Work difficulty evaluation system and work planning system

Abstract

To provide a system for appropriately evaluating the difficulty of a task. [Solution] The task difficulty evaluation system of the present invention is a system that evaluates the difficulty of a task to be evaluated when a work machine (100) that outputs operations according to input performs the task, and is equipped with an operation information acquisition unit (11) that acquires data related to the output when performing the task, a nonlinearity estimation unit (12) that includes prediction data means that predicts and accumulates the input of the work machine (100) when performing the task to be evaluated based on the data related to the output, and evaluates the degree of nonlinearity of the accumulated prediction data, and a task difficulty estimation unit (13) that evaluates the difficulty of the task to be evaluated based on the degree of nonlinearity.

Description

[Technical Field]

[0001] The present invention relates to a technique for evaluating the difficulty of a task. [Background technology]

[0002] At a work site (hereinafter simply referred to as a work site), operators are required to operate work machines and efficiently complete various tasks. For example, at a construction site using construction machinery such as hydraulic excavators, it is desirable to complete various work tasks (types) such as excavation and loading, leveling and shaping the ground, slope leveling and shaping, building demolition, traveling on uneven ground, and traveling on muddy ground in a short amount of time to achieve the desired condition. Therefore, if the difficulty of the tasks can be evaluated, the site manager can appropriately allocate work machines and workers to the site. A known example of work management that sets the difficulty of the tasks themselves is described in, for example, paragraph 0109 of Japanese Patent No. 7159797 (Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7159797 Paragraph 0109 Summary of the Invention [Problem to be solved by the invention]

[0004] In the past, it was thought that the difficulty of a task was primarily determined by external information such as on-site drawings, on-site photographs, task content (type), task time, etc., and work plans were drawn up based on external information. However, when the work was actually carried out on-site, the work did not proceed according to the original work plan, and delays often occurred, requiring changes to the work plan.

[0005] This is because the difficulty of a task is determined by various factors other than the task's external information. For example, the required operation of the task varies depending on the size and performance of the work machine (the machine's mechanical and body characteristics). For example, even when excavating a site with the same volume and shape, the task (or the operation of the work machine assigned to that task) can be easy or difficult depending on the hydraulic excavator's size, power, bucket shape, speed of operation, operating range, and various other factors. (Although it may be thought that greater size and power generally equates to easier operation, this is not always the case.) Furthermore, in actual work sites, even when excavating a site with the same shape and volume, the difficulty of the task is likely to change depending on internal factors that can only be learned once the task begins, such as the soil quality of the ground to be excavated, and slope erosion or softening of the ground due to rainwater inflow. Furthermore, the difficulty of the task is likely to change depending on the work environment, which is neither external nor internal, such as weather conditions. For this reason, in the past, when setting the difficulty level of a task, it was sometimes the case that the difficulty level was not set appropriately or fairly.

[0006] In particular, given the recent demand for labor-saving and unmanned operations, remote operation of work machines due to the development of communication networks, and automatic, autonomous, and self-driving work machines, even if the work machines on hand are assigned to each task based solely on the external difficulty of the task, work efficiency will not improve if the mechanical characteristics or machine characteristics (which can also be considered ease of use) of the work machines are poor.

[0007] In view of the above-mentioned circumstances, the present invention aims to provide a system that can appropriately evaluate and predict the difficulty of a task, taking into consideration not only external information about the task but also operability (task condition 1) resulting from the work machine, vehicle characteristics, and mechanical characteristics of the work attachments of each individual work machine, as well as internal information about the task (task condition 2) and the work environment (task condition 3). [Means for solving the problem]

[0008] For this purpose, the difficulty assessment system according to the present invention is a system that assesses the difficulty of a task to be assessed when a work machine that outputs an operation in accordance with an input performs the task, and is equipped with an operational data acquisition means that acquires data related to the output when the task is performed, a prediction data means that predicts and accumulates the input when the task to be assessed is performed based on the data related to the output, a nonlinearity assessment means that assesses the degree of nonlinearity of the accumulated prediction data, and a difficulty assessment means that assesses the difficulty of the task to be assessed based on the degree of nonlinearity.

[0009] According to the present invention, regardless of the work content (type), the mechanical characteristics (machine characteristics) of the work machine performing the work to be evaluated are taken into consideration, making it possible to perform a more appropriate and fair difficulty evaluation than in the past. The work to be evaluated may be a work in progress that is currently being performed, or a work that will be performed in the future but has not yet been performed. The work to be evaluated may also be a type of work that has never been performed before, or a type of work of the same type as a work that has been completed in the past and for which a track record has been accumulated, or a type of work similar to a type of work that has been completed in the past.

[0010] The operation data acquisition means may acquire data relating to the output of an in-progress task being evaluated, or may acquire data relating to the output of a task evaluated in the past. In one aspect of the present invention, the difficulty assessment means predicts the difficulty of the task to be evaluated based on the degree of similarity between the working conditions of one or more tasks previously evaluated by the nonlinearity assessment means and the working conditions of the task to be evaluated by the nonlinearity assessment means. According to this aspect, the difficulty assessment means assesses the degree of nonlinearity taking into account various working conditions, such as the machine characteristics (class, working mode, etc.) and machine characteristics (type of work attachment) of the work machine related to the previously evaluated task, the work content (type, working time, etc.), the soil quality of the ground, and the working environment (whether or not there is precipitation), thereby enabling an accurate assessment of the difficulty of the task. Therefore, even if the work site changes or the specifications of the work machine change, making operability (usability) more difficult, the difficulty of the task to be performed or the working area to be performed can be predicted if the working conditions are similar to those previously evaluated.

[0011] In a preferred aspect of the present invention, the device further comprises a means for evaluating the operating skill of a work machine operator based on data relating to the output accumulated in the operation data acquisition means and the degree of nonlinearity. According to this aspect, for example, if good machine behavior can be achieved even under highly nonlinear working conditions, i.e., conditions where the work is highly difficult, this indicates that the operator is able to perform operations in response to this nonlinearity, and it is possible to paradoxically evaluate that the operator has high skill without having to use normative data or understand the details of the work.

[0012] The work planning system of the present invention includes the above-described difficulty assessment system and a work planning unit that stores work plans, and the work planning unit changes the work plan in accordance with the predicted difficulty level. According to this aspect, the work plan is changed based on the predicted task difficulty, and the work area or operator is changed, thereby improving work efficiency and safety. [Effects of the Invention]

[0013] As described above, the present invention makes it possible to fairly evaluate the difficulty of a task regardless of external information about the task, such as the task content (type). This improves work efficiency and safety, and contributes to shortening work time. It also contributes to improving work efficiency in automated or autonomous driving that does not depend on the driver's operating skill. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a block diagram showing work (including the operation of a work machine) as an object to be controlled according to an embodiment of the present invention. [Figure 2A] 10A and 10B are graphs illustrating the variance value σ2 of the predicted input value and an example of a nonlinearity evaluation value. [Figure 2B] 10 is a graph showing a time chart of the machine output from the hydraulic excavator G and the predicted input value to the hydraulic excavator G. FIG. [Figure 3] FIG. 10 is a graph illustrating an example of the relationship between a nonlinearity evaluation value and task difficulty. [Figure 4] FIG. 10 is a graph illustrating an example of the relationship between a nonlinearity evaluation value and task difficulty. [Figure 5] FIG. 10 is a graph illustrating an example of a graded evaluation of task difficulty. [Figure 6] 1 is an overall diagram showing the configuration of a difficulty level assessment system and an operation planning system according to an embodiment of the present invention. [Figure 7] 10 is a table showing various work conditions related to past work stored in a server. [Figure 8] FIG. 1 is a schematic diagram showing a specific example of work (excavation work). [Figure 9] FIG. 10 is a schematic diagram illustrating operation assistance based on difficulty level evaluation. [Figure 10] 10A and 10B are schematic diagrams illustrating changes in a work plan and changes in work difficulty. [Figure 11] FIG. 10 is a block diagram showing a work as a control target according to another embodiment of the present invention. [Figure 12] FIG. 10 is an overall view showing the configuration of a difficulty level assessment system and an operation planning system according to another embodiment of the present invention. [Figure 13] FIG. 2 is a plan view showing a procedure for calculating a traveling position P(t) and a target traveling position P hat (t). [Figure 14] 10 is a scatter plot showing the absolute value of the difference in traveling position, the difficulty level d(P), and the deviation tolerance value. [Figure 15] FIG. 10 is a plan view showing how a difficult-to-travel work area (area A) is set based on the threshold value of difficulty d(P). [Figure 16] FIG. 2 is a plan view illustrating an example of an area A and a travel route of a work machine. [Figure 17] FIG. 10 is a plan view showing the issuance of a warning to the operator of the work machine depending on whether the position and course of the work machine are inside or outside area A. DETAILED DESCRIPTION OF THE INVENTION

[0015] An embodiment of the present invention will be described in detail below with reference to the drawings. As a specific example of an evaluation target task for which the difficulty assessment system of this embodiment will evaluate the difficulty, an excavation task performed by a work machine 100 as shown in Figure 8 (solid line, fine dashed line, coarse dashed line; the same applies to Figures 6, 9, and 12 described below) will be described.

[0016] The work machine 100 is, for example, a hydraulic excavator, and includes a lower traveling body 101, an upper rotating body 102 which is a rotating body mounted on the lower traveling body 101 so as to be rotatable around a rotation axis X, a boom 104, an arm 105, and an excavation bucket 106 which constitute a work device provided on the upper rotating body 102, a boom cylinder 107, an arm cylinder 108, and a bucket cylinder 109 which serve as hydraulic actuators for operating these components, a rotation motor (not shown) which is a hydraulic motor that drives the upper rotating body 102 around the rotation axis X relative to the lower traveling body 101, a cabin 111 provided on the upper rotating body 102, and an on-board controller (not shown) which controls each of the cylinders 107 to 109 based on the operation of an operator in the cabin 111.

[0017] The boom 104 has a base end connected to the upper rotating body 102 so as to be rotatable in the hoisting direction, and a tip end on the opposite side. The arm 105 has a base end connected to the tip end of the boom 104 so as to be rotatable, and a tip end on the opposite side, and can swing in a push / pull direction, moving toward or away from the upper rotating body 102.

[0018] The bucket 106 is rotatably attached to the tip of the arm 105 and can excavate and remove topsoil. The boom cylinder 107 is interposed between the boom 104 and the upper rotating body 102 so that it extends when hydraulic oil is supplied from a hydraulic circuit (not shown) or contracts when hydraulic oil is discharged to the hydraulic circuit, and raises and lowers the boom 104 as it extends and contracts.

[0019] Similarly, an arm cylinder 108 is interposed between the boom 104 and the arm 105 so as to swing the arm 105 in a pushing and pulling direction by its extension and retraction, and a bucket cylinder 109 is interposed between the arm 105 and the bucket 106 so as to rotate the bucket 106 by its extension and retraction. The arm 105, the bucket 106, the arm cylinder 108, and the bucket cylinder 109 constitute a work attachment 112 for excavation work.

[0020] An IMU (Inertial Measurement Unit) sensor that detects the attitude angle of the attached members relative to the horizontal plane is attached to each of the boom 104, arm 105, and bucket 106. Alternatively, a potentiometer that detects the relative angle between the two rotatably connected members is attached to the rotating part that connects the boom 104 or bucket 106 to each other.

[0021] In the case of an IMU sensor, an on-board controller (not shown) that controls each of the cylinders 107-109 calculates the stroke amount of the cylinders 107-109 (the stroke amount of the piston inside the cylinder; the same applies below) from information about the work attachment 112, such as the rotation axis position of each member of the work attachment 112 and the dimensions of each member, the connection relationship of each member of the work attachment 112, the attachment position dimensions of the cylinders 107-109 attached to each member of the work attachment 112, the installation position of the IMU sensor relative to each member of the work attachment 112, and attitude angle information of each member of the work attachment 112 output from the IMU sensor.

[0022] In the case of a potentiometer, the on-board controller (not shown) of the work machine 100 calculates the stroke amount of each of the cylinders 107 to 109 from information about the work attachment 112, such as the rotation axis positions of each member of the work attachment 112 and the dimensions of each member, the connection relationships of each member of the work attachment 112, the mounting position dimensions of each of the cylinders 107 to 109 attached to each member of the work attachment 112, and the relative angle information of each member of the work attachment 112 output from the potentiometer.

[0023] An operator in a cabin 111 manually and skillfully operates a plurality of controls such as control levers to operate the boom 104, arm 105, and bucket 106 (so-called manual operation), and excavates the topsoil on the ground surface with the bucket 106.

[0024] In this excavation work, the difficulty level assessment system of this embodiment uses the speed data of the work attachment 112 TIFF2026042465000002.tif5169 is acquired. t is time or duration. For the speed data of the work attachment 112, the stroke speed of the arm cylinder 108, which contributes greatly to excavation, the combined center of gravity speed of the work attachment 112, and the stroke speeds of the other cylinders 107, 109 are used as representative values. FIG. 1 is a block diagram showing work according to one embodiment of the present invention as an object to be controlled. In the block diagram, the work machine 100 or work attachment 112 performing excavation work is represented as a hydraulic excavator G in FIG. 1. The output yo of the hydraulic excavator G is, for example, the arm cylinder speed yo of the arm cylinder 108 (FIG. 8). Furthermore, the block diagram of FIG. 1 includes a reference model Gm to which a target value r of the speed of the arm cylinder 108 during excavation is input, an FIR type controller Cr* to which an output ym of the reference model Gm is input, and a PID type controller Ce to which a positive-negative difference between the output ym and the output yo is input, and a predicted input value obtained by adding the output of the FIR type controller Cr* and the output of the PID type controller Ce is calculated. TIFF2026042465000003.tif6170 is input into hydraulic excavator G.

[0025] The difficulty assessment system of this embodiment monitors the output yo at different times t via the on-board controller of the work machine 100, acquires multiple outputs yo, and calculates speed data TIFF2026042465000004.tif7170 is obtained. Using the data from TIFF2026042465000005.tif6169, a feedforward controller was designed to achieve the desired output characteristics. At the same time, a response prediction method called ERIT (Estimated Response Iterative Tuning), which is a method for predicting the input generated by the designed controller, was used to predict the predicted input value input to the hydraulic excavator G as shown on the left in Figure 2. TIFF2026042465000006.tif5169 (vertical axis on the left of Figure 2A). The horizontal axis on the left of Figure 2A is time t, which represents the time from the start of the task to the present or the end of the task.

[0026] Based on this, multiple predicted input values ​​at multiple times t from the start of work to the present are calculated. Each predicted input value for TIFF2026042465000007.tif5168 Variance value σ of TIFF2026042465000008.tif5169 2 Calculate.

[0027] Furthermore, the linear equation on the right side of Figure 2A is referenced to determine the nonlinearity evaluation value (the evaluation value of the degree of nonlinearity; the larger this value, the stronger the degree of nonlinearity, and conversely, the smaller this value, the weaker the degree of nonlinearity). The nonlinearity evaluation value is calculated by the variance value σ 2 Alternatively, the variance value σ 2 σ may be set to change exponentially with respect to σ. 2 If is small, the nonlinearity evaluation value will be small. Conversely, the variance value σ 2 In the proposed method using the ERIT method, the variance value σ 2 is an example of an index for evaluating linear strength (nonlinearity), and the variance value σ 2 The advantage of this method is that it allows us to evaluate the "degree of nonlinearity" (small variance = strong linearity, i.e., weak nonlinearity; large variance = weak linearity, i.e., strong nonlinearity).

[0028] The nonlinearity evaluation value is the predicted input value This indicates the degree of nonlinearity of TIFF2026042465000009.tif5168, and can be expressed as a percentage. Anything other than linear is nonlinear, regardless of the degree. Predicted input value If there is no variation in TIFF2026042465000010.tif5168, it is perfectly linear (nonlinearity evaluation value = 0). If the variation in TIFF2026042465000011.tif5168 is extremely large and there is no linearity, the nonlinearity evaluation value is 100.

[0029] The above-mentioned speed data TIFF2026042465000012.tif6169 shows an example of operational data for a work machine such as a hydraulic excavator. TIFF2026042465000013.tif6169 is the predicted value of the operation input required for hydraulic excavator G to achieve the desired output. TIFF2026042465000014.tif6168 indirectly represents the mechanical characteristics or body characteristics of the work machine. As shown in Figure 1, it is a predicted value calculated from a block diagram including the reference model Gm, FIR controller Cr*, and PID controller Ce, and does not need to be an actual value. Predicted input value TIFF2026042465000015.tif5168 is determined in correspondence with two values: the actual output yo from the hydraulic excavator G (working machine) and the reference output or desired output ym. The reference output or desired output is a value that is uniquely determined by the designer, and is a predicted input value calculated depending on the conditions set, such as the mechanical characteristics (machine characteristics) of the hydraulic excavator G at the time of evaluation, as well as low-response output and high-response output. TIFF2026042465000016.tif5168 can take the form of a clean first-order lag waveform, or it can become oscillatory, as shown in the graph in Figure 2B. The machine output, which is the vertical axis in the upper graph of Figure 2B, is the output from the hydraulic excavator G (work machine).

[0030] Next, the difficulty level evaluation system of this embodiment determines whether the task difficulty level d is high if the nonlinearity evaluation value x is high. The relationship is defined so that the relationship is linear ( TIFF2026042465000017.tif4170) and exponential functions ( TIFF2026042465000018.tif5169), or may be expressed by an approximate formula (solid line) based on the nonlinearity evaluation value of past tasks and actual data on task difficulty, which are displayed as a point cloud in Figure 4.

[0031] The task difficulty level does not have to be continuous as shown in FIGS. 3 and 4, but may be expressed in stages such as low, medium, and high as shown in FIG.

[0032] In Figure 5, the nonlinearity evaluation value gradually increases from the sixth trial to the ninth trial. In this case, the ninth trial with a nonlinearity evaluation value of 90% and the tenth trial with a nonlinearity evaluation value of 90% are judged to be high in work difficulty. This may be due to a deterioration in on-site work conditions, such as changes in the excavated soil or worsening weather conditions.

[0033] When work is performed at a new site, the work conditions such as the machine specifications of the work machine to be used for the work (for example, the machine class displayed in machine weight [tons] (Fig. 7 described below), machine output [kW], or length of the work attachment (standard, long)), work content (type), and ground conditions (sand, soil, or rock, influence of rain, rivers, groundwater, etc.) are either automatically read from the work plan or entered by the worker or manager, and a calculation method for the work difficulty (Figs. 3 to 5, etc.) is determined based on the similarity with the work conditions of work for which difficulty has been evaluated in the past, and the work difficulty of the work at the new site is predicted.

[0034] Here, past evaluations are not limited to work performed by a specific work machine, but can also be work results from multiple work machines or different work sites, and the work difficulty and its work conditions can be accumulated and saved through communication between an information aggregation server or the like and each work machine, and can be used for future predictions. As will be described in more detail below, machine specifications for multiple work machines, difficulty evaluation results and work conditions at work sites other than the site of the work being evaluated, and difficulty evaluation results and work conditions for past work being evaluated are shown as condition B, condition C, etc. in Figure 6, and examples are shown in Figure 7. There is no particular limit to the number of past data information.

[0035] The information aggregation server (operation information storage unit 14 in FIG. 6, which will be described later) that is connected to the on-board controller of the work machine 100 via a network means stores, for example, (1) the machine class of the work machine, (2) the work mode (emphasis on workability, emphasis on fuel economy, etc.), (3) the work content (type), (4) ground characteristics (soil quality, presence or absence of precipitation), and the difficulty level information at that time (FIG. 7).

[0036] The evaluation of the above-mentioned similarity can be done based on the number of matching working conditions and evaluation points, and the working conditions can be, for example, as the mechanical and machine characteristics of the work machine itself, (Working Condition 1) (1) class is 13 ton class, (2) working mode is fuel efficiency emphasis, as the external information of the work, (Basic Working Condition) (3) work content (type) is excavation work, as the internal information of the work, (Working Condition 2) (4)-1 soil type is sandy, and as the working environment, (Working Condition 3) (4)-2 precipitation presence / absence is no precipitation. Based on the similarity f (condition match 1, mismatch 0) with the past results stored on the server (Figure 7), No. 3 is selected, which has the highest evaluation value J of similarity f, and the difficulty under these conditions is used as the predicted difficulty. TIFF2026042465000019.tif6169If there are multiple task conditions with the same number of matches, the average difficulty value of those conditions can be used as the predicted difficulty value.

[0037] In addition, instead of simply evaluating similarity based on the number of matches, a weight is set for each item of work conditions to calculate an evaluation value J as shown in the formula below, and the difficulty of the condition for which the evaluation value J is the largest is set as the predicted difficulty. Alternatively, multiple tasks can be selected in descending order of the evaluation value J, and the difficulty of the selected task conditions can be used as a factor to calculate the difficulty using the weighted linear average (LWA), in which task conditions with a higher evaluation value J contribute more. If a threshold is set for the task difficulty, some kind of confirmation or action can be taken if this threshold is exceeded.

[0038] Examples include (1) changes to the work plan or the operation of the work machine, (2) changes to the work machine, and (3) changes to the worker. (1) In the case of changes to the work plan or the operation of the work machine, the work operation is modified to a relatively light load so that the difficulty of the work (nonlinearity) is reduced. For example, in the case of excavation work, the depth of the excavation trajectory is changed according to the soil quality (in the case of automatic driving, the depth is set according to the class).

[0039] For example, as shown in Fig. 9, when the ground is clayey or rocky, the difficulty of the work (which can be rephrased as the difficulty of operating the work machine) can be reduced by instructing the operator on display device 31 (when the operator is actually on board and operating the work machine) to work shallower than in general soil. Display device 31 is an operation assistance control terminal installed in cabin 111 shown in Fig. 8.

[0040] For example, in situations where repeated excavation is required on hard ground, the predicted input value of the work machine is low due to the characteristics of the machine, such as low output. If TIFF2026042465000021.tif5168 changes rapidly (i.e., the nonlinear evaluation value is large) and the difficulty of the work is high, the work planning unit 21 prompts the work manager to change to a machine class that is expected to bring the difficulty of the work below the threshold, or rewrites the work plan. Figure 10 is a graph showing, with arrows, a comparison of the difficulty of the work before and after changing the machine class from (13t) to (20t). According to the work planning system of this embodiment, (2) taking the measure of changing the work machine, the difficulty of the work decreases, contributing to improved work efficiency and reduced work time.

[0041] Figure 6 is a system configuration diagram of a work planning system including the difficulty assessment system of this embodiment. In Figure 6, the difficulty assessment system of this embodiment comprises an operation information acquisition unit 11, a nonlinearity estimation unit 12, a task difficulty estimation unit 13, and an operation information storage unit 14, and assesses the difficulty of a task to be assessed that involves condition A as a task condition. The difficulty assessment system or work planning system may be installed in the on-board controller of the work machine 100, or may be installed outside the work machine 100, for example, in a site office or a management company building, or may be installed in a cloud on a communications network.

[0042] The nonlinearity estimation unit 12 estimates the predicted input value TIFF2026042465000022.tif6169 was calculated, and the variance value σ was calculated as shown in Figure 2. 2 and calculates a non-linearity evaluation value. The task difficulty level estimation unit 13 evaluates the difficulty level as shown in Fig. 3, 4 or 5 described above. The operation information storage unit 14 stores information including task difficulty levels, such as the working conditions of tasks for which the difficulty level has been evaluated in the past, for one work machine or other work machines, as described above.

[0043] The work planning system of this embodiment includes the components of the difficulty assessment system described above, a work planning unit 21, an operation information comparison calculation unit 22, and a work difficulty prediction unit 23. The work planning unit 21 stores information related to the above-mentioned work plans and rewrites (changes) the stored work plans. As described above, the operation information comparison calculation unit 22 compares the work content (data information) related to the current work to be evaluated that is to be undertaken with the work content (data information) related to work evaluated in the past. The work difficulty prediction unit 23 predicts the difficulty of the current work to be evaluated that is to be undertaken based on the similarity with the data information related to the past evaluations, as described above.

[0044] FIG. 11 is a block diagram showing a modified example of the controlled object shown in FIG. 1, and the basic configuration of the block diagram of the modified example is the same as that of FIG. 1 described above, so that a repetitive explanation will be avoided. In the block diagram shown in FIG. 11, the PID controller Ce in FIG. 1 represents the operator (person) of the work machine 100 (hydraulic excavator G) as one that can be expressed by a PID controller. The sum of the output of the operator (person) and the output of the FIR type controller Cr* is the predicted input value input to the hydraulic excavator G. TIFF2026042465000023.tif6169. The output of hydraulic excavator G is the actual machine behavior yo(t).

[0045] Figure 12 is an overall diagram showing the configuration of a modified example of the system shown in Figure 6. The basic configuration of the overall diagram of the modified example is the same as that of Figure 6 described above, so redundant explanation will be avoided. The difficulty assessment system shown in Figure 12 further includes a skill assessment unit 15 that assesses the operator's operational skill. Furthermore, the work planning unit 21 of the work planning system changes an already stored work plan based on the operator's operational skill assessed by the skill assessment unit 15. A specific example is (3) changing the operator.

[0046] For example, if the task is evaluated as being difficult because the nonlinearity evaluation value is high despite (1) changes to the work plan or the operation of the work machine and (2) changes to the work machine, or if it is clear that the operator's skills are inappropriate, it is desirable to assign an operator with appropriate operating skills to the operator of the work machine involved in the excavation work. This is because whether or not the excavation work can be completed depends on the operator's skill. Furthermore, the level of completion and work time (task) required for the excavation work also depend on the operator's operating skill.

[0047] Conversely, the same applies when the level of difficulty is low, and there is no need to assign an experienced operator to an excavation task with low difficulty.

[0048] Therefore, among the data stored in the server of the difficulty assessment system, we focus on data that shows excellent work results despite the work being considered to have a high level of work difficulty, i.e., a high nonlinearity assessment value. Here, excellent data is, for example, time series data such as the cylinder speed yo(t) during excavation and the arm top height during ground leveling associated with excavation. Generally, a pilot with high operational skills can be regarded as a nonlinear controller with high control performance that can adaptively respond to various situations and multiple work conditions. As shown in Figure 11, the machine behavior yo(t) is the desired behavior, but the predicted input value If the TIFF2026042465000024.tif5169 is oscillatory, the nonlinear evaluation value is high. Furthermore, if the nonlinear evaluation value is high, the operator (person) can be considered to be behaving as a nonlinear controller with high control performance. In other words, the operating skill of the operator who is the source of the data can be evaluated by paradoxically grasping the degree of nonlinearity (Figures 11 and 12). Therefore, by reassigning the task difficulty (nonlinearity evaluation value) of the task to be evaluated now to the operator who is the source of the data and has a higher difficulty (nonlinearity evaluation value) and excellent results, the task can be fully accomplished.

[0049] Next, as a second example, the difficulty level evaluation of the task of machine body traveling (operation of the work machine) will be described.

[0050] Construction machinery such as hydraulic excavators, bulldozers, wheel loaders, crawler cranes, and aerial work platforms, as well as other self-propelled non-construction machinery, may experience lift when the vehicle runs over uneven ground, or frictional force may be reduced due to mud, causing the tracks and wheels of the undercarriage 101 ( FIG. 8 ) of the work machine 100 to spin. If such spinning increases, precise driving of the hydraulic excavator becomes difficult. This makes it particularly difficult to control the work machine using a remote control system or an automated driving system. In situations where such spinning occurs frequently, the estimated input value (corresponding to the input command in control) fluctuates even with respect to the amount of movement when controlling multiple hydraulic excavators with the same output. In other words, the nonlinearity evaluation value described above increases. Therefore, as with the excavation operation described above, the difficulty of driving operation (hereinafter referred to as driving operation) can be expressed using the nonlinearity evaluation value.

[0051] In the second embodiment, each time a travel operation is performed, the values ​​input by the operator to the controls at that time, such as the lever operation amount or travel pilot pressure, and the values ​​output by the work machine (specifically, the amount of movement of the hydraulic excavator) are each acquired. A nonlinearity evaluation value at that point is calculated from this input / output relationship. Then, by linking this nonlinearity evaluation value to the position where the excavator was traveling at that time, it is possible to evaluate the skill of the travel operation in the same way as the excavation operation described above.

[0052] Furthermore, a work area where the nonlinearity evaluation value is large and where driving operation is estimated to be difficult can be defined as "area A," and area A can be stored in the work planning unit 21 shown in Figure 6 or otherwise developed on a system such as Figure 6, and by determining whether the work machine is within area A or is approaching area A, a driving route can be planned that avoids area A and presented to the operator of the work machine on a display device 31 (Figure 9) or the like. Alternatively, an automatic driving route that avoids area A can be set. Figure 16 is a plan view showing area A and how the work machine 100 plans a driving route that avoids area A.

[0053] When a work machine is about to enter area A, the difficulty assessment system of the second embodiment can prohibit driving operations toward area A, add driving assist control, or present a message urging the operator to pay attention to driving operations. In addition, while the work machine is traveling in area A, a warning can be presented to the operator that the work machine is currently traveling in an area with a high level of difficulty.

[0054] Next, an example of means for determining a difficult area related to driving operations will be described.

[0055] Figure 13 is a plan view showing the travel of a work machine. First, the actual travel history P(t) of a specific operator over the most recent period (several days to several months) is obtained from, for example, GNSS data. Then, after time Δt has passed, the difference between the actual travel history P(t+Δt) and the target travel path P^(t+Δt) is calculated and the absolute value is taken (see the vertical axis of Figure 14).

[0056] Next, the travel difficulty calculated based on the nonlinearity evaluation value calculated while traveling at a certain point P(t) is designated as d(P), and the above-mentioned absolute value of the travel position difference and the difficulty d(P) are plotted as a scatter plot as shown in FIG. 14. Here, the allowable limit (also referred to as the deviation tolerance, or simply the tolerance) for travel deviation is determined based on the width and danger level of the surrounding roads on which the work machine 100 (hydraulic excavator) to be acquired is travelling. For example, in a highly dangerous location such as a narrow mountain road with cliffs just outside the road, this tolerance will be small. On the other hand, in a situation where the vehicle can travel freely over a wide, flat area, this tolerance may be large. The smallest d(P) that may exceed this tolerance is set as the difficulty threshold for that operator.

[0057] Next, the value of driving difficulty d(P) is plotted at its coordinate P on the plan view of the work site shown in Figure 15. In Figure 15, points that exceed the threshold are shown with black circles, and points that do not are shown with white circles. At this time, since it is impossible to obtain data for all points on the two-dimensional plane shown in Figure 15, for points where there is no data, predicted values ​​from similar conditions are used, or an estimate is made by linear interpolation from data for points surrounding P. When making this estimate, the data used also references the driving history of other work machines, including the own machine, and other operators, as represented by the multiple arrows in Figure 15. Then, the range that exceeds the threshold is calculated, and this range is set as "Area A."

[0058] Since this area A is considered to be an area where driving operation is difficult, a route that avoids area A is recommended to the operator of the work machine 100 (whether inside the cabin 111 or remotely operated from outside the cabin 111) as a driving route, as shown by the solid line in Fig. 16. Furthermore, by setting a similar area A in automatic driving control (or even more advanced autonomous driving control), it is possible to determine that driving through area A is difficult for that automatic driving control and to use a route that avoids area A. Furthermore, it is also possible to determine whether the own machine is driving through or about to enter area A, as shown in Fig. 17, and to issue a warning to the operator (whether inside the cabin 111 or remotely operated from outside the cabin 111). For example, the following warnings are issued.

[0059] As shown in FIG. 17, when the aircraft is in area A, the pilot is notified by a warning message displayed on the display device and an alarm sound.

[0060] Furthermore, as shown in Figure 17, when the aircraft is outside area A but within a certain distance from area A, and area A is in the direction of travel, the pilot is notified of the issuance of a warning by displaying a warning message on the display device and sounding an alarm. The content of the warning notification (content of the message, frequency of the alarm sound, etc.) is different from the warning issued when the aircraft is within area A described above.

[0061] Although the embodiments of the present invention have been described above with reference to the drawings, the present invention is not limited to the illustrated embodiments. Various modifications and variations can be made to the illustrated embodiments within the same scope as the present invention or within an equivalent scope. For example, some components may be extracted from one embodiment described above, and other components may be extracted from another embodiment described above, and these extracted components may be combined. [Industrial Applicability]

[0062] The present invention can be advantageously used in mechanized construction and mechanized work using work machines, manual operation of work machines by operators, and automatic or autonomous operation that does not depend on the operating skill of the operator. [Explanation of symbols]

[0063] 11 Operation information acquisition unit (operation data acquisition means), 12 Nonlinearity estimation unit (prediction data means + nonlinearity evaluation means), 13 work difficulty estimation unit (difficulty evaluation means), 14 operation information storage unit, 15 Skill evaluation section (means for evaluating the pilot's operating skills), 21 work planning unit, 22 operation information comparison calculation unit, 23 Work difficulty prediction unit (difficulty evaluation means for predicting the difficulty of the work to be evaluated) 100 Work machine, 101 Undercarriage, 105 Arm, 108 Arm cylinder, 112 Work attachment.

Claims

1. A system for evaluating the difficulty of a task to be evaluated when a work machine that outputs an operation according to an input performs the task, comprising: an operation data acquisition means for acquiring data relating to the output when performing a task; a prediction data means for predicting and storing the input when the evaluation target work is performed based on the data related to the output; a nonlinearity evaluation means for evaluating the degree of nonlinearity of the accumulated prediction data; A task difficulty evaluation system comprising: a difficulty evaluation means for evaluating the difficulty of the task to be evaluated based on the degree of nonlinearity.

2. 2. The task difficulty evaluation system of claim 1, wherein the difficulty evaluation means predicts the difficulty of the task to be evaluated based on the similarity between the task conditions for one or more tasks previously evaluated by the nonlinearity evaluation means and the task conditions for the task to be evaluated that the nonlinearity evaluation means is about to evaluate.

3. 2. The work difficulty evaluation system according to claim 1, further comprising a means for evaluating the driving skill of a driver of the work machine based on the data relating to the output accumulated in the operation data acquisition means and the degree of nonlinearity.

4. A task difficulty assessment system according to claim 2, comprising: a task planning unit that stores task plans; The work planning system, wherein the work planning unit changes the work plan in accordance with the predicted difficulty level.

Images