. Figure 6 is a diagram illustrating a method for generating a teaching comment by combining words. Figure 7 is a diagram illustrating a presented message. Figure 8 is a flowchart showing the process of presenting key work items using count information. Figure 9A is a diagram illustrating the state of count information. Figure 9B is a diagram illustrating the state of count information. Figure 10 is a diagram illustrating presented key work items. Figure 11A is a diagram showing the configuration of a control system for a work machine according to the second embodiment of this disclosure. Figure 11B is a diagram illustrating the data structure of operation data. Figure 12 is a flowchart illustrating an example of the operation of a control system for a work machine. Figure 13A shows multiple work data. Figure 13B shows work data after the data lengths have been standardized. Figure 13C shows the average value of the work data. Figure 13D shows the average value of the work data and data from a first-order lag system. Figure 14 shows an example of extracting work data from an interval with a high frequency distribution. Figure 15 is a flowchart illustrating the process of updating control parameters. Figure 16 is an example of a user interface screen when specific data can be selected according to the operator's will. Figure 17 is a flowchart illustrating the process related to the evaluation of the operator's skill. Figure 18 is an example of managing to limit the number of normative data to three for each similar condition (identical condition). Figure 19 is an example of a method for assigning control parameters. Figure 20 is an example of a method for assigning control parameters.

[0009] The embodiments of this disclosure will be described below with reference to the attached drawings.

[0010] <First Embodiment> Figure 1 is a diagram showing the configuration of a skill evaluation system according to the first embodiment of this disclosure.

[0011] As shown in Figure 1, the skill evaluation system 10 of this embodiment comprises an operation data acquisition unit 11, an evaluation unit 12, a presentation unit 13, and a storage unit 15. The operation data acquisition unit 11 acquires operation data of the work machine. The evaluation unit 12 evaluates the skill of the operator of the work machine in a series of tasks based on the difference between the operation data acquired by the operation data acquisition unit 11 and reference data for comparison with the operation data. The presentation unit 13 presents information based on the skill evaluated by the evaluation unit 12. The storage unit 15 stores various data. The skill evaluation system 10 is capable of communicating with the control device 50 of the work machine. The control device 50 controls the drive unit (actuator, etc.) of the work machine, which will be described later, and inputs control command signals to the drive unit.

[0012] The evaluation unit 12 evaluates the skill level corresponding to each of the multiple time-series work units that constitute a series of operations performed by the work machine. The presentation unit 13 presents only the information corresponding to the specific work unit for which the evaluation unit 12 had the worst skill evaluation result among the multiple work units when the series of operations is completed. The information presented by the presentation unit 13 when the series of operations is completed includes a message corresponding to the specific work unit. The timing of the presentation of the above information includes the time at the end of the operation or after the operation has finished. This presentation timing is defined as the time at the end of the operation.

[0013] Examples of the types of work involved in this series of operations include, for instance, excavation, lifting and rotating, soil removal, and return rotation when using a hydraulic excavator.

[0014] In this disclosure, the term "working machinery" is optional and includes heavy machinery and construction machinery. It also includes cranes.

[0015] Figure 2 is a flowchart illustrating the processing in the skill evaluation system 10. The following explanation will use the case of performing work with a hydraulic excavator as a work machine as an example. The hydraulic excavator comprises a lower traveling body including a travel device, an upper slewing body supported by the lower traveling body so as to be able to slewing relative to the lower traveling body, and an attachment supported by the upper slewing body. The attachment includes a boom that is mounted on the upper slewing body so as to be able to be raised and lowered, an arm that is mounted on the boom so as to be able to rotatably, and a tip attachment such as a bucket that is mounted on the arm so as to be able to rotatably. The hydraulic excavator further comprises a plurality of actuators that operate by receiving the supply of hydraulic fluid discharged from a hydraulic pump, including a boom cylinder for raising and lowering the boom, an arm cylinder for rotating the arm, a tip attachment cylinder for rotating the tip attachment, a slewing motor for slewing the upper slewing body relative to the lower traveling body, and a travel motor for moving the traveling device. The hydraulic circuit from the hydraulic pump to each actuator constitutes the drive unit of the work machine. The hydraulic circuit includes a control valve that receives a command signal and adjusts the flow rate of hydraulic fluid to each actuator. The command signal is set according to the control parameters (control characteristics) described later.

[0016] For example, suppose the excavation site and the soil removal site are located at different positions in the rotation direction of the upper rotating body. In this case, lifting rotation means the upper rotating body rotates from the excavation site to the soil removal site while holding soil in the bucket, and returning rotation means the upper rotating body rotates from the soil removal site to the excavation site after soil removal.

[0017] In step S102 of Figure 2, the operation data acquisition unit 11 acquires operation data during work using the work machine, for example, from the control device 50 of the work machine. The operation data acquisition unit 11 continues to acquire operation data until the series of operations is completed. The acquired operation data is stored in the storage unit 15 as appropriate.

[0018] Here, the operational data includes, for example, the magnitude of the operator's input (the amount of operation received by the control lever) to operate the boom cylinder, arm cylinder, bucket cylinder, slewing motor, and travel motor of the hydraulic excavator, as well as time-series data of multiple types related to the magnitude of the operation (operating speed) of the boom cylinder, arm cylinder, bucket cylinder, slewing motor, and travel motor, which are the machine output of the hydraulic excavator. In addition, time-series data related to the position (amount of movement) and speed of attachments may also be used as operational data. Depending on the type of work machine, various single or multiple types of time-series data can be used as operational data.

[0019] In step S104, the evaluation unit 12 determines whether the operational data acquired in step S102 is data that is subject to evaluation by the evaluation unit 12. If the determination is affirmative, the process proceeds to step S106; if the determination is negative, the process ends.

[0020] Here, the evaluation unit 12 determines the corresponding work type from the operational data acquired in step S102 and determines whether the identified work type is subject to evaluation by the evaluation unit 12. The work type can be determined, for example, based on lever operation patterns or using a classifier built with machine learning based on past operational data.

[0021] In step S106, the evaluation unit 12 sequentially evaluates the time-series skill of the operator of the work machine for a series of tasks based on the difference between the operational data acquired in step S102 and the normative data. The evaluation unit 12 calculates the skill evaluation result as a score. A higher score indicates a better skill evaluation result. For example, the evaluation unit 12 evaluates the score higher the smaller the difference between the operational data acquired in step S102 and the normative data. Alternatively, the evaluation unit 12 may calculate the similarity (norm) between the operational data acquired in step S102 and the normative data, and evaluate the score higher the higher this similarity.

[0022] Here, the normative data is, for example, operational data such as the magnitude of the operator's movements, the magnitude of the hydraulic excavator's movement, and the position and speed of the attachment, obtained from work performed by a skilled operator. The skilled operator's work described above corresponds to the same type of work as the type of work for which the operational data determined to be subject to evaluation by the evaluation unit 12 in step S104 is to be evaluated. The normative data is stored, for example, in the storage unit 15 or in other storage means and is used for evaluation by the evaluation unit 12.

[0023] The evaluation unit 12 recognizes complex operational data based on each of the multiple types of data included in the operational data, such as the magnitude of the operator's operation, the magnitude of the hydraulic excavator's movement, and the position (amount of movement) and speed of the attachment. It also recognizes complex normative data based on each of the multiple types of data included in the operational data, such as the magnitude of the operator's operation, the magnitude of the hydraulic excavator's movement, and the position (amount of movement) and speed of the attachment. The evaluation unit 12 then sequentially evaluates the time-series skill of the operator of the work machine for a series of tasks based on the difference between the complex operational data and the complex normative data. "Complex data" here refers to data that comprehensively represents the work state by combining multiple types of data (operation amount, machine movement amount, position, speed, etc.) at the same time or within a certain period of time.

[0024] In step S107, the evaluation unit 12 determines whether the work of the corresponding work type has been completed. If the determination is affirmative, the process proceeds to step S108; if the determination is negative, the process proceeds to step S102.

[0025] In step S108, the evaluation unit 12 determines whether the average value or the minimum value of the scores calculated in step S106 is below a predetermined threshold. If the determination is positive, the process proceeds to step S110; otherwise, the process ends.

[0026] In step S110, the presentation unit 13 searches for the time when the score is at its minimum. This process corresponds to the process of searching for the time when the skill evaluation result by the evaluation unit 12 was the worst.

[0027] Figures 3A, 3B, and 3C are examples of time series for operation data and scores, respectively. Curve 60 shows the time series of operation data starting from the start of work, acquired in step S102. Curve 62 shows the time series of scores starting from the start of work, acquired in step S106. Figure 3A shows the case where the score takes its lowest value in interval A, Figure 3B shows the case where the score takes its lowest value in interval B, and Figure 3C shows the case where the score takes its lowest value in interval C.

[0028] Figures 3A and 3C show an example of dividing a series of operations into three chronologically arranged operation units, with each operation unit corresponding to section A, section B, and section C. Here, for example, section A is the initial section from the start of the operation until the actuator reaches a steady speed, section B is the middle section where the actuator maintains a steady speed, and section C is the later section from when the actuator decelerates from a steady speed until it moves on to the next operation.

[0029] In the example in Figure 3A, the time ta, when the score is lowest, belongs to interval A. In this case, the presentation unit 13 extracts time ta as the time when the skill evaluation result by the evaluation unit 12 was the worst. In the example in Figure 3B, the time tb, when the score is lowest, belongs to interval B. In this case, the presentation unit 13 extracts time tb as the time when the skill evaluation result by the evaluation unit 12 was the worst. Similarly, in the example in Figure 3C, the time tc, when the score is lowest, belongs to interval C. In this case, the presentation unit 13 extracts time tc as the time when the skill evaluation result by the evaluation unit 12 was the worst. If there are multiple times with the lowest score, the evaluation unit 12 extracts the time with the lowest score first among the multiple times as the time with the worst evaluation result.

[0030] In step S111, a comparison is made between each of the multiple types of operational data and each of the multiple types of normative data at the time when the score identified in step S110 takes its lowest value.

[0031] Figure 4 shows a method for comparing each of the multiple types of operational data with each of the multiple types of normative data.

[0032] In Figure 4, the vertical axis of the graph shows the values ​​of the operational data, and the horizontal axis shows the time starting from the start of the work. Curve 61 shows the time series of one of the multiple reference data sets, and curve 60 shows the time series of one of the multiple operational data sets acquired in step S102. The operational data shown in curve 60 and the reference data shown in curve 61 are time series of either operation inputs such as boom raising operations, arm pulling operations, and bucket excavation operations, or machine outputs such as boom cylinder speed, arm cylinder speed, and bucket cylinder speed.

[0033] The evaluation unit 12 calculates the difference between each of the operational data acquired in step S102 and the respective reference data used for comparison with each operational data at time t when the score identified in step S110 takes its lowest value.

[0034] For example, the evaluation unit 12 calculates the difference between each of the multiple types of operational data and each of the normative data. Here, for example, the skill may be evaluated based on the difference e(t) in a single type of operational data. When performing the evaluation, the difference e(t) related to the operational data itself may be associated with the skill. In this case, the smaller the difference e(t), the better the evaluation result. Alternatively, when performing the evaluation based on the difference e(t) in multiple types of operational data, the similarity (norm) between the multiple types of operational data and the corresponding multiple types of normative data may be calculated based on the difference e(t) in the multiple types of operational data, and this similarity may be associated with the skill for evaluation. In this case, the higher the similarity, the better the evaluation result.

[0035] In step S112, the presentation unit 13 generates a teaching comment based on the calculation result in step S111.

[0036] The following describes a specific example of how to generate instructional comments in step S112.

[0037] Figure 5A is an example of a table showing the correspondence between calculation results and messages to be output. The table is stored, for example, in the storage unit 15, and the presentation unit 13 recognizes the table by referring to it from the storage unit 15. In the example of Figure 5A, messages A1, A2, and A3 are messages to be output when the time at which the score reaches its lowest value falls within interval A. Messages B1 and B2 are messages to be output when the time at which the score reaches its lowest value falls within interval B, and messages C1, C2, and C3 are messages to be output when the time at which the score reaches its lowest value falls within interval C.

[0038] Each message that may be determined to be output has predetermined conditions for determining whether it should be output. Specifically, these predetermined conditions are stored in separate frames for each row containing each message in the table, which are divided into sections for intervals, operation inputs, and aircraft outputs. Within the interval frame in the table, one of intervals A to C is stored. Interval A stored within the interval frame indicates that the time when the score reached its lowest value belongs to interval A, interval B stored within the interval frame indicates that the time when the score reached its lowest value belongs to interval B, and interval C stored within the interval frame indicates that the time when the score reached its lowest value belongs to interval C. Furthermore, for each row containing each message in the table, which is divided into sections for operation inputs and aircraft outputs, a black circle is placed inside the frame if the condition that the operation input or aircraft output at the time when the score reached its lowest value is insufficient or excessive is present, and the black circle is not placed inside the frame if the condition is not present, thus indicating that the predetermined conditions are stored (set). Here, the conditions that the operation input or machine output is insufficient or excessive correspond to the case where the difference between the operational data and the reference data exceeds a predetermined threshold (the difference is large). Specifically, the condition that the machine output is excessive corresponds to the case where the value of the operational data exceeds the value of the reference data and the difference between the operational data and the reference data exceeds a predetermined threshold, and the condition that the operation input or machine output is insufficient corresponds to the case where the value of the operational data falls below the value of the reference data and the difference between the operational data and the reference data exceeds a predetermined threshold. As shown in Figure 5A, predetermined conditions regarding the difference between the operational data and the reference data include the difference in operation input in boom operation, arm operation, and bucket operation (difference from the reference data), and the difference in boom cylinder speed, arm cylinder speed, and bucket cylinder speed as machine output (difference from the reference data).

[0039] The table indicates that predetermined conditions for determining which message should be output are met when all predetermined conditions for the interval, operation input, and machine output are met. For example, in message A1, the condition that the time when the score reached its lowest value belongs to interval A is stored in the interval frame, the black circle 811 representing the condition that boom operation (boom raising operation) is insufficient is stored in the operation input frame, and the black circle 812 representing the condition that boom cylinder speed is insufficient is stored in the machine output frame. In other words, the table indicates that message A1 should be output when the time when the score reached its lowest value belongs to interval A, boom operation (boom raising operation) is insufficient, and boom cylinder speed is insufficient. In message A2, the condition that the time when the score reached its lowest value belongs to interval A is stored in the interval frame, the black circle 821 representing the condition that arm operation (arm pulling operation) is insufficient is stored in the operation input frame, and the black circle 822 representing the condition that arm cylinder speed is insufficient is stored in the machine output frame. In other words, the table indicates that message A2 should be output if the time when the score is at its lowest falls within interval A, there is insufficient arm operation (arm pulling operation), and the arm cylinder speed is insufficient.

[0040] Message B1 stores the condition that the time when the score reached its lowest value belongs to interval B within the interval frame, the black circle 831 representing the condition that the boom operation (boom raising operation) is excessive within the operation input frame, the black circle 832 representing the condition that the boom cylinder speed is excessive within the machine output frame, and the black circle 833 representing the condition that the arm cylinder speed is excessive within the machine output frame. In other words, the table indicates that message B1 should be output when the time when the score reached its lowest value belongs to interval B, the boom operation (boom raising operation) is excessive, the boom cylinder speed is excessive, and the arm cylinder speed is excessive.

[0041] Figure 5B is a diagram illustrating the content of the instructional comments indicated by each message (Messages A1 to A3, Messages B1 to B2, Messages C1 to C3).

[0042] The instruction comments shown in Figure 5B are advice indicating areas for improvement to reduce the differences in operation input and machine output compared to the reference data, as shown in Figure 5A. The left frame of the instruction comments stores the time-series position of the work unit (sections A to C), with "In the initial stages of excavation," corresponding to section A, "During excavation," corresponding to section B, and "Towards the end of excavation," corresponding to section C. The right frame of the instruction comments stores advice indicating areas for improvement. The advice corresponds to conditions where the operation input or machine output is insufficient or excessive. For example, the instruction comment 81a for message A1 in Figure 5B, "In the initial stages of excavation, you have penetrated too far. A little more...", is advice indicating areas for improvement to prevent at least one of the conditions indicated by the black circle 811, which represents insufficient boom operation (boom raising operation), and the black circle 812, which represents insufficient boom cylinder speed. Similarly, the instruction comment 82a for message A2 in Figure 5B, "During the initial stage of excavation, excavation is slow. More...", is advice indicating improvements to prevent at least one of the following from being marked: black circle 821, which indicates insufficient arm operation (arm pulling operation); and black circle 822, which indicates insufficient arm cylinder speed. The instruction comment 83a for message B1 in Figure 5B, "During excavation, boom raising operation is excessive. Gentle operation...", is advice indicating improvements to prevent at least one of the following from being marked: black circle 831, which indicates excessive boom operation (boom raising operation); black circle 832, which indicates excessive boom cylinder speed; and black circle 833, which indicates excessive arm cylinder speed.

[0043] In step S112, the presentation unit 13 respectively determines which of sections A to C the time at which the score reaches the minimum value belongs to, whether the operation input is insufficient or excessive, and whether the machine body output is insufficient or excessive. Further, the presentation unit 13 refers to the table in FIG. 5A to determine whether the determination result satisfies any of the predetermined conditions for determining the message to be output shown in the table. Then, the presentation unit 13 determines the message for which the determination result is affirmative as the message to be output, and generates the teaching comment 81a based on the teaching comment in the determined message. For example, when the presentation unit 13 determines that the time at which the score reaches the minimum value belongs to section A, the boom operation (boom raising operation) is insufficient, and the boom cylinder speed is insufficient, the presentation unit 13 determines the message A1 that satisfies the predetermined condition corresponding to the determination result as the message to be output, and generates the teaching comment 81a, which is the content of the teaching comment in message A1, as shown in FIG. 5B.

[0044] In FIGS. 5A and 5B, operations related to booms, arms, and buckets, and cylinder speeds are illustrated as targets of operation data (time-series data) for calculating differences from reference data. However, it may also be assumed that teaching comments based on differences from reference data are generated for time-series data such as the combined center-of-gravity speed of the attachment representing the movement of the machine body during work and the tip coordinate position (such as height) of the attachment. The "teaching comment based on a difference from reference data" mentioned herein is a comment including improvement instructions for reducing deviations in operation input and machine body output obtained by comparing operation data with reference data. For example, when the operation input or machine body output in the operation data is insufficient compared with the reference data, a comment such as "Please increase the operation amount" is generated. The combined center-of-gravity speed of the attachment and the tip coordinate position of the attachment can be calculated, for example, based on information from angle sensors attached to each part of the work machine or information from the control device 50.

[0045] Furthermore, although FIGS. 5A and 5B use a plurality of time-series data as targets for calculating differences from reference data, evaluation based on the difference of one piece of time-series data may be performed. For example, the aforementioned attachment combined centroid velocity and the tip coordinate position of the attachment can be time-series data that more directly indicates the quality and efficiency of work, compared to operations related to booms, arms, and buckets and cylinder velocities. Therefore, depending on the type of work, etc., differences from reference data may be calculated and evaluated only for the attachment combined centroid velocity or the tip coordinate position of the attachment.

[0046] FIG. 6 illustrates a method of generating teaching comments by combining words.

[0047] In the example of FIG. 6, words that specify the type of work (e.g., "excavation", "grading"), words that specify the operation (e.g., "boom up", "arm draw"), words that indicate the improvement direction (e.g., instructions for increase / decrease), and words that indicate the degree (e.g., "greatly", "slightly") are prepared. Then, teaching comments are generated by combining these words. The teaching comments shown in FIG. 5B can also be generated by the same method.

[0048] Furthermore, instead of combining words, a comment generator constructed by machine learning or the like may be used, and teaching comments may be generated by inputting evaluation results into the comment generator.

[0049] In step S114 of FIG. 2, the presentation unit 13 presents a message (information) including a teaching message corresponding to the teaching comment for the operator based on the teaching comment generated in step S112.

[0050] FIG. 7 is a diagram illustrating an example of a message as a teaching comment presented by the presentation unit 13.

[0051] The message is presented to the operator, for example, by being displayed on a monitor screen in the operator's cab. Furthermore, the message may be conveyed by voice instead of, or in addition to, displaying the message. That is, the presentation unit 13 outputs the message to a display device having a monitor screen mounted in the operator's cab of the work machine, or to a speaker that outputs sound and is mounted in the operator's cab of the work machine. If the work machine is remotely controlled, the presentation unit 13 may output the message to a display device or speaker located outside the work machine.

[0052] In the example shown in Figure 7, the generated instruction comments are displayed in area 71 as instruction messages indicating specific areas for improvement, and the score calculated and output in step S106 is displayed in area 72 as the skill score.

[0053] Here, the presentation unit 13 presents only messages containing a teaching message corresponding to the specific work unit in which the skill evaluation result by the evaluation unit 12 was the worst among the multiple work units, which are sections A to C. This teaching message indicates the work unit to which the time with the lowest skill evaluation score in the series of tasks belongs, i.e., the specific work unit (one of sections A to C). This teaching message also functions as advice to the operator on how to improve the score in the specific work unit.

[0054] For example, the teaching message in region 71 shown in Figure 7 corresponds to the case where the difference in section B corresponding to "During excavation" (for example, the difference related to boom raising operation) is the largest. In this case, the teaching message "During excavation, the boom raising operation is excessive. Adjust the operation amount..." is displayed in region 71. This teaching message corresponds to the teaching comment 83a shown in Figure 5B.

[0055] Thus, in this embodiment, when a series of tasks is completed, only information including a teaching message for the specific task unit to which the time when the evaluation result of the evaluation unit 12 for skill was worst belongs is presented. As a result, only the most important teaching message is presented. In other words, information overload caused by the simultaneous presentation of multiple teaching messages can be prevented, and teaching can be focused on the points where the operator is most weak, thereby promoting efficient skill improvement.

[0056] Furthermore, if the evaluation unit 12's judgment in step S108 of Figure 2 is rejected, the system may be configured to display the skill score in area 72 instead of displaying the instruction comment in area 71.

[0057] Next, in step S116 of Figure 2, the evaluation unit 12 stores the evaluation information, which associates the evaluation results with the operator and time information, in the storage unit 15, and terminates the process.

[0058] Here, the evaluation information includes a message number that identifies the teaching message presented in area 71 (Figure 7) by the presentation unit 13, and a skill score presented in area 72 (Figure 7) by the presentation unit 13.

[0059] Furthermore, the evaluation information can broadly include evaluation results such as instructional comments (Figure 5B) that are not included in the instructional messages shown in Figure 7. By storing evaluation results other than instructional messages as evaluation information in the storage unit 15, the evaluation unit 12 can present various evaluation results, including instructional comments other than instructional messages, to the operator in a timely manner. This allows the operator to examine in detail their own skill-related problems, for example, making it possible to effectively use the evaluation information to improve their skills.

[0060] Furthermore, the evaluation information may also include the operation data acquired by the operation data acquisition unit 11.

[0061] Furthermore, count information indicating the number of times the teaching message presented by the presentation unit 13 has been presented may be added to the evaluation information.

[0062] In this embodiment, when a series of operations is performed multiple times, the display unit 13 displays the instruction messages that are displayed with high frequency as key focus items (important information).

[0063] Figure 8 is a flowchart showing the process of presenting priority areas of focus using count information.

[0064] In step S202 of Figure 8, the presentation unit 13 obtains count information included in the corresponding operator evaluation information from the storage unit 15.

[0065] Figures 9A and 9B illustrate the state of the count information.

[0066] As shown in Figure 9A, the count information is information that associates a message number that identifies a teaching message with a count number. The count number indicates the number of times the teaching message corresponding to the message number has been presented in the past. In other words, the count information is updated each time a series of operations is completed and a teaching message is presented (step S114 in Figure 2).

[0067] In step S204 of Figure 8, the presentation unit 13 determines, based on the acquired count information, whether there is a message number whose count has reached a threshold (for example, 10). If the determination is affirmative, the process proceeds to step S206; if the determination is negative, the process ends.

[0068] In step S206, the presentation unit 13 determines whether or not an improvement in skill corresponding to the teaching message for the message number whose count has reached a threshold (for example, 10) has been observed. For example, when a series of tasks is performed multiple times, the presentation unit 13 determines whether or not the difference e(t) between the score and the normative data at the time the teaching message presented as a key area of ​​focus is presented falls below a predetermined threshold. If this determination result is positive for a predetermined number of times, the presentation unit 13 determines that an improvement in skill has been achieved. If the determination is positive, the presentation unit 13 proceeds to step S208; if the determination is negative, it proceeds to step S210. Specific examples of this determination will be described later.

[0069] In step S210, the presentation unit 13 presents the teaching message corresponding to the message number whose count has reached the threshold as a priority item, and then terminates the process.

[0070] Figure 10 is a diagram illustrating the key initiatives presented.

[0071] In the example shown in Figure 10, the message presented in step S114 (Figure 2) includes key action items, and the key action items are displayed in area 75 shown in Figure 10.

[0072] In this example, in Figure 9A, the instruction message corresponding to message number "5" is highlighted as a priority item by using a different font color and background color than the instruction messages that do not correspond to message number "5". In the example in Figure 9A, the threshold used as the basis for judgment in step S204 is 10, and the count for message number "5" has reached 10. Therefore, the instruction message corresponding to message number "5" is displayed as a priority item.

[0073] Furthermore, multiple thresholds may be set for the count, and the way in which priority items are highlighted may be changed depending on the threshold reached.

[0074] The timing of presenting the priority areas of focus is at the discretion of the user. For example, displaying them at all times would ensure that users are always aware of them.

[0075] In step S206 described above, for example, the presentation unit 13 makes a judgment on improvement based on the evaluation information. For example, the judgment may be affirmed if the evaluation result of the skill by the evaluation unit 12 has improved, such as when the teaching message (for example, the teaching message corresponding to message number "5") has been presented only infrequently for a certain period of time.

[0076] As described above, if step S206 is affirmed, the process proceeds to step S208.

[0077] In step S208, the display unit 13 resets the count of the message number whose count has reached a threshold (for example, 10) and terminates the process. As shown in Figure 9B, the count reset makes the count of that message number (for example, message number "5") temporarily zero. Note that Figure 9B shows the state after the process in Figure 2 has been performed following the count reset, and several counts (counts other than message number "5") have been updated.

[0078] If step S206 is affirmed, the presentation unit 13 may present a message indicating that the evaluation result of the skill by the evaluation unit has improved. In this case, for example, the message may be generated using some of the words that make up the teaching comment shown in Figure 6. For example, using some of the words that make up the teaching comment, a comment supporting the operator's operation, such as "You did a good job raising the boom for 'excavation'," may be generated and presented. This can lead to an improvement in the operator's motivation by clearly showing that the improvements attempted based on the teaching message have yielded good results as intended.

[0079] Thus, in this embodiment, the items that are presented with the most frequent instruction messages and where no improvement is observed are presented to the operator as priority areas for work. For example, consider a scenario where multiple types of instruction messages are presented to the operator due to operational variability during long work periods or the diversity of work content. In this case, it can be difficult to determine which instruction messages should be given priority. However, in this embodiment, particularly important matters are presented as priority areas for work, making it possible to clarify the matters that the operator should prioritize and focus on.

[0080] As described above, in the above embodiment, only the message corresponding to the work unit with the largest difference among the multiple work units is presented when the series of tasks is completed. Therefore, information overload caused by the simultaneous presentation of multiple teaching messages can be prevented, and appropriate teaching can be provided to the operator at the appropriate time.

[0081] <Second Embodiment> Next, a second embodiment of the present disclosure will be described. Figure 11A is a diagram showing the configuration of the control system of the work machine of this embodiment.

[0082] As shown in Figure 11A, the control system 10 of the work machine in this embodiment includes an operation data acquisition unit 11 that acquires operation data of the work machine, a storage unit 120 that stores a plurality of reference data, and a characteristic setting unit 130 (setting unit) that determines the characteristics of the control device 50 of the work machine. The control device 50 is configured to collect operation data from the work machine and transmit it to the operation data acquisition unit 11. In addition, a transmission unit that collects operation data from the work machine and transmits it to the operation data acquisition unit 11 may be provided separately from the control device 50.

[0083] Furthermore, the control system 10 for the work machine includes a productivity evaluation unit 14 that evaluates productivity based on the operational data acquired by the operational data acquisition unit 11.

[0084] Furthermore, the control system 10 for the work machine includes a presentation unit 150 that presents the characteristics of a plurality of normative data extracted by the characteristic setting unit 130, a reception unit 16 that accepts an operation to select one specific data from a plurality of specific data whose characteristics have been presented by the presentation unit 150, and a skill evaluation unit 17 (evaluation unit) that evaluates the operator's skill based on the operation data acquired by the operation data acquisition unit 11 and the specific data extracted by the characteristic setting unit 130.

[0085] As shown in Figure 11A, data necessary for the operation of the control system 10 and data generated by the control system 10 are stored in a storage device 20 that can communicate with the control system 10. Note that the storage unit 120 shown in Figure 11A may be configured as part of the storage device 20. Alternatively, the storage unit 120 or the storage device 20 may be configured using an external database or a storage area provided in the control device 50 of the work machine.

[0086] Figure 11B illustrates the data structure of operational data. In this example, the operational data includes time-series work data that can be represented by representative characteristics, including responsiveness, decay, or productivity.

[0087] In the example shown in Figure 11B, the operational data includes machine ID, work ID, work type, date and time information, soil type, weather, responsiveness, damping characteristics, productivity, work data, and control parameters.

[0088] The machine ID is information that identifies the type of work machine and individual machines. In this disclosure, the type of work machine is arbitrary, and work machine is a concept that includes heavy machinery and construction machinery. Work machines also include hydraulic excavators and cranes. A hydraulic excavator as a work machine comprises a lower traveling body including a traveling device, an upper slewing body supported by the lower traveling body so as to be rotatable relative to the lower traveling body, and an attachment supported by the upper slewing body. The attachment includes a boom mounted to the upper slewing body so as to be able to be raised and lowered, an arm mounted to the boom so as to be able to rotate, and a tip attachment such as a bucket mounted to the arm so as to be able to rotate. The hydraulic excavator further comprises a plurality of actuators that are operated by the supply of hydraulic fluid discharged from a hydraulic pump, including a boom cylinder for raising and lowering the boom, an arm cylinder for rotating the arm, a tip attachment cylinder for rotating the tip attachment, a slewing motor for slewing the upper slewing body relative to the lower traveling body, and a traveling motor for traveling the traveling device.

[0089] The work ID is information that identifies operational data for each type of work, such as excavation or land leveling.

[0090] The work type is information that indicates the type of work, such as excavation or leveling.

[0091] Date and time information indicates the date and time the work was performed. Soil type information indicates the soil type (e.g., hardness) suitable for the work, such as excavation or leveling. Weather information indicates the weather conditions at the time of the work (e.g., sunny, rainy).

[0092] Responsiveness and damping characteristics are parameters that indicate the characteristics (dynamic characteristics) of the data in the corresponding operation performed by the operator operating the work machine. In this embodiment, the characteristics of work data and normative data can be uniquely represented by responsiveness and damping characteristics, making it easier to select normative data (specific data) that is suitable for the work data of the operator being supported. Note that parameters other than responsiveness and damping characteristics may be used as parameters that indicate the characteristics of work data and normative data.

[0093] Productivity is information that indicates the results for each corresponding task, and includes, for example, the amount of soil excavated by the work machine and the flatness of the ground surface after leveling work. Productivity can be evaluated by the productivity evaluation unit 14.

[0094] Work data is data related to the operator's operations (work), and includes, for example, the operator's operations (operations on the control levers) to operate the boom cylinder, arm cylinder, bucket cylinder, slewing motor, and travel motor of a hydraulic excavator, as well as parameters such as the operation (operating speed) of the boom cylinder, arm cylinder, bucket cylinder, slewing motor, and travel motor of the hydraulic excavator, the bucket cylinder speed, and the pump pressure of each part. In addition, time-series data related to the position and speed of the attachment may be used as operational data. Work data is time-series data of the values ​​of these parameters. That is, as operational data in Figure 11B, responsiveness and damping characteristics are stored so as to be associated with the work data, which is time-series data. Similarly, productivity, which is evaluated in relation to the work data, is also associated.

[0095] Control parameters are information that indicates the characteristics of the control device 50, and are set for the control device 50 by the characteristic setting unit 130. By setting control parameters for the control device 50, the operation of the work machine, which operates based on the operator's input, is changed to characteristics corresponding to the set control parameters, thereby providing operational support to the operator. For example, operational support is provided by changing the response speed of the work machine's operation to the operator's input, or by changing the magnitude of the work machine's operating speed in response to the operator's input. In other words, in the operation data of Figure 11B, appropriate control parameters for supporting the work represented by the work data are stored, associated with the work data, which is time-series data.

[0096] Furthermore, the operational data may include an operator ID to identify the operator, and the operator ID may be associated with data characteristics (representative characteristics that represent the data characteristics of that operator).

[0097] Next, the operation of the control system 10 of the work machine will be described.

[0098] The characteristic setting unit 130 extracts specific data from among the multiple reference data stored in the storage unit 120 that have characteristics similar to the operation data acquired by the operation data acquisition unit 11, and sets the characteristics of the control device 50 by setting the control parameters in the control device 50 based on the extracted specific data. In other words, the specific data has characteristics similar to the operation data acquired by the operation data acquisition unit 11.

[0099] Figure 12 is a flowchart showing an example of the operation of a control system for a work machine.

[0100] In step S102 of Figure 12, the operation data acquisition unit 11 acquires operation data related to a specific task performed by a work machine that operates based on the operator's input, and stores it in the storage unit 120.

[0101] Here, a specific task is, for example, a task belonging to a particular type of work, such as excavation at a fixed location. The type of work can be determined, for example, based on operation patterns as data related to the operator's actions (tasks), or using a task classifier constructed using machine learning or the like. The data related to the operator's actions mentioned above is indicated by the task data included in the operational data.

[0102] In step S104, the operational data acquisition unit 11 determines whether operational data for the specific task has already been acquired a predetermined number of times in step S102. If this determination is affirmative, the operational data acquisition unit 11 proceeds to step S106; otherwise, it proceeds to step S102.

[0103] Here, the predetermined number of times can be a set number, but it may also be set according to the instructions of the operator operating the work machine. For example, the operator may instruct the start and end of acquiring operational data (work data) by operating a switch or the like at the start and end of the work.

[0104] In step S106, the characteristic setting unit 130 performs an averaging process for the work data included in the operational data acquired in step S102 and stored in the storage unit 120.

[0105] Here, we will explain an example of the process of averaging work data.

[0106] Figure 13A shows multiple work data sets, Figure 13B shows work data after the data lengths have been standardized, Figure 13C shows the average value of the work data, and Figure 13D shows the average value of the work data and data from a first-order lag system. Figure 14 shows an example of extracting work data from an interval with a high frequency distribution.

[0107] As shown in Figures 13A and 13B, if the data lengths (working time) of the work data included in multiple operational data acquired by the operational data acquisition unit 11 differ from one another, the data lengths may be adjusted according to the type of work data (e.g., boom operation amount). In the examples in Figures 13A and 13B, the data lengths are adjusted to the shortest data length Lmin by excluding a portion of the data 61. In these figures, the vertical axis of the graph shows the value of the work data, and the horizontal axis shows the time starting from the start of work. The values ​​of the work data include, for example, the magnitude of the operator's operation on the hydraulic excavator (amount of operation received by the operating lever), the magnitude of the hydraulic excavator's movement (operating speed), and the position of the attachment (amount of movement).

[0108] Alternatively, as shown in Figure 14, the work data contained in each of the multiple operational data sets can be classified according to their data length Lmin, and the frequency N for each data length Lmin can be aggregated to extract only the data with a high frequency distribution. The average value can then be calculated and approximated using a first-order lag system. In the example in Figure 14, only the work data for tasks with the highest frequency N and a work time in the range of 3.5 to 4.0 seconds is extracted.

[0109] Next, the representative data 62 shown in Figure 13C is calculated by calculating the average value of the work data.

[0110] Furthermore, if only one operational data is acquired by the operational data acquisition unit 11, the averaging process for the work data is unnecessary, and the work data value included in that operational data is used as the representative data 62.

[0111] In step S108 of Figure 12, the characteristic setting unit 130 searches for and extracts normative data similar to the representative data 62. Here, as normative data, for example, operational data acquired in advance by the operational data acquisition unit 11 and stored in the storage unit 120, which pertains to a specific task performed by a work machine operated based on the operations of a skilled operator in the past (step S102), can be used. In this case, the normative data is extracted from data that shares at least the same type of work as the operational data (representative data) being compared in this case. Furthermore, normative data whose responsiveness and damping (characteristics) are similar to the operator's work data in this case is extracted.

[0112] As normative data, a single operational data point may be used, or, similar to representative data 62, data obtained by averaging operational data (work data) based on multiple operations (representative data) may be used.

[0113] The characteristic setting unit 130, for example, when the representative data 62 and the normative data are approximated by data from a first-order lag system, extracts the normative data that the data approximates.

[0114] Figure 13D shows the case where representative data 62 is approximated by data 63 from a first-order lag system. The first-order lag system data 63 shown in Figure 13D corresponds to data whose responsiveness and attenuation match those of representative data 62. Note that the responsiveness is a time constant T, and the attenuation characteristic is a value of 1 or greater.

[0115] In this case, the characteristic setting unit 130 can extract one or more reference data that approximates the representative data 62, i.e., reference data in which the work data approximates the data 63 of a first-order lag system, as reference data that approximates the representative data 62. In this embodiment, work data that approximates the representative data 62 is searched for via first-order lag system data corresponding to responsiveness and attenuation, but the method for searching for approximate reference data is arbitrary.

[0116] In the extraction of normative data (step S108), substantial similarity of the operational data can be ensured by considering not only the similarity of the work data but also the commonality of other information contained in the operational data. For example, only normative data with common soil type, weather (Figure 11B), etc., in the operational data may be extracted, or only normative data with similar control parameters (Figure 11B) may be extracted. When extracting normative data, commonality of work dates and times (for example, commonality of time zones) may also be considered.

[0117] Furthermore, in the above example, the normative data is basically based on the operational data of skilled operators, but productivity (Figure 11B) may also be considered when extracting normative data (operational data). That is, priority may be given to those with good productivity, or the extraction may be limited to those with good productivity. In addition, when extracting normative data, the aircraft type indicated by the aircraft ID (Figure 11B) and the commonalities and similarities of individual aircraft may also be considered.

[0118] In step S110, the characteristic setting unit 130 determines whether or not there is normative data similar to the representative data 62. If the determination is affirmative, the process proceeds to step S112; otherwise, the process ends. This determination is affirmed in step S108 when normative data similar to the representative data 62 is extracted, for example, when normative data with a similarity value exceeding a predetermined threshold is found.

[0119] In step S112, the characteristic setting unit 130 calculates the degree of similarity between the extracted normative data and the representative data 62.

[0120] The degree of similarity between the extracted normative data (A(1), A(2), A(3), ..., A(N)) and the representative data 62 (A'(1), A'(2), A'(3), ..., A'(N)) can be evaluated using the norm L2(N) by equation (1).

[0121] Here, N is the data length, and A and A' represent the work type (Figure 11B). There may be multiple work types, and multiple work types, such as B and B' or C and C', may be added as targets for evaluation using the norm L2(N).

[0122] In step S114, the characteristic setting unit 130 sets the control parameters to be set in the control device 50 based on specific data.

[0123] In this case, if only one normative data is extracted, this normative data can be designated as specific data, and the control parameters (Figure 11B) attached to the normative data can be directly set in the control device 50.

[0124] Furthermore, if there are multiple extracted normative data (specific data), the control parameters may be modified by weighting each of the control parameters attached to the normative data according to the degree of approximation (norm L). For example, if three normative data are used, in equation (2), the numbers 1 to 3 in the subscripts attached to norm L and control gain Kp correspond to each of the normative data. The same applies to control parameters other than control gain Kp. In equation (2), the subscript kihan means normative.

[0125] In this embodiment, the operation data acquisition unit 11 can acquire operation data of a work machine whose control characteristics of the control device 50 have been set based on specific data by the characteristic setting unit 130. Furthermore, the productivity evaluation unit 14 evaluates productivity based on the operation data of the work machine whose control characteristics have been set, and the characteristic setting unit 130 can extract new specific data and update the characteristics of the control device 50 according to the evaluation result by the productivity evaluation unit 14.

[0126] For example, the productivity evaluation unit 14 can evaluate the productivity after the control characteristics of the control device 50 have been set based on specific data, based on the amount of soil excavated and the progress of the work. In this case, if the evaluation result has not reached a desired state (criteria), the characteristic setting unit 130 may update the control parameters applied to the control device 50 based on new specific data (operational data) stored in the storage unit 120.

[0127] Figure 15 is a flowchart illustrating the process of updating control parameters.

[0128] In step S202 of Figure 15, the productivity evaluation unit 14 evaluates the productivity after the control characteristics of the control device 50 have been set based on specific data, based on the amount of soil excavated and the progress of the work.

[0129] In step S204, the productivity evaluation unit 14 determines whether the productivity evaluation result meets the criteria. If the determination is affirmative, the process proceeds to step S202; if the determination is negative, the process proceeds to step S206.

[0130] In step S206, the characteristic setting unit 130 updates the control parameters applied to the control device 50 based on the new specific data (operational data) stored in the storage unit 120, and proceeds to step S202. Here, for example, the control parameters applied to the control device 50 based on the new specific data (operational data) can be updated by performing the same processing as in steps S108 to S114 in Figure 12. Alternatively, in step S112 in Figure 12, new control parameters may be set (updated) based on operational data that has been found to have a high degree of approximation but was not adopted as a control parameter in step S114 (for example, operational data with the second highest degree of approximation). Notifying the operator of the update of the control parameters can avoid causing any operational discomfort.

[0131] By updating control parameters, even if the initially applied parameters did not provide effective support, the likelihood of ultimately applying control parameters suitable for the operator increases. Furthermore, it becomes possible to provide support that flexibly adapts to the operator's physical condition and work situation. It also becomes possible to update control parameters in real time in response to changes in the operator's physical condition and work situation.

[0132] When selecting control parameters for the control device 50, the operator may select specific data to be applied according to their preference. In this case, for example, the presentation unit 150 may present the operator with the characteristics (e.g., responsiveness and damping characteristics) of a plurality of normative data extracted by the characteristic setting unit 130, and the reception unit 16 may accept the operator's operation to select one specific data. When an operator is aware of their own operational characteristics, specifying specific data that matches those characteristics can provide more effective operational support. Furthermore, it becomes possible to provide support that is flexibly adapted to the operator's physical condition and work situation.

[0133] Figure 16 illustrates a user interface screen presented by the presentation unit 150 when specific data can be selected according to the operator's wishes. The presentation unit 150 is composed of, for example, a display device installed in the driver's cab of a work machine in which the operator is seated. The display device is equipped with buttons and a touch panel for operating the display device, and the reception unit 16 is composed of, for example, these buttons and the touch panel.

[0134] In the example shown in Figure 16, on screen 71, the operator can select either "Automatic Setting" or "Manual Setting" via the reception unit 16. If "Automatic Setting" is selected, screen 72 is displayed, and the process of automatically setting the control parameters of the control device 50 (corresponding to step S114) begins. If "Cancel" is selected on screen 72, the process of automatically setting the control parameters is canceled, and the system returns to screen 71.

[0135] When "Manual Setting" is selected on screen 71, screen 73 is displayed, allowing the operator to select a control characteristic via the reception unit 16. On screen 73, one of three control characteristics can be selected. Once a control characteristic is selected on screen 73, the control parameters of the control device 50 are updated according to the operator's specifications. Then, screen 75 is displayed via screen 74, which indicates that a control characteristic has been selected and updated.

[0136] If the operator selects "Reselect" via the reception unit 16 on screen 75, the update of the control parameters is canceled, and the screen returns to screen 73. If "Cancel" is selected on screen 75, the update of the control parameters is canceled, and the screen returns to screen 71.

[0137] In this embodiment, the skill evaluation unit 17 evaluates the operator's skill based on the operator's operational data acquired by the operational data acquisition unit 11 and specific data extracted by the characteristic setting unit 130, and can add the evaluation result to the operational data. For example, by assigning a skill score indicating skill to the standard data in advance, it becomes possible to preferentially refer to standard data that is close to the working conditions and the operator's characteristics. Therefore, a highly accurate skill evaluation can be performed. As a skill score for the standard data, for example, 100 points could be set for standard data from skilled operators such as professional operators, and 60 points for the operator's own operational data (work data).

[0138] Figure 17 is a flowchart showing the process related to the evaluation of operator skills.

[0139] In step S302 of Figure 17, the skill evaluation unit 17 acquires operational data related to a specific task from the storage unit 120 via the operational data acquisition unit 11.

[0140] In step S304, the skill evaluation unit 17 determines the type of work corresponding to the operational data acquired in step S302. The type of work can be determined by, for example, an operation input pattern such as lever operation, or by a work discriminator constructed using machine learning.

[0141] In step S306, the skill evaluation unit 17 determines whether the identified work type is one that can be evaluated for skill. If the determination is affirmative, the process proceeds to step S308; otherwise, the process ends.

[0142] In step S308, the skill evaluation unit 17 extracts only the normative data that is of the same work type as the operational data acquired in step S302 and that closely matches the work conditions and operator characteristics. The skill evaluation unit 17 also stores the number of normative data that have been extracted.

[0143] Here, the skill evaluation unit 17 extracts normative data from the storage unit 120 that is similar to the operation data acquired in step S302, based on the work type, operation data, and operator characteristics (responsiveness, damping characteristics) indicated by the normative data determined in step S304. Furthermore, the skill evaluation unit 17 may narrow down the normative data by considering the posture and operating speed of the work machine, based on the initial posture of the work machine at the start of work and the time required for one operation. In this case, for example, the similarity of data conditions may be judged by a rule such as considering the initial posture of the work machine as the same condition for a certain width (for example, a width of 100 mm or less).

[0144] In step S310, the skill evaluation unit 17 searches for the normative data (work data) extracted in step S308 and extracts normative data that approximates the acquired operational data (work data). Here, for example, the skill evaluation unit 17 may sort the normative data extracted in step S308 in descending order of similarity and narrow down the data with the highest degree of similarity to be used as normative data for the table.

[0145] In step S312, the skill evaluation unit 17 performs a skill evaluation on the operational data acquired in step S302 based on the normative data extracted in step S308.

[0146] In step S314, the skill evaluation unit 17 determines whether the number of normative data stored in step S308 is less than a certain number. If the determination is affirmative, the process proceeds to step S318; otherwise, the process proceeds to step S316.

[0147] Specifically, step S314 determines whether the number of normative data that satisfy the same similar conditions is below a certain number. If the number of data is large, for example, the number of normative data for each score interval divided into 5-point increments (0-5, 5-10, ..., 90-95, 95-100) may be counted, and the oldest data may be deleted from the interval with the highest count (frequency). Alternatively, the oldest data may be deleted from the entire set of normative data. Alternatively, data that has been referenced as normative data for evaluation infrequently may be deleted preferentially.

[0148] In step S316, the skill evaluation unit 17 deletes some of the normative data extracted in step S308 from the storage unit 120. Here, for example, older normative data is deleted from the storage unit 120 so that the number of normative data that satisfy the same conditions becomes a certain number (step S314).

[0149] In step S318, the skill evaluation unit 17 adds the results of the skill evaluation (step S312) to the operational data acquired in step S302, stores it in the storage unit 120 as new standard data, and terminates the process. By adding the results of the skill evaluation to the operational data, information indicating the characteristics of the standard data can be automatically provided. This makes it possible to make the standard data applicable to a wide range of tasks and conditions.

[0150] Figure 18 shows an example of managing the number of normative data points to be limited to three for each similar condition (identical condition).

[0151] In the example shown in Figure 18, if three data points (Data B, Data C, and Data P) that satisfy the similar condition indicated as "Condition 2" are already stored in the storage unit 120, then if the newly acquired operational data in step S302 also satisfies "Condition 2," there will be a total of four operational data points that satisfy "Condition 2." Therefore, in such a case, for example, in step S316, the oldest of the three standard data points already stored in the storage unit 120 is deleted. If the newly acquired operational data in step S302 satisfies "Condition 1," but the total number of operational data points that satisfy "Condition 1," including this data, is three or less, then no standard data is deleted. Through the above processing, the number of standard data points to which skill evaluations have been assigned can be appropriately managed. This management prevents the number of standard data points corresponding to the same condition from becoming excessive, thus allowing for the coverage of standard data corresponding to a wide range of conditions while keeping the total number of standard data points under control.

[0152] As described above, in the control system 10 of the work machine of this embodiment, the operation data (reference data) includes control parameters (Figure 11B). However, if it is desired to generate reference data based on manual operation without going through the control system 10 of the work machine, the reference data does not have control parameters attached to it in the first place. Therefore, in such cases, it is necessary to add the control parameters to the reference data retroactively.

[0153] Figures 19 and 20 illustrate methods for assigning control parameters.

[0154] In the example shown in Figure 19, the controller output uc calculated based on the control deviation e can be approximated by the manual operation input u. Therefore, min(j) can be solved using optimization calculations on the evaluation function j = Σ(u - uc), and the control parameters (e.g., control gain) can be calculated and set as reference data.

[0155] As shown in Figure 20, system C can be identified from the input / output data (u, y) of the operation that serves as the reference data. The controller's gains and other parameters can then be adjusted so that the output y' of system C (output of the approximate model of the hydraulic excavator) and the reference output y (output of the actual hydraulic excavator) are approximate, and these adjustments can be set as control parameters (e.g., control gains).

[0156] As described above, in the above embodiment, the characteristics of the control device 50 are set based on reference data whose characteristics approximate those of the operational data, so that characteristics of the control device 50 that are adapted to the characteristics of the operator can be obtained. For this reason, work efficiency can be effectively improved.

[0157] Although the embodiments of this disclosure have been described in detail above, the invention is not limited to any particular embodiment, and various modifications and changes are possible within the scope of the claims. Furthermore, it is possible to combine all or more of the components of the multiple embodiments described above. In particular, the first and second embodiments can complement each other.

[0158] The skill evaluation system relating to the first aspect of this disclosure comprises: an operation data acquisition unit that acquires operation data of a work machine; a storage unit that stores reference data for comparison with the operation data; and an evaluation unit that evaluates the skill of the operator of the work machine, wherein the evaluation unit evaluates the skill based on the difference between the operation data acquired by the operation data acquisition unit and the reference data stored in the storage unit.

[0159] The skill evaluation system relating to the second aspect of this disclosure further comprises a presentation unit that presents information based on the skill evaluated by the evaluation unit in the first aspect, wherein the operation data and the normative data are time-series data, the evaluation unit evaluates the time-series skill of the operator of the work machine in a series of operations consisting of a plurality of time-series work units based on the difference between the operation data and the normative data, the presentation unit presents the information corresponding to the skill at the time when the evaluation result of the skill by the evaluation unit was the worst, upon completion of the series of operations, and the information presented by the presentation unit includes a message corresponding to a specific work unit among the plurality of work units to which the time belongs.

[0160] In the third aspect of the present disclosure, the skill evaluation system, in the first or second aspect, presents, as important information, the information relating to the work unit corresponding to the message that was presented relatively frequently among the multiple work units when the series of operations is performed multiple times.

[0161] In the skill evaluation system relating to the fourth aspect of this disclosure, if the evaluation result of the skill for the specific work unit by the evaluation unit is improved in the first to third aspects, the presentation unit presents a message corresponding to the improvement.

[0162] The skill evaluation system relating to the fifth aspect of this disclosure, in the first to fourth aspects, the message presented by the presentation unit when the series of operations is completed includes information indicating the chronological position of the specific operation unit in the plurality of operation units.

[0163] The skill evaluation system relating to the sixth aspect of this disclosure, in the first to fifth aspects, the message presented by the presentation unit when the series of tasks is completed includes advice to the operator to reduce the difference in the specific task unit.

[0164] In the seventh aspect of the present disclosure, the skill evaluation system, in the first to sixth aspects, provides information regarding the specific work unit corresponding to the message that was presented relatively infrequently when the series of operations was performed multiple times.

[0165] In the eighth aspect of the skill evaluation system of this disclosure, in the first to seventh aspects, the presentation unit presents a message generated based on the difference.

[0166] A control system for a work machine according to the ninth aspect of the present disclosure comprises a skill evaluation system for the first to eighth aspects and a setting unit for setting control parameters for controlling the work machine, wherein the storage unit stores a plurality of normative data, and the setting unit extracts specific data from the plurality of normative data stored in the storage unit that are normative data having characteristics similar to the operation data acquired by the operation data acquisition unit, and sets the control parameters based on the extracted specific data.

[0167] A control system for a work machine relating to the tenth aspect of this disclosure, wherein, in the ninth aspect, the normative data includes time-series work data that can be represented by representative characteristics including responsiveness, decay, or productivity.

[0168] A control system for a work machine according to the eleventh aspect of the present disclosure further comprises a productivity evaluation unit that evaluates productivity based on the operation data acquired by the operation data acquisition unit in the ninth or tenth aspect, wherein the operation data acquisition unit reacquires the operation data of the work machine whose control parameters have been set in advance based on specific data, the productivity evaluation unit evaluates the productivity based on the operation data reacquired by the operation data acquisition unit, and the setting unit updates the control parameters by extracting new specific data according to the evaluation result by the productivity evaluation unit.

[0169] A control system for a work machine relating to the twelfth aspect of the present disclosure further comprises, in the ninth to eleventh aspects, a presentation unit that presents the characteristics of a plurality of normative data extracted by the setting unit, and a reception unit that receives an operation to select one of the plurality of specific data whose characteristics have been presented by the presentation unit, wherein the setting unit sets the control parameters based on the specific data selected via the reception unit.

[0170] In the control system for a work machine relating to the thirteenth aspect of this disclosure, in the ninth to twelfth aspects, the storage unit stores the operation data acquired by the operation data acquisition unit as the reference data.

Claims

1. A skill evaluation system comprising: an operation data acquisition unit for acquiring operation data of a work machine; a storage unit for storing reference data for comparison with the operation data; and an evaluation unit for evaluating the skill of the operator of the work machine, wherein the evaluation unit evaluates the skill based on the difference between the operation data acquired by the operation data acquisition unit and the reference data stored in the storage unit.

2. A skill evaluation system according to claim 1, further comprising a presentation unit that presents information based on the skill evaluated by the evaluation unit, wherein the operation data and the normative data are time-series data, the evaluation unit evaluates the time-series skill of the operator of the work machine in a series of operations composed of a plurality of time-series work units based on the difference between the operation data and the normative data, the presentation unit presents the information corresponding to the skill at the time when the evaluation result of the skill by the evaluation unit was the worst, upon completion of the series of operations, and the information presented by the presentation unit includes a message corresponding to a specific work unit among the plurality of work units to which the time belongs.

3. A skill evaluation system according to claim 2, wherein the presentation unit presents, as important information, the information relating to the work unit corresponding to the message that was presented at a relatively high frequency among the plurality of work units when the series of work is performed multiple times.

4. A skill evaluation system according to claim 2, wherein when the evaluation result of the skill performed by the evaluation unit for the specific work unit is improved, the presentation unit presents a message corresponding to the improvement.

5. A skill evaluation system according to claim 2, wherein the message presented by the presentation unit when the series of tasks is completed includes information indicating the chronological position of the specific task unit in the plurality of task units.

6. A skill evaluation system according to claim 2, wherein the message presented by the presentation unit when the series of tasks is completed includes advice to the operator for reducing the difference in the specific task unit.

7. A skill evaluation system according to claim 2, wherein the presentation unit presents information relating to the specific work unit corresponding to the message presented at a relatively low frequency when the series of operations is performed multiple times.

8. A skill evaluation system according to claim 2, wherein the presentation unit presents a message generated based on the difference.

9. A control system for a work machine comprising: a skill evaluation system according to claim 1; and a setting unit for setting control parameters for controlling the work machine, wherein the storage unit stores a plurality of normative data; and the setting unit extracts specific data from the plurality of normative data stored in the storage unit, which is normative data having characteristics similar to the operation data acquired by the operation data acquisition unit, and sets the control parameters based on the extracted specific data.

10. The control system for a work machine according to claim 9, wherein the normative data includes time-series work data that can be represented by representative characteristics including responsiveness, decay, or productivity.

11. A control system for a work machine according to claim 9, further comprising a productivity evaluation unit that evaluates productivity based on the operational data acquired by the operational data acquisition unit, wherein the operational data acquisition unit reacquires the operational data of the work machine whose control parameters have been set in advance based on specific data, the productivity evaluation unit evaluates the productivity based on the operational data reacquired by the operational data acquisition unit, and the setting unit updates the control parameters by extracting new specific data according to the evaluation result of the productivity evaluation unit.

12. A control system for a work machine according to claim 9, further comprising: a presentation unit that presents the characteristics of a plurality of normative data extracted by the setting unit; and a reception unit that receives an operation to select one of the plurality of specific data whose characteristics have been presented by the presentation unit, wherein the setting unit sets the control parameters based on the specific data selected via the reception unit.

13. The control system for a work machine according to claim 9, wherein the storage unit stores the operation data acquired by the operation data acquisition unit as the reference data.

Images