Shape estimation device, property estimation device, crane, crane control system, shape estimation method, and shape estimation program

Abstract

To provide useful information for crane control.SOLUTION: A shape estimation device (1) includes: a detection acquisition unit (101) for acquiring detection results of a plurality of distance sensors provided on a surface facing an object of a bucket, which are detected when a part of the object being deposited is gripped and lifted by the bucket of a crane; and a shape estimation unit (102) that estimates a shape of the object gripped by the bucket based on the detection result.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a shape estimation device that estimates the shape of an object lifted by a crane equipped with a bucket, etc.

Background Art

[0002] The garbage introduced into a garbage treatment plant is temporarily stored in a storage facility called a garbage pit and then introduced into an incinerator and incinerated. For the transportation of garbage in the garbage pit, generally, a crane equipped with a bucket is used. For example, Patent Document 1 below discloses a technique for automatically controlling a crane equipped with a bucket to stir the garbage in the garbage pit.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The automatic control described in Patent Document 1 is based on the premise that the same volume of waste will always be lifted when the crane's bucket is made to grasp the waste. However, when the crane is actually made to grasp waste in a waste pit, the volume and weight of the waste lifted are often inconsistent depending on the condition of the waste at the position where the bucket is lowered. For example, if the waste at the position where the bucket is lowered is sloped or compacted, the volume and weight of waste lifted by the bucket may be significantly less. In such cases, after the bucket is hoisted up, an inefficient control method is required, in which the bucket is lowered again and grasped again after it is determined that the amount of waste lifted is insufficient by a weighing scale attached to the bucket's wire. Furthermore, although the bulk density of waste is useful as an indicator of its quality (combustibility), it was difficult to calculate the bulk density because the volume of waste lifted by the bucket could not be measured.

[0005] Thus, conventionally, the information available for crane control has been limited to weight information obtained from weighing scales as described above, which has hindered further improvements in crane control. This is a common problem not only for transporting waste but also for cranes with buckets that transport any accumulated object. One aspect of the present invention aims to provide information useful for crane control. [Means for solving the problem]

[0006] To solve the above problems, a shape estimation device according to one aspect of the present invention includes a detection result acquisition unit that acquires detection results indicating the distance from a distance sensor to an object, which are detected by a plurality of distance sensors provided on the surface of the bucket facing the object when the bucket of the crane grips and lifts a part of an object that is piled up, and a shape estimation unit that estimates the shape of the object being gripped by the bucket, or the shape of the object that is about to be gripped by the bucket, based on the detection results.

[0007] Furthermore, in order to solve the above problems, a property estimation device according to one aspect of the present invention includes a detection result acquisition unit that acquires a detection result indicating the distance from a distance sensor to an object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object when the bucket of the crane grips and lifts a part of the object that is piled up, and a property estimation unit that estimates the properties of the object being gripped by the bucket based on the detection result.

[0008] Furthermore, in order to solve the above problems, a crane according to one aspect of the present invention is provided with a plurality of distance sensors on the surface of the bucket facing the object to be transported, for detecting the distance to the object being gripped by the bucket, or the distance to the object being about to be gripped by the bucket.

[0009] Furthermore, in order to solve the above problems, a shape estimation method according to one aspect of the present invention is a shape estimation method performed by one or more information processing devices, comprising: a detection result acquisition step of acquiring detection results indicating the distance from a distance sensor to an object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object when a portion of an object piled up is grasped and lifted by the bucket of a crane; and a shape estimation step of estimating the shape of the object being grasped by the bucket, or the shape of the object being grasped by the bucket, based on the detection results. [Effects of the Invention]

[0010] According to one aspect of the present invention, it becomes possible to provide information useful for controlling a crane. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block diagram showing an example configuration of a shape estimation device according to Embodiment 1 of the present invention. [Figure 2]This figure shows the configuration of the crane control system, including the shape estimation device described above. [Figure 3] This diagram shows how multiple distance sensors detect the distance to the waste being held by the crane. [Figure 4] This figure shows an example of estimating the shape of an object gripped by a simulated crane. [Figure 5] This flowchart shows an example of the process performed by the shape estimation device described above. [Figure 6] This figure shows an overview of a crane control system according to Embodiment 2 of the present invention. [Figure 7] This block diagram shows an example of the main components of a shape estimation device included in the crane control system described above. [Figure 8] This flowchart shows an example of the process performed by the shape estimation device described above. [Figure 9] This is a block diagram showing an example of the main components of a property estimation device according to Embodiment 3 of the present invention. [Figure 10] This is a block diagram showing an example of the main components of a crane control device according to Embodiment 4 of the present invention. [Figure 11] This figure shows an example of control using the crane control device described above. [Figure 12] This flowchart shows an example of the process performed by the crane control device described above. [Modes for carrying out the invention]

[0012] [Embodiment 1] (System Configuration) Figure 2 shows the configuration of the crane control system 7 according to this embodiment. As shown in the figure, the crane control system 7 includes a shape estimation device 1, a crane control device 3, and a crane 5. The crane control system 7 is a system that controls the crane 5 using the shape estimation device 1 and the crane control device 3.

[0013] FIG. 2 shows an example in which the crane control system 7 is applied to the transportation of waste stored in a pit P, which is a waste storage facility. The waste may be any material that can be transported by the crane 5, such as household waste or industrial waste. Further, the crane control system 7 can be used for transporting any object other than waste. Therefore, the "waste" in the following description can be read as any object.

[0014] In the crane 5, a plurality of distance sensors 6 for detecting the distance to the waste held by the bucket are provided on the surface of the bucket facing the waste for lifting the waste. Further, according to this distance sensor 6, not only the waste being held but also the distance to the waste that is about to be held by the bucket (waste that has not been completely held and has not been lifted) can be detected. The principle of distance detection of the distance sensor 6 is not particularly limited. For example, an infrared distance sensor or an ultrasonic distance sensor may be used as the distance sensor 6, and both an infrared distance sensor and an ultrasonic distance sensor may be included in the plurality of distance sensors 6 used. Also, the distance sensor 6 may be a so-called proximity sensor.

[0015] The detection result of the distance sensor 6 indicates the shape of the waste held by the bucket or the shape of the waste that is about to be held by the bucket, and is information useful for controlling the crane 5. That is, according to the crane 5, information useful for controlling the crane 5 can be provided.

[0016] Note that the shape of the waste refers to the overall outer shape of the waste held by the bucket or the outer shape of a part of the waste that is about to be held by the bucket (the part pulled by the bucket), rather than the shape of each individual article contained in the waste.

[0017] The shape estimation device 1 then estimates the shape of the waste being held by the bucket, or the shape of the waste being held by the bucket, based on the detection results of multiple distance sensors 6. The crane control device 3, which controls the operation of the crane 5, then controls the crane 5 according to the estimation results from the shape estimation device 1.

[0018] In general, waste materials consist of a mixture of various items, so their properties are not consistent, and even if the crane 5 performs the same gripping operation, the amount and shape of the waste that can be lifted are not consistent. However, the crane control system 7 estimates the shape of the waste using the detection results of the distance sensor 6 installed on the crane 5, and controls the crane 5 according to the estimation result, making it possible to stably perform automatic control of the crane 5.

[0019] (Device configuration) The configuration of the shape estimation device 1 will be explained based on Figure 1. Figure 1 is a block diagram showing an example of the main components of the shape estimation device 1. As shown in the diagram, the shape estimation device 1 includes a control unit 10 that controls all parts of the shape estimation device 1, and a storage unit 11 that stores various data used by the shape estimation device 1. The shape estimation device 1 also includes a communication unit 12 for the shape estimation device 1 to communicate with other devices, an input unit 13 that receives input of various data to the shape estimation device 1, and an output unit 14 for the shape estimation device 1 to output various data. The control unit 10 also includes a detection result acquisition unit 101, a shape estimation unit 102, a volume calculation unit 103, a bulk density calculation unit 104, and a crane control unit 105.

[0020] The detection result acquisition unit 101 acquires detection results indicating the distance from the distance sensors 6 to the waste, which are detected by multiple distance sensors 6 provided on the surface of the bucket facing the waste when the bucket of the crane 5 grasps and lifts a portion of the accumulated waste. The detection result acquisition unit 101 may acquire the above detection results from the distance sensors 6 via the communication unit 12. This communication may be wired or wireless. Alternatively, a user of the shape estimation device 1 may input the detection results via the input unit 13.

[0021] The shape estimation unit 102 estimates the shape of the waste being held by the bucket, or the shape of the waste being held by the bucket, based on the detection results acquired by the detection result acquisition unit 101. In this embodiment, an example is described in which the shape estimation unit 102 estimates the shape of the waste being held by the bucket. Further details of the shape estimation method will be described in the "Shape Estimation Method" section below.

[0022] The volume calculation unit 103 calculates the volume of the waste held by the bucket from the shape estimated by the shape estimation unit 102. For example, if the shape estimation unit 102 calculates a function of the curved surface that represents the outer edge of the waste held by the bucket, the volume calculation unit 103 calculates the volume of the space enclosed by that curved surface.

[0023] The bulk density calculation unit 104 calculates the bulk density of the waste from the volume calculated by the volume calculation unit 103 and the weight of the waste held by the bucket. Specifically, the bulk density calculation unit 104 calculates the bulk density of the waste by dividing the weight of the waste by its volume. Bulk density is also called apparent density. The weight of the waste held by the bucket can be measured, for example, by attaching a weighing scale to the wire that suspends the bucket, so the bulk density calculation unit 104 can obtain the measured value from such a weighing scale via the input unit 13 or the communication unit 12.

[0024] The crane control unit 105 controls the crane 5 according to the shape estimated by the shape estimation unit 102. More specifically, the crane control unit 105 controls the crane 5 based on the bulk density calculated based on the shape estimated by the shape estimation unit 102. For example, the crane control unit 105 may cause the crane 5 to re-grasp the waste if the calculated bulk density exceeds a predetermined upper limit, or if the calculated bulk density is below a predetermined lower limit. This allows the crane 5 to transport waste with a desired bulk density. This is an effective control method, for example, when loading waste into a hopper connected to an incinerator.

[0025] The crane 5 is controlled via the crane control device 3 shown in Figure 2. Of course, it is also possible to control the crane 5 without using the crane control device 3; in this case, the shape estimation device 1 will perform the functions of the crane control device 3.

[0026] As described above, the shape estimation device 1 includes a detection result acquisition unit 101 that acquires detection results indicating the distance from the distance sensors 6 to the waste, which are detected by a plurality of distance sensors 6 provided on the surface of the bucket facing the waste when the bucket of the crane 5 grasps and lifts a portion of the accumulated waste, and a shape estimation unit 102 that estimates the shape of the waste being grasped by the bucket, or the shape of the waste being grasped by the bucket, based on the said detection results.

[0027] With the above configuration, it becomes possible to estimate the shape of the object being gripped by the bucket, or the shape of the object being gripped by the bucket. Since the shape of the object is useful information for controlling the crane 5, the above configuration can provide useful information for controlling the crane 5.

[0028] Furthermore, as described above, the shape estimation device 1 may also include a volume calculation unit 103 that calculates the volume of the waste held by the bucket from the shape estimated by the shape estimation unit 102, and a bulk density calculation unit 104 that calculates the bulk density of the waste held by the bucket from the volume calculated by the volume calculation unit 103 and the weight of the waste held by the bucket.

[0029] This provides useful information for understanding the properties of waste, specifically its bulk density. The bulk density of the waste can be used to control the crane 5 as described above, and can also be used to predict the combustion state when incinerating waste.

[0030] (Shape estimation method) The shape estimation method by the shape estimation unit 102 will be explained with reference to Figure 3. Figure 3 shows how the distance to the waste grasped by the crane 5 is detected by distance sensors 6a to 6f. In the following, when it is not necessary to distinguish between distance sensors 6a to 6f, they will simply be referred to as distance sensor 6.

[0031] The crane 5 shown in Figure 3 comprises a wire 51 and a bucket 52 suspended from the wire 51. The bucket 52 comprises a base 521 connected to the wire 51 and claws 522 for grasping waste, which are rotatably connected to the base 521. Although only two claws 522 are shown in Figure 3 for illustrative purposes, only enough claws 522 should be provided to stably grasp the waste.

[0032] Furthermore, distance sensors 6 are provided on the inside (the side facing the waste) of the claw 522 of the crane 5 shown in Figure 3. Specifically, distance sensors 6a, 6b, and 6c are provided in order on one claw 522, from the side connected to the base 521 to the tip. Similarly, distance sensors 6d, 6e, and 6f are provided in order on the other claw 522, from the side connected to the base 521 to the tip.

[0033] The distance sensor 6 shown in Figure 3 is embedded inside the claw 522. By embedding the distance sensor 6 inside the claw 522, the possibility of damage to the distance sensor 6 due to contact with waste can be reduced. Of course, the distance sensor 6 may be placed on the surface of the claw 522 without being embedded. Even when embedded, a part of the distance sensor 6 may protrude from the surface of the claw 522. In addition, although the distance sensor 6 is not provided on the inside of the base 521 (the surface facing the waste) in the example of Figure 3, the distance sensor 6 may also be provided on the inside of the base 521.

[0034] Thus, the location and manner in which the distance sensors 6 are installed on the inner surface of the bucket 52 are arbitrary. Furthermore, the number of distance sensors 6 installed is not particularly limited. However, from the viewpoint of improving the accuracy of shape estimation, it is preferable to install the distance sensors 6 at various locations so that distance measurements can be taken from as many different directions as possible with respect to the waste held in the bucket 52.

[0035] In the example shown in Figure 3, the waste G is held in the bucket 52. In Figure 3, the distances from the distance sensors 6a to 6f to the waste G are shown by dashed lines. The position of a point on the surface of the waste G can be determined from the distance detected by the distance sensor 6 and the opening angle of the claws 522.

[0036] Specifically, the position of point p1 on the surface of waste G can be determined from the distance detected by distance sensor 6a. Similarly, the positions of points p2 to p6 on the surface of waste G can be determined from the distances detected by distance sensors 6b to 6f. These positions can also be represented by coordinate values. In the following, the set of points on the surface of waste G identified from the distance detected by distance sensor 6a will be referred to as point cloud data.

[0037] The shape estimation unit 102 can estimate the shape of the waste from the point cloud data identified as described above from the detection results of each distance sensor 6. Various techniques can be applied to estimate the shape from the point cloud data.

[0038] For example, the shape estimation unit 102 may estimate the shape using GPIS (Gaussian Process Implicit Surface). GPIS is a method that combines a Gaussian process, which learns a function from data, and an implicit surface, which represents the shape using the function. According to GPIS, the shape of the object surface can be estimated using point cloud data representing the object surface. Using GPIS is preferable because it makes it possible to estimate the shape of the waste G with high accuracy. However, since shape estimation using GPIS is computationally intensive, if you want to perform shape estimation using GPIS at high speed, it is preferable to use parallel computing using a GPU (Graphical Processing Unit).

[0039] Of course, the algorithm for shape estimation is not limited to GPIS. For example, it is possible to represent the shape using "Implicit Surface" in GPIS, learn the function used for shape representation using a machine learning method other than a Gaussian process (e.g., a neural network), and then estimate the shape. It is also possible to represent the shape using a representation method other than "Implicit Surface," and then estimate the shape using a shape estimation model learned by any machine learning method such as a Gaussian process or a neural network. In addition to these, it is also possible to estimate the shape using methods such as Deep Geometric Prior, the ball pivot algorithm, or the Delaunee partition method.

[0040] As described above, the shape estimation unit 102 may estimate the shape of the waste using a machine learning method from the detection result of the distance sensor 6. This makes it possible to estimate the shape of the waste with high accuracy. The detection result of the distance sensor 6 may be used as an explanatory variable as is, or features generated based on the detection result may be used as an explanatory variable. Furthermore, if a shape estimation model that requires prior training is used, the trained model is generated in advance and stored in the memory unit 11, etc. Also, if a function used for shape representation is trained in advance, that function is stored in the memory unit 11, etc.

[0041] (Example of distance sensor placement and shape estimation) An experiment was conducted to estimate the shape of an object using the detection results of a distance sensor. This will be explained with reference to Figure 4. Figure 4 shows an example of estimating the shape of an object X gripped by a simulated crane 5A.

[0042] As shown in IMG1 of Figure 4, the simulated crane 5A is a small crane created to resemble the crane 5, and has a configuration in which four claws 522A are connected to the base 521A of the bucket. Also, as shown in IMG2 of Figure 4, multiple distance sensors 6A are provided on the inner surface of the claws 522A. More specifically, seven distance sensors 6A are provided on the inner surface of the claws 522A. The distance sensors 6A are distributed fairly evenly across the entire curved inner surface of the claws 522A, so that the distance to the object being gripped can be measured from various directions. Similarly, seven distance sensors 6A are also provided on each of the other three claws 522A. In other words, the simulated crane 5A has a total of 28 distance sensors 6A.

[0043] The object X, gripped by the simulated crane 5A, is a sphere as shown in IMG3 of Figure 4. While the object X is gripped by the simulated crane 5A, the distance is detected by each distance sensor 6A, and the point cloud data obtained from these detection results is shown in IMG4 of Figure 4. Then, IMG5 of Figure 4 shows the estimated shape of the object X obtained from this point cloud data using GPIS. The estimated shape shown in IMG5 closely matches the shape of the object X shown in IMG3. Thus, it is possible to estimate the shape of the object X with high accuracy using GPIS.

[0044] (Process flow) The flow of the process (shape estimation method) performed by the shape estimation device 1 will be explained based on Figure 5. Figure 5 is a flowchart showing an example of the process performed by the shape estimation device 1. In the following, the process of feeding the waste in pit P shown in Figure 2 into the hopper connected to the incinerator will be explained.

[0045] In S11, the crane control unit 105 instructs the crane 5 to perform a gripping operation. For example, the crane control unit 105 may notify the crane control device 3 of the position in the pit P where the gripping operation will be performed, and instruct it to perform the following series of processes. The series of processes involves moving the bucket 52 to the notified position, opening the bucket 52 at that position, lowering the bucket 52 onto the surface of the waste in that state, closing the bucket 52 to grip the waste, and winding up the wire 51 to lift the gripped waste.

[0046] In S12, the detection result acquisition unit 101 acquires the detection results of the distances detected by each distance sensor 6 after the waste has been lifted by the processing in S11. These detection results show the distance from each distance sensor 6 to the lifted waste.

[0047] In S13, the shape estimation unit 102 estimates the shape of the waste held by the bucket 52 based on the detection result in S12. As described above, the method of estimating the shape is not particularly limited. For example, the shape estimation unit 102 may generate the point cloud data described above based on the detection result in S12 and estimate the shape of the waste held by the GPS from the generated point cloud data. The format in which the shape estimation result is expressed is arbitrary. For example, the shape estimation unit 102 may output a function as information indicating the estimation result.

[0048] In S14, the volume calculation unit 103 calculates the volume of the waste held by the bucket 52 based on the shape estimated in S13. Then, in S15, the bulk density calculation unit 104 calculates the bulk density of the waste from the volume calculated in S14 and the weight of the waste held by the bucket 52. As mentioned above, the weight of the waste may be measured using a weighing scale.

[0049] In S16, the crane control unit 105 determines whether the bulk density calculated in S15 is below a predetermined upper limit. If the result in S16 is YES, the process proceeds to S17; if the result in S16 is NO, the process proceeds to S18.

[0050] In S18, the crane control unit 105 instructs the crane control device 3 to open the bucket 52 of the crane 5. As a result, the waste that was held in the bucket 52 by the process in S11 is released, and the bucket 52 becomes empty. After this, the process returns to S11 and the gripping operation is performed again. The next gripping operation may be performed at the same location in the pit P as the previous gripping operation, or at a different location.

[0051] Meanwhile, in S17, the crane control unit 105 instructs the crane control device 3 to execute the process of feeding waste into the hopper, and the process shown in Figure 5 is completed. As a result, the waste that was held in the bucket 52 by the process in S11 is fed into the hopper and put to incineration. According to the process shown in Figure 5, since waste with a bulk density below a predetermined upper limit is fed into the hopper, combustion in the incinerator can be stabilized.

[0052] As described above, the shape estimation method according to this embodiment includes a detection result acquisition step (S12) in which detection results indicating the distance from the distance sensors 6 to the waste are obtained, which are detected by a plurality of distance sensors 6 provided on the surface of the bucket 52 facing the waste when the bucket 52 of the crane 5 grasps and lifts a portion of the accumulated waste, and a shape estimation step (S13) in which the shape of the waste being grasped by the bucket 52 is estimated based on the detection results of S12. Thus, it is possible to provide information useful for controlling the crane 5, such as the shape of the waste being grasped by the bucket 52. It is also possible to estimate the shape of the waste that is about to be grasped by the bucket 52 by a similar process.

[0053] [Embodiment 2] Other embodiments of the present invention are described below. For the sake of clarity, components having the same function as those described in the above embodiments will be denoted by the same reference numerals, and their descriptions will not be repeated. This also applies to Embodiment 3 and subsequent embodiments.

[0054] (overview) In this embodiment, we will describe a crane control system 7A that estimates the shape of the waste to be gripped by the crane's bucket, predicts the weight of the gripped waste from the change in shape during the gripping operation, and controls the crane based on this prediction result.

[0055] Figure 6 shows an overview of the crane control system 7A. As shown in the figure, the crane control system 7A includes a shape estimation device 1A and a crane control device 3. Similar to the shape estimation device 1 of Embodiment 1, the shape estimation device 1A acquires the detection result of the distance sensor 6 provided on the bucket 52, notifies the crane control device 3 of the control content corresponding to the detection result, and the crane control device 3 controls the crane 5 according to this notification.

[0056] The difference between the shape estimation device 1A of this embodiment and the shape estimation device 1 of Embodiment 1 is that the shape estimation device 1A estimates the shape of the waste to be grasped by the bucket 52 and predicts the weight of the grasped waste based on its shape during the grasping operation.

[0057] ST1 to ST3 in Figure 6 show the flow of the gripping operation. The gripping operation begins with lowering the bucket 52 onto the waste G in an open state, as shown in ST1 of Figure 6. By lowering the bucket 52 onto the waste G in an open state, the tips of the claws 522 bite into the waste G. Subsequently, by closing the claws 522, the waste G is scraped by the closing claws 522 (ST2), and the gripping operation ends when the opening angle of the claws 522 reaches a predetermined angle or greater than the angle at which the waste G can be lifted (ST3). After this, the waste G held by the claws 522 is lifted by winding up the bucket 52.

[0058] Here, even when the same control is used to make the crane 5 perform a gripping operation, the amount of waste G that can be lifted is not necessarily the same. For example, if the accumulated waste G is compacted, the claws 522 may not be able to sufficiently penetrate the waste G at ST1. In this case, the amount of waste G that is scooped up at ST2 will be less, and the amount of waste G that is held in place by the claws 522 at ST3 will also be less, resulting in a smaller amount of waste G being lifted in the end. Also, if the surface of the accumulated waste G is sloped, the claws 522 may not be able to sufficiently penetrate the waste G at ST1, which may also reduce the amount of waste G that can be lifted. Since transporting small amounts of waste G is inefficient, in such cases it is preferable to open the bucket 52 to release the waste G and then re-grip the waste G.

[0059] In conventional crane automated control systems, the recognition of a small amount of lifted waste could only be detected after the waste had been lifted and its weight was measured by a scale attached to the wire. As a result, re-grabbing of the waste was done after the waste had been lifted, causing a time loss due to the need to temporarily lift small amounts of waste that were not worth transporting.

[0060] The shape estimation device 1A can predict the weight of the waste G to be grasped before it is lifted, thus enabling efficient transportation without the time loss described above. This prediction utilizes the relationship between the weight of the waste G to be grasped and the shape of the waste G during the grasping operation.

[0061] In other words, when the weight of the waste G being grasped is small, the degree to which the waste G bulges during the grasping operation remains within a small range, while when the weight of the waste G being grasped is large, the degree to which the waste G bulges during the grasping operation remains within a large range. The shape estimation device 1A utilizes this relationship to estimate the shape of the waste G from the detection result of the distance sensor 6, and predicts the weight of the waste G being grasped from the estimated shape.

[0062] The above predictions can also be made using a weight prediction model constructed by machine learning, where features corresponding to the shape of the waste G are used as explanatory variables and the weight of the waste G is used as the dependent variable. For example, the volume of the raised portion scraped up by the claws 522 at a predetermined timing, such as in stage ST3, may be calculated and used as the above feature.

[0063] The training data for constructing the weight prediction model described above can also be automatically generated by repeatedly performing trials in which the crane 5 performs a gripping operation, the shape of the waste is estimated during the gripping operation, and the weight of the gripped waste is measured using the weighing scale described above after the completion of each gripping operation. In other words, training data can be generated by generating feature quantities to be used as explanatory variables from the results of the shape estimation, and then associating these feature quantities with the measurement results from the weighing scale as ground truth data.

[0064] Furthermore, the above explanatory variables may also be used as features that show the changes in the estimated shape. For example, the rate of increase in the volume of the part that is raised by being scraped up by the claws 522 may be calculated and used as the above feature. In this case, there is an advantage in that the weight can be predicted before the closing operation of the bucket 52 is completed (before ST3).

[0065] When using a weight prediction model that uses the rate of increase in volume of the raised portion scraped by the claws 522 as an explanatory variable, the shape estimation device 1A may acquire the detection result of the distance sensor 6 at ST1, when the claws 522 begin to close, and estimate the shape of the accumulated waste G. Then, the shape estimation device 1A may acquire the detection result of the distance sensor 6 again at ST2, when a predetermined time has elapsed after the claws 522 begin to close, and estimate the shape of the waste G. Then, the shape estimation device 1A may calculate the rate of increase in volume of the raised portion scraped by the claws 522 from the difference in the estimated shapes and input it into the weight prediction model to calculate the predicted weight.

[0066] Furthermore, for example, when using a weight prediction model that uses the volume of the raised portion scraped up by the claws 522 as an explanatory variable, the shape estimation device 1A may acquire the detection result of the distance sensor 6 at ST3, when the closing operation of the bucket 52 is completed, and estimate the shape of the accumulated waste. The shape estimation device 1A may then calculate the volume of the raised portion scraped up by the claws 522 from the estimated shape and input it into the weight prediction model to calculate the predicted weight.

[0067] (Device configuration) The configuration of the shape estimation device 1A will be explained based on Figure 7. Figure 7 is a block diagram showing an example of the main components of the shape estimation device 1A. The shape estimation device 1A differs from the shape estimation device 1 shown in Figure 1 in that the control unit 10 has been replaced by a control unit 10A. The control unit 10A includes a detection result acquisition unit 101, a shape estimation unit 102, a weight prediction unit 103A, and a crane control unit 105.

[0068] The weight prediction unit 103A predicts the weight of the waste grasped by the bucket 52 based on the shape estimated by the shape estimation unit 102 during the gripping operation of the bucket 52. As explained in the "Overview" section above, the method of weight prediction is not particularly limited, and for example, the weight prediction unit 103A may predict the weight using a weight prediction model constructed by machine learning.

[0069] As described above, the shape estimation device 1A includes a weight prediction unit 103A that predicts the weight of the waste to be grasped by the bucket 52 based on the shape estimated by the shape estimation unit 102 during the grasping operation of the bucket 52, when the shape estimation unit 102 estimates the shape of the waste to be grasped by the bucket 52.

[0070] This makes it possible to predict the weight of waste that will be lifted by the bucket 52 before the bucket 52 has finished gripping and lifting the waste. For example, it is possible to predict that the desired weight of waste may not be lifted before the bucket 52 has finished gripping the waste. Based on such predictions, it becomes possible to perform processing such as re-gripping at an earlier stage, thereby making the transportation of waste by the crane 5 more efficient.

[0071] (Process flow) The flow of the process (shape estimation method) performed by the shape estimation device 1A will be explained based on Figure 8. Figure 8 is a flowchart showing an example of the process performed by the shape estimation device 1A. In the following, the process of transferring waste from a predetermined position in pit P shown in Figure 2 to another position will be explained.

[0072] In S21, the crane control unit 105 moves the bucket 52 of the crane 5 to a predetermined position, lowers the bucket 52 onto the waste at that position, and then starts the gripping operation. Then, in S22, the detection result acquisition unit 101 acquires the distance detection results from each distance sensor 6, and in S23, the shape estimation unit 102 estimates the shape of the waste that is about to be gripped by the bucket 52 based on the detection results acquired in S22.

[0073] The S22 process is performed at a predetermined timing during the period from the start to the end of the gripping operation (from ST1 to ST3 in Figure 6) (for example, at a predetermined time after the closing operation of bucket 52 has started).

[0074] In S24, the detection result acquisition unit 101 determines whether or not to terminate the shape estimation. For example, if two shape measurements are performed during the period from the start to the end of the grasping operation, the detection result acquisition unit 101 determines to terminate the shape estimation (YES in S24) if the process in S22 has been executed twice, and determines not to terminate the shape estimation (NO in S24) if the process in S22 has been executed once. If YES is determined in S24, the process proceeds to S25, and if NO is determined in S24, the process returns to S22.

[0075] In S25, the weight prediction unit 103A predicts the weight of the waste to be gripped by the bucket 52 based on the shape estimated in S23. The method for predicting weight based on the estimated shape has already been explained, so it will not be explained again here.

[0076] In S26, the crane control unit 105 determines whether the weight predicted in S25 is above a predetermined lower limit. If the result in S26 is YES, the process proceeds to S27; if the result in S26 is NO, the process proceeds to S28. The process in S26 is to determine whether or not to re-grasp the waste. In addition, in S26, it may also be determined whether the predicted weight is above a predetermined upper limit, or whether the predicted weight is outside a predetermined appropriate range.

[0077] In S28, the crane control unit 105 instructs the crane control device 3 to open the bucket 52 of the crane 5 and then hoist it up. As a result, the waste that was held in the bucket 52 by the process in S21 is released, the bucket 52 becomes empty, and is hoisted up. After this, the process returns to S21 and the gripping operation is performed again. The next gripping operation may be performed at the exact same predetermined position in the pit P as the previous gripping operation, or at a position slightly shifted from the predetermined position.

[0078] Meanwhile, in S27, the crane control unit 105 instructs the crane control device 3 to hoist up the bucket 52 and transport the waste to a predetermined transfer destination. This completes the process shown in Figure 8. According to the process in Figure 8, it is possible to re-grasp the waste at an earlier timing according to the prediction result in S25, thus enabling efficient waste transfer.

[0079] [Embodiment 3] (Device configuration) The configuration of the property estimation device 2 according to this embodiment will be explained with reference to Figure 9. Figure 9 is a block diagram showing an example of the main components of the property estimation device 2. The property estimation device 2 is a device that estimates the properties of waste held by the bucket 52 based on the distance detection result by the distance sensor 6 provided on the bucket 52. The property estimation device 2 also has a function to control the crane 5 based on the property estimation result.

[0080] As shown in Figure 9, the property estimation device 2 includes a control unit 20 that controls all parts of the property estimation device 2, and a storage unit 21 that stores various data used by the property estimation device 2. The property estimation device 2 also includes a communication unit 22 for the property estimation device 2 to communicate with other devices, an input unit 23 that receives input of various data to the property estimation device 2, and an output unit 24 for the property estimation device 2 to output various data. The control unit 20 also includes a detection result acquisition unit 201, a property estimation unit 202, and a crane control unit 203.

[0081] Similar to the detection result acquisition unit 101 described in Embodiment 1, the detection result acquisition unit 201 acquires a detection result indicating the distance from the distance sensors 6 to the waste, which is detected by a plurality of distance sensors 6 provided on the surface of the bucket facing the waste when the bucket of the crane 5 grasps and lifts a portion of the accumulated waste.

[0082] The property estimation unit 202 estimates the properties of the waste held by the bucket 52 based on the detection results acquired by the detection result acquisition unit 201. The properties to be estimated are those whose differences in properties are reflected in the differences in the shape of the waste when held by the bucket 52. For example, the above properties may be the weight, volume, bulk density, moisture content, type, quality, etc. of the waste.

[0083] The properties can be estimated, for example, using a property estimation model constructed by machine learning the relationship between variables, with the detection results of the distance sensor 6 or features generated from those detection results as explanatory variables and the properties to be estimated as the dependent variable. For example, a property estimation model for estimating bulk density can be generated using training data that associates the bulk density of the waste held in the bucket 52 with the detection results of the distance sensor 6 installed on the bucket 52.

[0084] Furthermore, for example, a property estimation model for estimating the quality of waste can be generated using training data that associates the quality of the waste held in the bucket 52 with the detection results of the distance sensor 6 installed on the bucket 52. The criteria for evaluating the quality of waste are arbitrary; for example, the cumulative number of times the waste has been agitated can be estimated as a value indicating the quality of the waste. Here, agitation of waste refers to lifting the waste with the bucket 52 and dropping it. By performing such agitation, the waste is homogenized, for example, by tearing the bag containing the waste.

[0085] The crane control unit 203 controls the crane 5 based on the property estimation results from the property estimation unit 202. The property estimation results and the corresponding control content can be predetermined. For example, when the property estimation unit 202 calculates bulk density, the control performed by the crane control unit 203 is the same as that of the crane control unit 105 in Embodiment 1. The crane 5 may be controlled via the crane control device 3 shown in Figure 2, or it is possible to control the crane 5 without using the crane control device 3.

[0086] As described above, the property estimation device 2 includes a detection result acquisition unit 201 that acquires detection results indicating the distance from the distance sensors 6 to the waste, which are detected by a plurality of distance sensors 6 provided on the surface of the bucket 52 facing the waste when the bucket 52 of the crane 5 grasps and lifts a portion of the accumulated waste, and a property estimation unit 202 that estimates the properties of the waste grasped by the bucket 52 based on the detection results.

[0087] Since the properties of the waste held in the bucket 52 are reflected in the detection results of the distance sensor 6, the above configuration makes it possible to estimate the properties of the waste held in the bucket 52. Furthermore, since the properties of the waste are useful information for controlling the crane 5, the above configuration makes it possible to provide useful information for controlling the crane 5.

[0088] The property estimation device 2 also determines the control content of the crane 5 based on the estimated properties. However, the determination of the control content of the crane 5 may be performed by, for example, the operator of the crane 5, or by another device such as the crane control device 3. When the operator determines the control content of the crane 5, the property estimation device 2 should present the estimated properties to the operator by outputting them to the output unit 24, etc., so that the crane 5 can be controlled taking into account the presented properties.

[0089] Furthermore, the flow of the processing (property estimation method) performed by the property estimation device 2 is generally the same as the processing in Figure 5. However, after the detection result acquisition unit 201 acquires the detection result in S12 (detection result acquisition step), property estimation is performed by the property estimation unit 202 instead of the processing in S13 to S15 (property estimation step). The processing of the crane control unit 203 after the property estimation can be appropriately determined according to the property to be estimated.

[0090] [Embodiment 4] (Device configuration) In this embodiment, a crane control device 3A that controls the crane 5 according to the detection result of the distance sensor 6 without estimating the shape or properties of the waste will be described. Figure 10 is a block diagram showing an example of the main components of the crane control device 3A. The crane control device 3A is a device that controls the operation of the crane 5 based on the distance detection result of the distance sensor 6 provided on the bucket 52.

[0091] As shown in the figure, the crane control device 3A includes a control unit 30A that controls all parts of the crane control device 3A, and a storage unit 31A that stores various data used by the crane control device 3A. The crane control device 3A also includes a communication unit 32A for the crane control device 3A to communicate with other devices, an input unit 33A that receives input of various data to the crane control device 3A, and an output unit 34A for the crane control device 3A to output various data. The control unit 30A also includes a detection result acquisition unit 301A and a crane control unit 302A.

[0092] Similar to the detection result acquisition unit 101 described in Embodiment 1, the detection result acquisition unit 301A acquires a detection result indicating the distance from the distance sensors 6 to the waste, which is detected by a plurality of distance sensors 6 provided on the surface of the bucket facing the waste when the bucket of the crane 5 grips and lifts a portion of the accumulated waste.

[0093] The crane control unit 302A controls the operation of the crane 5 based on the detection results acquired by the detection result acquisition unit 301A. The control content can be determined, for example, using a control content estimation model constructed by machine learning the relationship between the detection results of the distance sensor 6 or features generated from those detection results as explanatory variables, and the content of the control to be performed when those detection results are obtained as the objective variable. As described in each of the embodiments above, various estimations related to the control content of the crane 5 (for example, estimation of the shape and properties of waste) can be made based on the detection results of multiple distance sensors 6. And, since the appropriate control content is determined according to these estimation results, it is also possible to determine the control content of the crane 5 based on the detection results of the distance sensor 6.

[0094] For example, the crane control device 3A can also decide whether or not to feed the grasped waste into the hopper by a process similar to that in Figure 5. In this case, instead of the processes in S13 to S16, the crane control unit 302A performs a process to decide whether or not to feed the grasped waste into the hopper based on the detection result obtained in S12. Alternatively, for example, the crane control device 3A can also decide whether or not to re-grasp the waste by a process similar to that in Figure 8. In this case, instead of the processes in S23 to S26, the crane control unit 302A performs a process to decide whether or not to re-grasp the waste based on the detection result obtained in S22.

[0095] As described above, the crane control device 3A includes a detection result acquisition unit 301A that acquires detection results indicating the distance from the distance sensors 6 to the waste, which are detected by a plurality of distance sensors 6 provided on the surface of the bucket facing the waste when the bucket of the crane 5 grasps and lifts a portion of the accumulated waste, and a crane control unit 302A that controls the operation of the crane 5 based on the detection results. With this configuration, automatic control of the crane 5 is realized using information useful for controlling the crane 5, which is the detection results of a plurality of distance sensors 6 provided on the bucket 52.

[0096] (Example of control) Figure 11 shows an example of control by the crane control device 3A. More specifically, Figure 11 shows an example in which the crane control device 3A controls the opening and closing of the bucket 52 during a scattering operation, which is the operation of scattering waste along the movement path of the bucket 52 while moving the bucket 52.

[0097] As described above, waste agitation refers to the process of lifting waste with bucket 52 and dropping it, with the aim of homogenizing the waste in pit P. Waste agitation can be performed by dropping the lifted waste in place, or it can be agitated by the scattering action described above. Hereafter, agitation by scattering action will be referred to as scattering agitation. In scattering agitation, it is required to scatter the waste evenly along the path of movement of bucket 52.

[0098] In the example shown in Figure 11, the bucket 52 of the crane 5 is moved from position A1 to A2. During this time, the bucket 52 is repeatedly opened and closed slightly, causing the waste to fall little by little. If the opening of the bucket 52 is too large or the time the bucket 52 is open is too long, a large amount of waste will fall at once, resulting in uneven distribution. Conversely, if the opening of the bucket 52 is too small or the time the bucket 52 is open is too short, a lot of waste will remain in the bucket 52 even when it reaches position A2, and this waste will fall at position A2, resulting in uneven distribution. Furthermore, since the quality of waste is generally not constant, even if the same control is used to pick up the waste, the same amount of waste may not be picked up. Also, even if the opening angle and opening time of the bucket 52 during distribution are the same, the same amount of waste may not fall.

[0099] Therefore, in the crane control device 3A, the detection result acquisition unit 301A continuously acquires detection results indicating the distance from the distance sensor 6 to the waste during scattering and mixing. Then, the crane control unit 302A controls the opening and closing of the bucket 52 so that the waste is evenly scattered along the movement path of the bucket 52, based on the changes in the detection results acquired by the detection result acquisition unit 301A. This makes it possible to automatically perform the difficult task of evenly scattering the waste along the movement path of the bucket 52.

[0100] The control for evenly distributing waste along the movement path of the bucket 52 can be determined using the control content estimation model described above. The control content only needs to be able to vary the amount of waste distributed. For example, the crane control unit 302A may adjust at least one of the bucket 52 opening degree, the execution time of the opening operation, and the execution time of the closing operation. In this case, the crane control unit 302A can use a control content estimation model that takes at least one of the bucket 52 opening degree, the execution time of the opening operation, and the execution time of the closing operation as the objective variable. It should be noted that, as will be explained below, it is also possible to determine the control content without using a control content estimation model.

[0101] (Process flow) Figure 12 is a flowchart showing an example of the process (crane control method) executed by the crane control device 3A. Note that Figure 12 shows the process during scattering and mixing. Furthermore, the following section describes an example of determining the control content without using the control content estimation model described above.

[0102] In S41, the crane control unit 302A causes the crane 5 to start the scattering operation. Specifically, the crane control unit 302A specifies a position in the pit P (the starting position of the scattering operation) and causes the crane 5 to grasp the waste at that position. The crane control unit 302A also specifies another position in the pit P (the ending position of the scattering operation) and causes the bucket 52 to start moving toward that position, as well as to start the opening and closing operation of the bucket 52.

[0103] In S42, the crane control unit 302A determines whether a single opening and closing operation, that is, a series of operations in which the bucket 52 is opened to a predetermined degree and then closed, has been completed. The process in S42 is repeated until it is determined that a single opening and closing operation has been completed (YES in S42). If YES is determined in S42, the process proceeds to S43.

[0104] In S43, the detection result acquisition unit 301A acquires the detection results of the distance sensor 6 detected during the period from the start to the end of one opening / closing operation which was determined to have ended in S42. The detection result acquisition unit 301A may acquire all detection results detected during the period from the start to the end of the opening / closing operation, or it may acquire only some of the detection results. For example, the detection result acquisition unit 301A may acquire the detection results at the start and end of the opening / closing operation.

[0105] In S44, the crane control unit 302A adjusts the content of the next opening and closing operation, which was determined to have ended in S42. For example, if the detection results for the start and end of the opening and closing operation are obtained in S43, the crane control unit 302A may calculate the difference between the average value of the distance detection results at the start of the opening and closing operation and the average value of the distance detection results at the end of the operation. If the calculated difference is greater than a predetermined standard value, the crane control unit 302A may adjust the operation so that the amount of waste dropped in the next opening and closing operation decreases, and if the calculated difference is less than a predetermined standard value, the operation may adjust the operation so that the amount of waste dropped in the next opening and closing operation increases.

[0106] The above reference value may be, for example, the difference between the average distance detection result at the start of an opening / closing operation and the average distance detection result at the end of the operation, when an ideal amount of waste falls in one opening / closing operation. Adjustments to reduce the amount of waste that falls include, for example, lowering the opening degree of the bucket 52, increasing the opening operation time (opening for a longer period of time), and decreasing the closing operation time (closing in a shorter period of time). To increase the amount of waste that falls, the opposite adjustments can be made.

[0107] In S45, the crane control unit 302A causes the crane 5 to perform the following opening and closing operation based on the adjustments made in S44. Thus, in S44 and S45, the crane control unit 302A controls the opening and closing of the bucket 52 so that the waste is evenly distributed along the movement path of the bucket 52, based on the progression of the detection results acquired by the detection result acquisition unit 301A.

[0108] In S46, the crane control unit 302A determines whether the number of times the bucket 52 has been opened and closed has reached a predetermined number. If the result in S46 is YES, the crane control unit 302A finishes adjusting the opening and closing operation and, when the bucket 52 reaches the end position of the scattering operation, causes the bucket 52 to open, and the process in Figure 12 ends. On the other hand, if the result in S46 is NO, the process returns to S42.

[0109] [Variation] The entity executing each process described in the above embodiments is arbitrary and not limited to the examples given above. In other words, the devices constituting the crane control systems 7 and 7A can be changed as appropriate, as long as each process described in the above embodiments can be executed.

[0110] For example, in the crane control system 7 shown in Figure 2, the shape estimation device 1 and the crane control device 3 are separate devices, but they can also be combined into a single device. The same applies to the crane control system 7A shown in Figure 6. Furthermore, the execution entity for each process described in Figures 5, 8, and 12 does not necessarily have to be a single device; these processes can be divided and executed by multiple arbitrary information processing devices (computers).

[0111] Furthermore, in each of the above embodiments, examples were described in which a trained model constructed by machine learning is used. The shape estimation device 1, 1A, property estimation device 2, and crane control devices 3, 3A may be equipped with a function to construct such a trained model or a function to generate training data to be used for training.

[0112] [Examples of implementation using software] The functions of the shape estimation devices 1 and 1A, the property estimation device 2, and the crane control devices 3 and 3A (hereinafter referred to as "devices") can be realized by programs that cause a computer to function as the device, and by programs that cause a computer to function as each control block of the device (especially each part included in the control unit 10 or 10A) (shape estimation program / property estimation program / crane control program).

[0113] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this control device and storage device, the functions described in each of the embodiments are realized.

[0114] The above program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the above device. In the latter case, the program may be supplied to the above device via any wired or wireless transmission medium.

[0115] Furthermore, some or all of the functions of each of the above control blocks can also be realized by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the above control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to realize the functions of each of the above control blocks by, for example, a quantum computer.

[0116] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of Symbols]

[0117] 1, 1A shape estimation device 101 Detection result acquisition unit 102 Shape estimation section 103 Volume Calculation Unit 103A Weight prediction unit 104 Bulk Specific Gravity Calculation Unit 105 Crane Control Unit 2 Property estimation device 201 Detection Result Acquisition Unit 202 Property Estimation Department 203 Crane Control Unit 3.3A Crane control device 301A Detection Result Acquisition Unit 302A Crane Control Unit 5 Cranes 52 buckets 6. Distance Sensor

Claims

1. When a crane's bucket grasps and lifts a portion of an object that is piled up, a detection result acquisition unit acquires a detection result indicating the distance from the distance sensors to the object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object. The system includes a shape estimation unit that estimates the shape of the object being gripped by the bucket based on the detection results, The distance sensor is a shape estimation device, wherein the bucket has multiple claws for grasping the object, and each of the multiple claws is rotatably mounted relative to the base of the bucket, and multiple claws are provided on the surface of the claw facing the object, from the side connected to the base to the tip.

2. The shape estimation device according to claim 1, wherein the shape estimation unit estimates the shape of the object using a machine learning method from the detection result of the distance sensor.

3. When a crane's bucket grasps and lifts a portion of an object that is piled up, a detection result acquisition unit acquires a detection result indicating the distance from the distance sensors to the object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object. The system includes a shape estimation unit that estimates the shape of the object being gripped by the bucket, or the shape of the object being gripped by the bucket, based on the detection results. The shape estimation unit estimates the shape of the object being held by the bucket, A volume calculation unit calculates the volume of the object being held by the bucket based on the shape estimated by the shape estimation unit, A shape estimation device comprising: a volume calculation unit that calculates the bulk density of an object held by a bucket from the volume calculated by the volume calculation unit and the weight of the object held by the bucket.

4. When a crane's bucket grasps and lifts a portion of an object that is piled up, a detection result acquisition unit acquires a detection result indicating the distance from the distance sensors to the object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object. The system includes a shape estimation unit that estimates the shape of the object being gripped by the bucket, or the shape of the object being gripped by the bucket, based on the detection results. The shape estimation unit estimates the shape of the object that is about to be gripped by the bucket, A shape estimation device comprising: a weight prediction unit that predicts the weight of the object to be gripped by the bucket based on the shape of the object estimated by the shape estimation unit during the gripping operation of the bucket.

5. When a crane's bucket grasps and lifts a portion of an object that is piled up, a detection result acquisition unit acquires a detection result indicating the distance from the distance sensors to the object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object. A property estimation device comprising: a property estimation unit that estimates the properties of the object being held by the bucket based on the detection results.

6. Multiple distance sensors are provided on the surface of the bucket facing the object being transported, for detecting the distance to the object being held by the bucket. The distance sensor is provided on a crane, which is a plurality of claws provided on a bucket for grasping an object, and each of the plurality of claws is rotatably mounted relative to the base of the bucket, with the distance sensor provided on the surface of the claw facing the object, from the side connected to the base to the tip.

7. The crane according to claim 6, A crane control device comprising: a detection result acquisition unit that acquires the detection result of the distance sensor provided by the crane according to claim 6; and a crane control unit that controls the operation of the crane based on the detection result, A crane control system including a crane control system.

8. The crane control system according to claim 7, wherein the crane control unit controls the opening and closing of the bucket so that the object is evenly scattered along the movement path of the bucket, based on the progression of the detection results, during a scattering operation which is an operation in which the bucket is moved and the object is scattered along the movement path of the bucket.

9. A shape estimation method performed by one or more information processing devices, A detection result acquisition step involves acquiring detection results indicating the distance from a distance sensor to an object, which is detected by a plurality of distance sensors provided on the surface of the bucket facing the object, when a portion of an object being piled up is grasped and lifted by the bucket of the crane, The process includes a shape estimation step of estimating the shape of the object being held by the bucket based on the detection results, Shape estimation method, wherein the distance sensor is provided in a plurality of claws of the bucket for grasping the object, and each of the plurality of claws is rotatably mounted relative to the base of the bucket, with a plurality of distance sensors provided on the surface of the claw facing the object, from the side connected to the base to the tip.

10. A shape estimation program for causing a computer to function as a shape estimation device according to claim 1, wherein the computer functions as the detection result acquisition unit and the shape estimation unit.

Images