JP2020064588A - Shape information generation device, control device, unloading device, distribution system, program, and control method - Google Patents

Abstract

To meet a request that collapse of a cargo be prevented during unloading even when an overall shape of the cargo is unknown in an effort to apply a picking system to cargo loading onto a loading space of a truck and a container, and a cargo unloading from such a loading space.SOLUTION: A shape information generation device comprises: a first information acquisition unit for acquiring three-dimensional information on a first region of a surface of a plurality of piled articles acquired by imaging or scanning the plurality of articles from a first point; a second information acquisition unit for acquiring three-dimensional information on a second region of the surface of the plurality of articles acquired by imaging or scanning the plurality of articles from a second point; and a synthesis unit for generating information indicating a three-dimensional shape of at least a part of the surface of the plurality of articles based on the three-dimensional information on the first region acquired by the first information acquisition unit and the three-dimensional information on the second region acquired by the second information acquisition unit.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a shape information generation device, a control device, a loading / unloading device, a physical distribution system, a program, and a control method.

A picking system is known in which a picking hand and a robot arm are used to take out a work stored in a container and place the work at a specific position in another container.
[Prior Art Document]
[Patent Document]
[Patent Document 1] JP 2016-091053 A [Patent Document 2] JP 2016-147330 A [Patent Document 3] JP 2017-033429 A

  It is considered to apply the picking system to loading and unloading luggage on a loading platform such as a truck and a container. In this case, even if the overall shape of the package is unknown, it is required to take out the package while suppressing the collapse of the package.

  In a first aspect of the present invention, a shape information generation device is provided. The above-mentioned shape information generation device acquires, for example, first information acquisition for acquiring three-dimensional information of a first region on the surface of a plurality of articles obtained by imaging or scanning a plurality of stacked articles from a first point. Section. The above-described shape information generation device includes, for example, a second information acquisition unit that acquires three-dimensional information of the second region of the surface of the plurality of articles, which is obtained by imaging or scanning the plurality of articles from the second point. . The above-mentioned shape information generation device, for example, based on the three-dimensional information of the first region acquired by the first information acquisition unit, and the three-dimensional information of the second region acquired by the second information acquisition unit, a plurality of articles And a combining unit that generates information indicating a three-dimensional shape of at least a part of the surface of the. In the above-mentioned shape information generating device, the positions of the first point and the second point may be different from each other. The positions of the first point and the second point may be different from each other. The combining unit may complement one of the three-dimensional information of the first area and the three-dimensional information of the second area with the other to generate information indicating the three-dimensional shape of at least a part of the surfaces of the plurality of articles.

  In a second aspect of the invention, a controller is provided. The control device described above is, for example, a control device for controlling a robot. The robot described above uses, for example, a manipulator to take out at least one article from among a plurality of stacked articles. The above-mentioned control device is provided with the 1st information acquisition part which acquires three-dimensional information on the 1st field of the surface of a plurality of articles, for example, which was obtained by imaging or scanning a plurality of articles from the 1st point. The above-mentioned control device is provided with the 2nd information acquisition part which acquires three-dimensional information on the 2nd field of the surface of a plurality of articles, for example, which was obtained by imaging or scanning a plurality of articles from the 2nd point. The control device described above, for example, based on the three-dimensional information of the first region acquired by the first information acquisition unit, and the three-dimensional information of the second region acquired by the second information acquisition unit, the surface of the plurality of articles And a synthesis unit that generates information indicating the three-dimensional shape of at least part of In the above control device, the positions of the first point and the second point may be different from each other. In the above control device, the synthesizing unit complements one of the three-dimensional information of the first region and the three-dimensional information of the second region with the other to indicate the three-dimensional shape of at least a part of the surfaces of the plurality of articles. May be generated.

  In the above control device, the combining unit may remove the information about the surface of the manipulator from the three-dimensional information of the first area acquired by the first information acquisition unit. In the above control device, the combining unit may remove the information about the surface of the manipulator from the three-dimensional information of the second area acquired by the second information acquisition unit. In the control device described above, the synthesizing unit complements one of the three-dimensional information of the first region and the three-dimensional information of the second region, from which the information about the surface of the manipulator has been removed, with the other surface of the plurality of articles. The information indicating at least a part of the three-dimensional shape may be generated.

  In the above control device, the synthesizing unit (a) for each of the plurality of voxels virtually set in the workspace of the robot based on the three-dimensional information of the first region acquired by the first information acquiring unit. Occupancy information indicating that some of the surfaces of the plurality of articles are present in the voxel, (b) Blank information indicating that the surfaces of the plurality of articles are not present in the voxel, or (c) In the voxel The first map information associated with unknown information indicating that presence or absence of a part of the surfaces of the plurality of articles is unknown may be generated. In the above control device, the combining unit may (a) occupy information, (b) blank information, or, for each of the plurality of voxels, based on the three-dimensional information of the second area acquired by the second information acquiring unit, or (C) The second map information associated with the unknown information may be generated. In the above control device, the combining unit complements one of the first map information and the second map information with the other, and for each of the plurality of voxels, (a) occupation information, (b) blank information, or ( c) You may generate synthetic | combination map information with which unknown information was matched.

  In the control device described above, the combining unit may generate the first map information based on the three-dimensional information of the first area at a plurality of time points acquired by the first information acquisition unit. In the control device described above, the combining unit may generate the second map information based on the three-dimensional information of the second area at a plurality of time points acquired by the second information acquisition unit.

  In the control device described above, the combining unit may generate the first map information based on the three-dimensional information of the first region at one or more time points acquired by the first information acquisition unit. In the control device described above, the combining unit may generate the second map information based on the three-dimensional information of the second area at one or more time points acquired by the second information acquisition unit. In the above control device, when the first map information and the second map information are different for the same voxel, the combining unit (i) uses the information that is newer in time to generate the combined map information. Alternatively, (ii) the synthetic map information may be generated using the information closer to the installation position of the manipulator.

  The control device may include a take-out target determination unit that determines an article that is a target of a take-out operation by the manipulator. The control device may include a take-out direction determination unit that determines a direction in which the article determined by the take-out target determination unit is taken out.

  In a third aspect of the invention, an unloading device is provided. The loading / unloading device includes, for example, the control device. The loading / unloading device described above includes, for example, a manipulator.

  In a fourth aspect of the present invention, a physical distribution system is provided. The distribution system described above includes, for example, the control device described above. The physical distribution system described above includes, for example, a manipulator. The physical distribution system described above includes, for example, a carrying device that carries at least one of the plurality of articles.

  In a fifth aspect of the present invention, a control method is provided. The above control method is, for example, a control method for controlling the robot. The robot described above uses, for example, a manipulator to take out at least one article from among a plurality of stacked articles. The control method described above includes, for example, a first information acquisition step of acquiring three-dimensional information of the first regions on the surfaces of the plurality of articles, which are obtained by imaging or scanning the plurality of articles from the first point. The above-mentioned control method has a 2nd information acquisition step which acquires three-dimensional information on the 2nd field of the surface of a plurality of articles, for example, which was obtained by imaging or scanning a plurality of articles from the 2nd point. The control method described above is based on, for example, the three-dimensional information of the first area acquired in the first information acquisition step and the three-dimensional information of the second area acquired in the second information acquisition step. A synthesis step for generating information indicative of the three-dimensional shape of at least a part of the surface of the. In the above control method, the positions of the first point and the second point may be different from each other. In the above control method, in the synthesizing step, one of the three-dimensional information of the first area and the three-dimensional information of the second area is complemented by the other, and information indicating the three-dimensional shape of at least a part of the surfaces of the plurality of articles is obtained. May be included.

  In a sixth aspect of the present invention, a program is provided. A non-transitory computer-readable medium storing the above program may be provided. The above program may be a program for causing a computer to function as the above shape information generation device. The above program may be a program for causing a computer to function as the above control device. The above program may be a program for causing a computer to execute the above control method.

  Note that the above summary of the invention does not enumerate all the necessary features of the invention. Further, a sub-combination of these feature groups can also be an invention.

1 schematically shows an example of the system configuration of a physical distribution management system 100. 1 schematically shows an example of the internal configuration of the vanning system 120. 1 schematically shows an example of the internal configuration of the devanning system 140. 1 schematically shows an example of the system configuration of the loading / unloading system 142. 1 schematically shows an example of the system configuration of the loading / unloading system 142. 1 schematically shows an example of the internal configuration of the system control unit 210. An example of point cloud data is shown schematically. An example of point cloud data is shown schematically. An example of point cloud data is shown schematically. An example of information processing in the synthesis map generation unit 630 is schematically shown. 1 schematically shows an example of a data table 1100. 1 schematically shows an example of the internal configuration of the management server 160. 1 schematically shows an example of a data table 1300.

  Hereinafter, the present invention will be described through embodiments of the invention, but the following embodiments do not limit the invention according to the claims. In addition, not all of the combinations of features described in the embodiments are essential to the solving means of the invention. In the drawings, the same or similar parts may be denoted by the same reference numerals, and redundant description may be omitted.

[Outline of physical distribution management system 100]
FIG. 1 schematically shows an example of the system configuration of the physical distribution management system 100. In the present embodiment, the physical distribution management system 100 includes a vanning system 120, a devanning system 140, and a management server 160. In this embodiment, the vanning system 120 has an unloading system 122. In the present embodiment, the devanning system 140 has an unloading system 142.

  The physical distribution management system 100 may be an example of a physical distribution system. The devanning system 140 may be an example of a physical distribution system. The unloading system 122 may be an example of an unloading device and robot. The unloading system 142 may be an example of an unloading device and robot. As used herein, the term “one or more” means “one or more”.

  In this embodiment, the vanning system 120 utilizes the unloading system 122 to load the truck 12 with one or more loads. The vanning system 120 may obtain, for each of the one or more packages loaded on the truck 12, information regarding the size of the package. The vanning system 120 may acquire, for each of the one or more packages loaded on the truck 12, information regarding feature points for recognizing the package. The vanning system 120 may obtain, for each of the one or more packages loaded on the truck 12, information regarding the mass of the package.

  The method by which the vanning system 120 acquires the above-mentioned information regarding each package is not particularly limited. In one embodiment, the above information is input to a terminal of a packager of a package or a terminal of a delivery company of a package and is transmitted from the terminal to the vanning system 120 or the management server 160. In another embodiment, the above information is obtained by a sensor located on the vanning system 120. The sensor may be a sensor for acquiring an image of a package or a point cloud on the surface of a package. The sensor described above may be a sensor for obtaining the load mass.

  The vanning system 120 transmits the above information to the management server 160 via the communication network 10, for example. Details of the vanning system 120 will be described later.

  In this embodiment, the devanning system 140 utilizes the unloading system 142 to remove one or more loads from the truck 12. The devanning system 140, for example, accesses the management server 160 via the communication network 10 and acquires from the management server 160 information regarding the size of each of the one or more packages loaded on the truck 12. The devanning system 140 may acquire, from the management server 160, information regarding each feature point of one or more packages loaded on the truck 12. The devanning system 140 may obtain, from the management server 160, information regarding the mass of each of the one or more loads loaded on the truck 12.

  The devanning system 140 may take out one or more packages from the truck 12 by using various information acquired from the management server 160. Note that the devanning system 140 may acquire the above-described various types of information from the vanning system 120, the terminal of the cargo shipper, or the terminal of the parcel delivery company. Details of the devanning system 140 will be described later.

  In the present embodiment, the management server 160 sends and receives information to and from the vanning system 120 and the devanning system 140. For example, the management server 160 acquires information regarding the load of the truck 12 from the vanning system 120. The management server 160 may store information regarding the load of the truck 12 in the storage device. The management server 160 may transmit information regarding the load of the truck 12 to the devanning system 140 in response to a request from the devanning system 140, for example. Details of the management server 160 will be described later.

  In the present embodiment, the communication network 10 may be a wired communication transmission path, a wireless communication transmission path, or a combination of a wireless communication transmission path and a wired communication transmission path. . The communication network 10 may include a wireless packet communication network, the Internet, a P2P network, a dedicated line, a VPN, a power line communication line, or the like. The communication network 10 may include (i) a mobile communication network such as a mobile phone network, and (ii) wireless MAN, wireless LAN, Bluetooth (registered trademark), Zigbee (registered trademark), NFC (Near Field Communication). It may include a wireless communication network such as. The wireless MAN may be WiMAX®. The wireless LAN may be WiFi (registered trademark).

  In the present embodiment, as an example, the loading / unloading system 122 of the vanning system 120 loads the cargo on the truck 12, and the loading / unloading system 142 of the devanning system 140 takes the cargo from the truck 12, as an example. Will be described in detail. However, the unloading system 122 and the unloading system 142 are not limited to this embodiment. In other embodiments, the unloading system 122 and / or the unloading system 122 may perform both loading and unloading operations for the truck 12. Is configured as follows.

[Specific configuration of each part of the distribution management system 100]
Each unit of the physical distribution management system 100 may be implemented by hardware, software, or hardware and software. At least a part of each unit of the physical distribution management system 100 may be realized by a single server or may be realized by a plurality of servers. At least a part of each unit of the physical distribution management system 100 may be realized on a virtual server or a cloud system. At least a part of each unit of the physical distribution management system 100 may be realized by a personal computer or a mobile terminal. Examples of the mobile terminal include a mobile phone, a smartphone, a PDA, a tablet, a notebook computer or a laptop computer, a wearable computer, and the like. Each unit of the physical distribution management system 100 may store information using a distributed ledger technology such as a block chain or a distributed network.

  When at least a part of the constituent elements of the physical distribution management system 100 is realized by software, the constituent elements realized by the software define the operation related to the constituent elements in an information processing device having a general configuration. It may be realized by activating the program. The information processing apparatus having the above-described general configuration includes (i) a data processing apparatus having a processor such as a CPU and a GPU, a ROM, a RAM, and a communication interface, and (ii) a keyboard, a pointing device, a touch panel, a camera, and voice input. Device, gesture input device, various sensors, input device such as GPS receiver, (iii) display device, voice output device, output device such as vibration device, and (iv) storage device such as memory, HDD, SSD (external (Including a storage device). In the information processing device having the above general configuration, the data processing device or the storage device may store the program. The above program causes the information processing apparatus to execute the operation defined by the program when executed by the processor. The above program may be stored in a non-transitory computer-readable recording medium. The above program may be stored in a computer-readable medium such as a CD-ROM, a DVD-ROM, a memory, or a hard disk, or may be stored in a storage device connected to a network.

  The above program may be a program for causing a computer to function as the physical distribution management system 100 or a part thereof. The above program may include a module that defines the operation of each unit of the physical distribution management system 100. These programs or modules work on a data processing device, an input device, an output device, a storage device, or the like to cause a computer to function as each unit of the physical distribution management system 100, or to make a computer perform an information processing method in each unit of the physical distribution management system 100. Let it run. The above program may be installed in a computer forming at least a part of the physical distribution management system 100 from a computer-readable medium or a storage device connected to a network. The computer may function as at least a part of each unit of the physical distribution management system 100 by executing the above program. The information processing described in the above program is a concrete example in which the software related to the program and the various hardware resources of the distribution management system 100 or a part thereof cooperate with each other when the program is read into a computer. Function as an intelligent means. Then, the above-mentioned specific means realizes the calculation or processing of information according to the purpose of use of the computer in this embodiment, whereby the physical distribution management system 100 according to the purpose of use is constructed.

  The above program may be a program for causing a computer to execute the information processing method in the physical distribution management system 100 or a part thereof. In one embodiment, the information processing method in the physical distribution management system 100 or a part thereof may be a method of generating information indicating the three-dimensional shape of an article.

  The method of generating the above information is, for example, first information for acquiring three-dimensional information of the first region of the surface of the plurality of articles obtained by imaging or scanning the plurality of stacked articles from the first point. Has an acquisition stage. The method of generating the above information includes, for example, a second information acquisition step of acquiring three-dimensional information of the second region of the surface of the plurality of articles, which is obtained by imaging or scanning the plurality of articles from the second point. Have. The method of generating the above information is based on, for example, the three-dimensional information of the first area acquired in the first information acquisition step and the three-dimensional information of the second area acquired in the second information acquisition step, There is a compositing step for generating information indicative of a three-dimensional shape of at least a part of the surfaces of the plurality of articles. In the above method of generating information, the positions of the first point and the second point may be different from each other. In the method for generating the above information, the combining step complements one of the three-dimensional information of the first area and the three-dimensional information of the second area with the other to form the three-dimensional shape of at least a part of the surfaces of the plurality of articles. May be generated.

  In another embodiment, the information processing method in the distribution management system 100 or a part thereof may be a control method for controlling a robot. The robot described above uses, for example, a manipulator to take out at least one article from among a plurality of stacked articles.

  The control method described above includes, for example, a first information acquisition step of acquiring three-dimensional information of the first regions on the surfaces of the plurality of articles, which are obtained by imaging or scanning the plurality of articles from the first point. The above-mentioned control method has a 2nd information acquisition step which acquires three-dimensional information on the 2nd field of the surface of a plurality of articles, for example, which was obtained by imaging or scanning a plurality of articles from the 2nd point. The control method described above is based on, for example, the three-dimensional information of the first area acquired in the first information acquisition step and the three-dimensional information of the second area acquired in the second information acquisition step. A synthesis step for generating information indicative of the three-dimensional shape of at least a part of the surface of the. In the above control method, the positions of the first point and the second point may be different from each other. In the above control method, in the synthesizing step, one of the three-dimensional information of the first area and the three-dimensional information of the second area is complemented by the other, and information indicating the three-dimensional shape of at least a part of the surfaces of the plurality of articles is obtained. May be included.

  FIG. 2 schematically shows an example of the internal configuration of the vanning system 120. In the present embodiment, the vanning system 120 includes a loading / unloading system 122 and a system control unit 210 that controls the loading / unloading system 122. The vanning system 120 may include a conveyor 270 that carries the load 20. The vanning system 120 may include a three-dimensional camera 280 that images or scans the parcel 20 carried by the conveyor 270.

  In the present embodiment, the loading / unloading system 122 includes a manipulator 220, a pedestal 230, a rail 232, a drive unit 240, and a drive control unit 250. In the present embodiment, the manipulator 220 has a robot hand 222 and a robot arm 224.

  In this embodiment, the unloading system 122 uses the manipulator 220 to load the container 22 of the truck 12 with at least one load 20. The luggage 20 may be an example of an article.

  For example, the loading / unloading system 122 first identifies the shape and current position of the package 20 conveyed by the conveyor 270. The loading / unloading system 122 acquires, for example, an image captured by the three-dimensional camera 280, and identifies the shape and current position of the luggage 20 based on the image. The three-dimensional camera 280 may acquire a distance image of the luggage 20 or point cloud data.

  The unloading system 122 then determines the placement of the load 20 inside the container 22 based on the shape of the load 20. As the arrangement of the luggage 20, the position and posture of the luggage 20 are exemplified. The loading / unloading system 122 may acquire, for example, from the management server 160, information regarding the arrangement of the packages 20 inside the container 22. The unloading system 122 then picks up the load 20 from the conveyor 270 and places it at a specific location inside the container 22.

  The unloading system 122 may include a single manipulator 220 or multiple manipulators 220. The unloading system 122 may use a single manipulator 220 to load a single load 20 or multiple manipulators 220 to load a single load 20. The unloading system 122 may use a single manipulator 220 to load multiple loads 20 at a time.

  In the present embodiment, the robot hand 222 of the manipulator 220 holds the luggage 20. The gripping method of the robot hand 222 is not particularly limited. Examples of the gripping method of the robot hand 222 include a method of adsorbing the load 20 with a low pressure, a method of gripping the load 20 with a plurality of finger modules, and the like. The robot hand 222 may be replaced depending on the work content. The robot hand 222 may have various sensors such as a contact sensor, a mass measurement sensor, and a load measurement sensor. The robot hand 222 may have a camera. The camera may include a three-dimensional camera that acquires range images or point cloud data. The three-dimensional camera may have a plurality of image pickup devices. The three-dimensional camera may have a laser irradiator for distance measurement.

  In the present embodiment, the robot arm 224 of the manipulator 220 is connected to the robot hand 222 and adjusts the position and posture of the robot hand 222. The structure of the robot arm 224 is not particularly limited, but, for example, a multi-joint robot arm having a plurality of joints is used. The robot arm 224 may have various sensors such as a mass measurement sensor and a load measurement sensor. The load measurement sensor may be a sensor that measures the torque, current value, or voltage value of the drive unit 240 that drives each joint of the robot arm 224.

  In the present embodiment, the pedestal 230 supports the manipulator 220. The pedestal 230 may be a fixed pedestal or a movable pedestal. In the present embodiment, the pedestal 230 can move on the rail 232 along the extending direction of the rail 232. As a result, the robot hand 222 can reach the inside of the container 22. The pedestal 230 may house at least a part of the driving unit 240 therein. The pedestal 230 may house at least a part of the drive control unit 250 therein.

  In the present embodiment, the drive unit 240 drives the manipulator 220. In one embodiment, the driving unit 240 drives the robot hand 222. For example, the drive unit 240 supplies air to the robot hand 222 or sucks air from the robot hand 222. The drive unit 240 may be a pump. The driver 240 may have one or more pumps. In another embodiment, the driving unit 240 drives the robot arm 224. For example, the drive unit 240 adjusts the angle of each joint of the robot arm 224. The driving unit 240 may be a motor. The drive unit 240 may include one or more motors.

  In the present embodiment, the drive unit 240 drives the pedestal 230. For example, the drive unit 240 adjusts the relative position between the container 22 and the pedestal 230. The driving unit 240 may be a motor. The drive unit 240 may include one or more motors.

  The drive control unit 250 controls the drive unit 240. For example, the drive controller 250 receives a command from the system controller 210 to control the operation of the unloading system 122. The drive control unit 250 controls the drive unit 240 based on a command from the system control unit 210.

  The command for controlling the operation of the unloading system 122 may be information indicating the trajectory of the robot hand 222. The information indicating the trajectory of the robot hand 222 includes (i) initial placement, passing placement and final placement of the robot hand 222, and information indicating movement time, (ii) robot hand 222 in initial placement, passing placement and final placement. Information indicating the angle of each joint and the moving time, (iii) information indicating the placement of the robot hand 222 at each time during the moving time, and (iv) information indicating each joint of the robot hand 222 at each time during the moving time. Information indicating an angle can be exemplified. The arrangement of the robot hand 222 is specified by the position and posture of the robot hand 222, for example.

  FIG. 3 schematically shows an example of the internal configuration of the devanning system 140. In the present embodiment, the devanning system 140 includes a loading / unloading system 142 and a system control unit 210 that controls the loading / unloading system 142. The devanning system 140 may include a conveyor 270 that carries the load 20. The devanning system 140 may include a three-dimensional camera 280 that images or scans the parcel 20 carried by the conveyor 270.

  In the present embodiment, the loading / unloading system 142 includes a manipulator 220, a pedestal 230, a rail 232, a drive unit 240, and a drive control unit 250. In the present embodiment, the manipulator 220 has a robot hand 222 and a robot arm 224.

  The system control unit 210 may be an example of a shape recognition device and a control measure. The drive control unit 250 may be an example of a shape recognition device and a control measure.

  In the present embodiment, the unloading system 142 removes at least one load 20 from among the plurality of loads 20 stacked inside the container 22 of the truck 12. The unloading system 142 places the loads 20 unloaded from the container 22 on the conveyor 270, for example.

  The unloading system 142 may include a single manipulator 220 or multiple manipulators 220. The unloading system 142 may use a single manipulator 220 to remove a single load 20 and may use multiple manipulators 220 to remove a single load 20. The unloading system 142 may use a single manipulator 220 to remove multiple loads 20 at a time.

  Each part of the unloading system 142 may have the same configuration as each part of the unloading system 122. Details of the unloading system 142 will be described later. In addition, each part of the loading / unloading system 122 may have the same configuration as each part of the loading / unloading system 142 as long as there is no technical contradiction.

  In this embodiment, the system control unit 210 controls the loading / unloading system 142. Details of the system control unit 210 will be described later. The system control unit 210 of the loading / unloading system 122 may have the same configuration as the system control unit 210 of the loading / unloading system 142 as long as there is no technical contradiction.

  An example of the system configuration of the loading / unloading system 142 will be described with reference to FIGS. 4 and 5. FIG. 4 schematically illustrates an example of a side view of the unloading system 142. FIG. 5 schematically illustrates a top view of the unloading system 142.

  According to the embodiments described in connection with FIGS. 4 and 5, the unloading system 142 includes a 3D camera 260, a 3D camera 462, a 3D camera 464, a 3D camera 466 and a 3D camera 468. Prepare In this embodiment, the three-dimensional camera 462, the three-dimensional camera 464, the three-dimensional camera 466, and the three-dimensional camera 468 are arranged at different positions on the mount 430, respectively.

  One of the three-dimensional camera 462, the three-dimensional camera 464, the three-dimensional camera 466, and the three-dimensional camera 468 may be an example of the first information acquisition unit. The other one of the three-dimensional camera 462, the three-dimensional camera 464, the three-dimensional camera 466, and the three-dimensional camera 468 may be an example of the second information acquisition unit. The position of each camera on the gantry 430 may be an example of the first point or the second point.

  The three-dimensional camera 462 images or scans the inside of the container 22 from the installation position of the camera. Thereby, three-dimensional information about a partial area of the surface of the plurality of loads 20 stacked inside the container 22 is acquired. Examples of the three-dimensional information include point cloud data and range images. The point cloud data may be a set of three-dimensional coordinate values of measurement points. The three-dimensional camera 462 may acquire three-dimensional information using a plurality of image pickup devices. The three-dimensional camera 462 may include a laser irradiator.

  Similarly, the three-dimensional camera 464 images or scans the inside of the container 22 from the installation position of the camera. The three-dimensional camera 466 images or scans the inside of the container 22 from the installation position of the camera. The three-dimensional camera 468 images or scans the inside of the container 22 from the installation position of the camera. Each of the three-dimensional camera 464, the three-dimensional camera 466, and the three-dimensional camera 468 may have the same configuration as the three-dimensional camera 462.

  FIG. 6 schematically shows an example of the internal configuration of the system control unit 210. In the present embodiment, the system control unit 210 includes a point cloud information acquisition unit 620, a combination map generation unit 630, an extraction work determination unit 640, a robot control unit 650, and a storage unit 660.

  The point cloud information acquisition unit 620 may be an example of the first information acquisition unit and the second information acquisition unit. The synthesis map generation unit 630 may be an example of a synthesis unit. The extraction work determination unit 640 may be an example of an extraction target determination unit and an extraction direction determination unit.

  In the present embodiment, the point cloud information acquisition unit 620 is information obtained by the 3D camera 462, the 3D camera 464, the 3D camera 466, and the 3D camera 468 imaging or scanning the inside of the container 22. To get. For example, the point cloud information acquisition unit 620 acquires the point cloud data inside the container 22 acquired by each of the three-dimensional camera 462, the three-dimensional camera 464, the three-dimensional camera 466, and the three-dimensional camera 468.

  Accordingly, the point cloud information acquisition unit 620 can acquire a plurality of sets of point cloud data regarding the surfaces of the plurality of loads 20 stacked inside the container 22. Each of the plurality of sets of point cloud data is obtained by imaging or scanning a part of the surface of the plurality of loads 20 from different points.

  In a small work picking device, a work contained in a container is mainly imaged or scanned from above the work. Further, the robot hand of the picking device approaches the work from above the work. On the other hand, according to the present embodiment, the plurality of loads 20 stacked in the vertical direction are imaged or scanned mainly from the side surface side of the loads 20. The robot hand 222 approaches the luggage 20 from the side of the luggage 20.

  In the present embodiment, the composite map generation unit 630 is a single unit that indicates the internal situation of the container 22 based on a plurality of sets of point cloud data that is acquired by the point cloud information acquisition unit 620 and that indicates the internal situation of the container 22. Point cloud data (sometimes referred to as composite point cloud data) is generated. The synthetic map generation unit 630 may generate the synthetic point cloud data by complementing the point cloud data of one set from the plural sets of point cloud data with the point cloud data of the other set. The synthetic map generation unit 630 removes (i) point cloud data relating to the surface of the manipulator 220 from each of the plural sets of point cloud data, and (ii) the synthetic point cloud based on the plural sets of point cloud data. Data may be generated.

  According to the combined point cloud data, the three-dimensional shape of at least a part of the surfaces of the plurality of loads 20 inside the container 22 is shown. For example, the composite point group data includes (i) a three-dimensional shape of a part of the upper surfaces of the plurality of loads 20 and (ii) a three-dimensional shape of a side surface of the plurality of loads 20 that faces the outlet side. Indicates.

  The composite map generation unit 630 determines the surface inside the container 22 based on (i) a plurality of sets of point cloud data or (ii) composite point cloud data generated by using the plurality of sets of point cloud data. A three-dimensional map showing a part of the shape or a part of the surface shape of the plurality of loads 20 may be generated. For example, the synthesis map generation unit 630 associates (a) occupation information, (b) blank information, or (c) unknown information with each of a plurality of voxels virtually set inside the container 22. Generate map information (sometimes referred to as a synthetic depth map).

  (A) The occupancy information may be information indicating that a point group corresponding to a part of the surface of the luggage 20 exists inside the voxel. This indicates that a part of the surface of the luggage 20 exists at the position of the voxel. (B) The blank information may be information indicating that a point group corresponding to a part of the surface of the luggage 20 does not exist inside the voxel. The blank information may be information indicating that the point cloud does not exist inside the voxel. This indicates that the surface of the luggage 20 does not exist at the position of the voxel. (C) The unknown information may be information indicating that the presence or absence of the point cloud corresponding to a part of the surface of the luggage 20 is unknown inside the voxel. This indicates that the presence or absence of a part of the surface of the luggage 20 at the position of the voxel is unknown.

  The plurality of loads 20 stacked inside the container 22 include loads of different sizes. Therefore, due to the unevenness of the surfaces of the plurality of loads 20, there may be an area on the surfaces of the plurality of loads 20 where the point cloud data cannot be acquired by a single three-dimensional camera. Further, each of the three-dimensional cameras of the loading / unloading system 122 performs the work for taking out the container 22 arranged inside the container 22 not only when the manipulator 220 is located outside the container 22. While inside, the inside of the container 22 is imaged or scanned. Therefore, when the manipulator 220 is reflected in the angle of view of each 3D camera, there may be an area on the surface of the plurality of packages 20 where the point cloud data cannot be acquired by a single 3D camera.

  According to the present embodiment, the synthetic map generation unit 630 synthesizes a plurality of sets of point cloud data obtained by imaging or scanning from different points to obtain a three-dimensional image of at least a part of the surfaces of the plurality of packages 20. Composite point cloud data indicating the shape is generated. Thereby, for example, the system control unit 210 can control the loading / unloading system 142 based on the point cloud data regarding a wider area on the surface of the plurality of loads 20.

  In the present embodiment, for the purpose of simplifying the description, the synthetic map generation unit 630 determines, based on the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466, The details of the synthesis map generation unit 630 will be described by taking the case of generating the synthesis point cloud data as an example. However, the synthesis map generation unit 630 is not limited to this embodiment. In another embodiment, the synthetic map generation unit 630 may generate synthetic point cloud data based on a plurality of sets of point cloud data related to another combination of four 3D cameras. The synthetic map generation unit 630 may generate synthetic point cloud data based on three or more sets of point cloud data.

[Combining multiple point cloud data]
For example, the synthetic map generation unit 630 acquires the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466. In one embodiment, the synthesis map generation unit 630 complements one of the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466 with the other point cloud data to combine the data. Generate point cloud data.

  In another embodiment, the synthetic map generation unit 630 removes the point cloud data regarding the surface of the manipulator 220 from the point cloud data acquired by the three-dimensional camera 462. The synthetic map generation unit 630 removes the point cloud data regarding the surface of the manipulator 220 from the point cloud data acquired by the three-dimensional camera 466. In one embodiment, the synthetic map generation unit 630 removes, erases, or excludes specific point cloud data to remove the specific point cloud data. In another embodiment, the synthetic map generation unit 630 removes the point cloud data by rewriting the distance information included in the specific point cloud data to information indicating that the distance is unknown.

  Next, the synthesis map generation unit 630 removes (i) the point cloud data acquired by the three-dimensional camera 462 from which the point cloud related to the surface of the manipulator 220 has been removed, and (ii) the point cloud related to the surface of the manipulator 220. One of the point cloud data acquired by the three-dimensional camera 466 is complemented by the other to generate composite point cloud data. The complementing process is generated by the following procedure, for example.

  According to the present embodiment, the synthetic map generation unit 630, from the storage unit 660, for each of the three-dimensional camera 462 and the three-dimensional camera 466, an arbitrary position of the loading / unloading system 142 (may be referred to as a reference position). .) Is obtained in the three-dimensional space. The point cloud data acquired by the three-dimensional camera 462 indicates, for example, the angle and distance from the three-dimensional camera 462 regarding each point on the surface of the object existing inside the container 22. The point cloud data acquired by the three-dimensional camera 466 indicates, for example, the angle and distance from the three-dimensional camera 466 regarding each point on the surface of the object existing inside the container 22.

  The synthetic map generation unit 630 determines the three-dimensional shape based on the information indicating the relative positions of the three-dimensional camera 462 and the three-dimensional camera 466 and the point cloud data acquired by the three-dimensional camera 462 and the three-dimensional camera 466. The position of the point indicated by each point cloud data acquired by the camera 466 and the position of the point indicated by each point cloud data acquired by the three-dimensional camera 466 are calculated. The synthetic map generation unit 630 compares the position of the point indicated by each point cloud data acquired by the three-dimensional camera 466 with the position of the point indicated by each point cloud data acquired by the three-dimensional camera 466. Based on the comparison result, the synthetic map generation unit 630 removes (i) the point cloud related to the surface of the manipulator 220, the point cloud data acquired by the three-dimensional camera 462, and (ii) the point cloud related to the surface of the manipulator 220. By combining or merging with the point cloud data obtained by the three-dimensional camera 466 from which the points have been removed, the point cloud data showing points different from the points indicated by the point cloud data obtained by the three-dimensional camera 462. , Generate composite point cloud data.

[Generation of synthetic depth map]
For example, the synthetic map generation unit 630 acquires the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466. Based on the point cloud data acquired by the three-dimensional camera 462, the composite map generation unit 630, for each of the plurality of voxels virtually set inside the container 22, (a) occupation information, (b) blank information. Alternatively, (c) the first depth map associated with the unknown information may be generated. Based on the point cloud data acquired by the three-dimensional camera 466, the composite map generation unit 630, for each of a plurality of voxels virtually set inside the container 22, (a) occupancy information, (b) blank information. Alternatively, (c) the second depth map associated with the unknown information may be generated.

  The synthesis map generation unit 630 complements one of the first depth map and the second depth map with the other, and (a) occupancy information, (b) blank information, or (c) unknown for each of the plurality of voxels. A combined depth map may be generated with associated information. For example, the synthesis map generation unit 630 combines or merges the information of the first depth map and the information of the second depth map for each of the plurality of voxels virtually set inside the container 22, Generate a composite depth map.

  The interior of container 22 may be an example of a workspace for unloading system 142. The first depth map may be an example of the first map information. The second depth map may be an example of second map information. The combined depth map may be an example of combined map information.

  In another embodiment, the synthetic map generation unit 630 may generate the first depth map after removing the point cloud related to the surface of the manipulator 220 from the point cloud data acquired by the three-dimensional camera 462. The synthetic map generation unit 630 may generate the second depth map after removing the point cloud related to the surface of the manipulator 220 from the point cloud data acquired by the three-dimensional camera 466.

  In another embodiment, the synthetic map generation unit 630 may generate the first depth map based on the point cloud data acquired by the three-dimensional camera 462 at a plurality of time points. The synthetic map generation unit 630 may generate the second depth map based on the point cloud data obtained by the three-dimensional camera 466 at a plurality of time points.

  For example, when the point cloud related to the surface of the manipulator 220 is removed from the point cloud data acquired by the three-dimensional camera 462 at time t2, the number of point groups related to the surface of the luggage 20 may be extremely reduced. Therefore, the synthesis map generation unit 630 uses the point cloud data acquired by the three-dimensional camera 462 at time t1 that is a time before time t2, and the point cloud data acquired by the three-dimensional camera 462 at time t2. Complement. Specifically, for example, the synthetic map generation unit 630 outputs at least a part of the “unknown information” of the first depth map generated based on the point cloud data acquired by the three-dimensional camera 462 at time t2 at time t1. In the above, based on the point cloud data acquired by the three-dimensional camera 462, it is replaced with “occupancy information” or “blank information”.

  In another embodiment, the synthetic map generation unit 630 may generate the first depth map based on the point cloud data acquired by the three-dimensional camera 462 at one or more time points. The synthetic map generation unit 630 may generate the second depth map based on the point cloud data obtained by the three-dimensional camera 466 at one or more time points.

  When the composite map generation unit 630 generates a composite depth map based on the first depth map and the second depth map, the data of the first depth map and the data of the second depth map are the same for the same voxel. It can be different. In addition, regarding the same voxel, the data of the first depth map at the first time and the data of the first depth map at the second time may be different. Similarly, for the same voxel, the data of the second depth map at the third time and the data of the second depth map at the fourth time may be different.

  Therefore, according to one embodiment, the synthetic map generation unit 630 generates the synthetic depth map by using the information that is newer in time. In another embodiment, the synthetic map generation unit 630 may generate the synthetic depth map by using the information on the side closer to the installation position of the manipulator 220.

  In the present embodiment, the case has been described in which the composite depth map is generated based on the first depth map and the second depth map. However, the method of generating the composite depth map is not limited to this embodiment. In another embodiment, the synthetic map generation unit 630 acquires the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466. The synthetic map generation unit 630 generates synthetic point cloud data based on the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466. The synthetic map generation unit 630 may generate a synthetic depth map based on the synthetic point cloud data.

  In the present embodiment, the extraction work determination unit 640 determines the content of the work of the manipulator 220. For example, the take-out work determination unit 640 determines the luggage 20 that is the target of the take-out work by the manipulator 220. The take-out work determination unit 640 may determine the direction in which the parcel 20 targeted for the take-out work is taken out (sometimes referred to as the take-out direction). When the take-out work determination unit 640 cannot determine at least one of the baggage 20 and the take-out direction of the baggage 20 to be taken out, it sends information indicating that to the terminal of the administrator or the user of the distribution management system 100. You may

  In general, in the container 22, the closer to the installation position of the manipulator 220 (sometimes referred to as the outlet side of the container 22), the closer the parcel 20 is to the parcel 20 through the three-dimensional camera 260. A large amount of information can be obtained. In addition, the amount of information obtained via the three-dimensional camera 260 is larger as the luggage 20 is located above. Therefore, for example, the unloading work determination unit 640 selects, as the target of the unloading work, the load 20 located on the uppermost surface and closest to the outlet side of the container 22, and the selected load 20 is selected. 20 is taken out toward the take-out port side of the container 22. However, the determination of the extraction target and the determination of the extraction direction are not limited to this embodiment.

  In one embodiment, the unloading work determination unit 640 determines the luggage 20 to be the target of the unloading work, based on the positional relationship between each of the stacked packages 20. The unloading work determination unit 640 may determine the garage 20 to be the target of the unloading work, based on the vertical positional relationship among the stacked packages 20. For example, the unloading work determination unit 640 determines the garage 20 to be the target of the unloading work so that the higher the load 20 is, the more the target of the unloading work becomes. The unloading work determination unit 640 targets the unloading work so that the luggage 20 arranged below the center is preferentially targeted for the unloading work as compared with the luggage arranged in the middle position. You may decide the baggage 20 which becomes. The priority may be set so that the luggage 20 arranged on the uppermost surface has the highest priority, or the luggage 20 arranged at a position other than the uppermost surface may have the highest priority. .

  The unloading work determination unit 640 may determine the luggage 20 to be the target of the unloading work, based on the positional relationship in the depth direction of the container 22 among the plurality of stacked luggage 20. For example, the unloading work determination unit 640 determines the garage 20 to be the target of the unloading work so that the garage 20 disposed on the unloading side of the container 22 is more preferentially the target of the unloading work. It should be noted that the priority may be set so that the luggage 20 arranged closest to the outlet side of the container 22 is given the highest priority, and the luggage 20 arranged in a specific area of the container 22 is It may be set to have the highest priority.

  Next, the take-out work determination unit 640 determines the take-out direction of the parcel 20 targeted for the take-out work. The take-out direction of the luggage 20 is not particularly limited.

  The load 20 may be taken out vertically upward or may be one direction substantially parallel to the horizontal direction. The take-out direction of the luggage 20 may be inclined with respect to the vertical direction or the horizontal direction. The take-out direction of the luggage 20 may be substantially parallel to the depth direction of the container 22 (horizontal direction in FIG. 3), or in the width direction of the container 22 (direction perpendicular to the paper surface of FIG. 3). It may be substantially parallel. The take-out direction of the luggage 20 may be inclined with respect to the depth direction or the width direction of the container 22. The take-out direction of the luggage 20 may be substantially parallel to the longitudinal direction of the luggage 20 or substantially parallel to the lateral direction of the luggage 20. The take-out direction of the luggage 20 may be substantially parallel to the longitudinal direction or the lateral direction of the luggage 20.

  It should be noted that the extraction direction being “upward” includes not only the case where the extraction direction is upward in the vertical direction but also the case where the angle formed by the extraction direction and the vertical direction is less than 90 degrees. The angle formed by the extraction direction and the vertical direction may be 60 degrees or less, or 45 degrees or less. In addition, the take-out direction is "depth direction" or "width direction", or that the take-out direction is "depth direction" or "width direction" is not limited to the case where the take-out direction is parallel to the horizontal direction, It also includes the case where the angle between the take-out direction and the horizontal direction is less than 90 degrees. The angle formed by the extraction direction and the horizontal direction may be 60 degrees or less, or 45 degrees or less.

  When there are a plurality of candidates related to the take-out direction of the luggage 20, the take-out work determination unit 640 does not interfere with the luggage 20 and the manipulator 220, which are the objects of the take-out work, and the wall of the other luggage 20 or the container 22, It may be determined whether or not there is a route through which the package 20 can be taken out (sometimes referred to as a take-out route). The take-out work determination unit 640 may calculate an index indicating the certainty of the determination result regarding the presence or absence of the take-out route. The take-out work determination unit 640 performs the take-out work when the package 20 is taken out along each take-out route. You may estimate the time required.

  The unloading work determination unit 640 may determine, as the unloading direction, a candidate that can establish the unloading route among the plurality of candidates related to the unloading direction of the luggage 20. When the number of candidates for which the extraction route can be established is two or more, the extraction work determination unit 640 selects a candidate corresponding to a more reliable extraction route based on the index indicating the certainty of the above determination result. The extraction direction may be determined. The take-out work determination unit 640 may establish the take-out route among the plurality of candidates related to the take-out direction of the package 20, and may determine the candidate having the shortest time for the take-out work as the take-out direction.

  The take-out work determination unit 640 determines the take-out direction of the packages 20 based on, for example, the positional relationship between the stacked packages 20. The unloading work determination unit 640 may determine the unloading direction of the packages 20 based on the vertical positional relationship among the stacked packages 20. For example, when the luggage 20 arranged on the uppermost surface is taken out, the taking-out work determination unit 640 decides to take out the luggage 20 so that the luggage 20 is pulled out upward. When the baggage 20 arranged at a position lower than the center is taken out, the take-out work determination unit 640 causes the baggage 20 to be pulled out toward the manipulator 220 side or the baggage 20 in the width direction of the container 22. It may be decided to remove the load 20 so that it is pulled along.

  The unloading work determination unit 640 may determine the unloading direction of the packages 20 based on the positional relationship in the depth direction of the containers 22 in the plurality of stacked packages 20. For example, when the luggage 20 arranged closest to the outlet side of the container 22 is taken out, the take-out work determination unit 640 determines that the luggage 20 is taken out so that the luggage 20 is pulled out toward the take-out side. decide. When another package 20 exists between the package 20 that is the target of the package extraction work and the outlet side of the container 22, the package extraction work determination unit 640 causes the package 20 to be pulled out upward, or It is decided to remove the load 20 so that the load 20 can be pulled out along the width direction of the container 22.

  In the present embodiment, the unloading work is taken as an example, in which at least one of the unloading work 20 and the unloading direction of the garage 20 is determined based on the positional relationship of each of the stacked plurality of loads 20. The details of the decision unit 640 have been described. However, the extraction work determination unit 640 is not limited to this embodiment.

  In another embodiment, the unloading work determination unit 640 may determine at least one of the luggage 20 to be the target of the unloading operation and the unloading direction of the luggage 20 based on the size of the luggage 20. For example, the taking-out work determination unit 640 takes out the taking-out work so that the smaller the size of the parcels 20 located on the uppermost surface of each of the plurality of rows or columns, the parcel 20 having a smaller size is the target of the taking-out work. The luggage 20 to be the target of work is determined. The unloading work determination unit 640 may preferentially target the unloading work for the stuff 20 having a size in a specific range among the plurality of stuffs 20 located on the uppermost surface of each of the plurality of rows or columns. Alternatively, the baggage 20 to be taken out may be determined.

  The take-out work determination unit 640 selects the one piece of luggage 20 as an object to be taken out based on the size of the one piece of luggage 20 and the size of the other piece of luggage 20 located on the exit side of the one piece of luggage 20. You may decide whether or not. The other luggage 20 may be the luggage 20 arranged in the same row as the one luggage 20 or may be the luggage 20 arranged in a row different from the one luggage 20. Good.

  For example, when a large-sized baggage 20 exists near the outlet of the container 22, when the manipulator 220 takes out another baggage 20 or the three-dimensional camera 260 images or scans the inside of the container 22, It is possible that the large baggage 20 interferes with the above work. Therefore, when the size of one piece of luggage 20 is larger than the size of another piece of luggage 20, the take-out work determination unit 640 decides not to select the one piece of luggage 20 as the take-out target. As a result, the larger-sized luggage 20 is taken out first. As a result, the time required to take out the plurality of loads 20 loaded in the container 22 is shortened. In addition, the blind spot of the three-dimensional camera 260 caused by the luggage 20 disappears, and the content of the composite point cloud data or the composite depth map is further enhanced.

  In another embodiment, the take-out work determination unit 640 acquires information regarding the size of each of the plurality of packages 20 loaded in the container 22 from the management server 160. The take-out work determination unit 640 takes out the baggage 20 and the baggage 20 targeted for the take-out work based on the information regarding the sizes of the plurality of packages 20 and the combined point cloud data or the combined depth map of the plurality of packages 20. At least one of the directions may be determined.

  For example, the take-out work determination unit 640 analyzes at least part of the boundary between one package 20 and another package 20 by analyzing the composite point group data or the composite depth map of the plurality of packages 20. The take-out work determination unit 640 estimates the length of at least one side of the one package 20 based on the detected boundary information. The take-out work determination unit 640 estimates the size of one package 20 based on the information on the estimated length of at least one side and the information on the size of each of the plurality of packages 20.

  Thereby, even when the three-dimensional camera 260 cannot image or scan a part of one baggage 20, the take-out work determination unit 640 uses the information regarding the size of each of the plurality of bags 20 to reduce the It is possible to narrow down the candidates for the size of the luggage 20. As a result, the take-out work determination unit 640 overlaps the one package 20 with the other package 20 based on the information indicating the estimated value of the size of the one package 20 and the combined point cloud data or the combined depth map. The condition can be estimated.

  The above-mentioned degree of overlap may be indicated by a statistical value of values narrowed down as candidates for the size of one package 20. Examples of statistical values include maximum value, minimum value, average value, median value, mode value, and the like. For example, the above-described degree of overlap is indicated by at least one of the minimum value and the maximum value of the estimated value of the length of the overlapping portion between one package 20 and another package 20.

  For example, the take-out work determination unit 640 estimates the degree of overlap of each of the one or more packages 20 recognized based on the information acquired by the three-dimensional camera 260 with the other packages 20 existing above. The unloading work determination unit 640 may determine the baggage 20 to be the target of the unloading work, based on the degree of overlap with other packages 20. For example, the unloading work determination unit 640 determines the garage 20 to be the target of the unloading work so that the garage 20 having a smaller degree of overlap with the other garage 20 is the target of the unloading work more preferentially.

  When the baggage B is stacked on the baggage A, and the degree of overlap between the baggage A and the baggage B is lighter than a predetermined degree, the take-out work determination unit 640 causes the baggage A to be loaded. It may be decided to take out the luggage B before the luggage B, and it may be decided to take out the luggage B before the luggage A. On the other hand, when the degree of overlap between the baggage A and the baggage B is more severe than a predetermined degree, the take-out work determination unit 640 determines that the baggage B is taken out before the baggage A.

  The take-out work determination unit 640 may determine the take-out direction of the package 20 based on information indicating the degree of overlap between the one package 20 and another package 20 existing above the one package 20. When the above-mentioned degree of overlap is lighter than a predetermined degree, the unloading work determination unit 640 may determine an arbitrary direction as the unloading direction of one piece of luggage 20. On the other hand, when the above-mentioned degree of overlap is more severe than a predetermined degree, the unloading work determination unit 640 may determine a direction other than the upward direction as the unloading direction of the one package 20.

  The extraction work determination unit 640 may determine the allowable range or the allowable value of the angle formed by the extraction direction and the vertical direction based on the above-described degree of overlap. The take-out work determination unit 640 may determine whether or not there is a take-out route when the one package 20 is pulled out in a specific direction, based on the information indicating the degree of overlap.

  In another embodiment, the take-out work determination unit 640 evaluates the shadow at the boundary between the one package 20 and the other package 20. The extraction work determination unit 640 may evaluate the shadow based on at least one of the shape of the shadow, the size of the shadow, and the density of the shadow. The take-out work determination unit 640 may determine at least one of the luggage 20 to be the target of the take-out work and the take-out direction of the luggage 20 based on the evaluation regarding the shadow.

  For example, the take-out work determination unit 640 analyzes the image captured by the three-dimensional camera 260 and recognizes a shadow formed at the boundary or gap between the one package 20 and another package 20. The shape, size, and darkness of the above shadows depend on the position and number of light sources, the size of one package 20, the size of another package 20, and the positional relationship between one package 20 and another package 20. It changes accordingly.

  Therefore, the take-out work determination unit 640 estimates the length of at least one side of the one package 20 or the other package 20 based on at least one of the shadow shape, the shadow size, and the shadow density. Good. The take-out work determination unit 640 may estimate the size of one package 20 based on the information on the estimated length of at least one side and the information on the size of each of the plurality of packages 20. Further, the take-out work determination unit 640 may estimate the degree of overlap between the one package 20 and the other package 20 based on at least one of the shadow shape, the shadow size, and the shadow density. . The unloading work determination unit 640 may determine at least one of the load 20 to be the target of the unloading work and the unloading direction of the load 20 by a procedure similar to the procedure described above.

  As described above, the consideration factor for determining at least one of the luggage 20 to be the target of the unloading operation and the unloading direction of the luggage 20 includes (i) the positional relationship between the stacked plurality of luggage 20, and (ii) Each size of the plurality of packages 20, (iii) analysis result of information obtained by capturing or scanning the inside of the container 22 by the three-dimensional camera 260 (for example, synthetic point cloud data or synthetic depth map), ( iv) Analysis result (for example, shadow evaluation result) of the image obtained by the inside of the container 22 captured by the three-dimensional camera 260, (v) Overlapping degree of one package 20 and another package 20 , (Vi) the certainty of the determination result regarding the presence / absence of the extraction route, and (vii) the time required for the extraction work. The unloading work determination unit 640 may determine at least one of the load 20 to be the target of the unloading work and the unloading direction of the load 20 based on a combination of two or more consideration elements.

  In the present embodiment, the robot controller 650 controls the manipulator 220. For example, the robot control unit 650 controls the manipulator 220 to take out the parcel 20 determined by the take-out work determination unit 640 from the plurality of parcels 20 and arrange the parcel 20 on the conveyor 270. The robot controller 650 may determine the trajectory of the robot hand 222 for performing the above work. The robot controller 650 may generate a command indicating the trajectory of the robot hand 222 and send the command to the drive controller 250.

  In the present embodiment, the storage unit 660 stores various kinds of information. For example, the storage unit 660 may acquire information regarding the size of each of the plurality of packages 20 loaded in the container 22 from the management server 160 and store the information. The storage unit 660 may acquire information regarding each feature point of the plurality of packages 20 loaded in the container 22 from the management server 160 and store the information.

  The point cloud data will be described with reference to FIGS. 7, 8 and 9. 7 and 8 schematically show an example of point cloud data at a position of a specific height. FIG. 7 shows an example in which the manipulator 220 is not within the angle of view of the camera. FIG. 7 shows an example in which the manipulator 220 is within the angle of view of the camera. FIG. 9 schematically shows an example of point cloud data at a plurality of time points.

  FIG. 7 shows the point cloud data 702 acquired by the three-dimensional camera 466, along with the luggage 72, the luggage 74, and the luggage 76 arranged inside the container 22. Similarly, FIG. 7 shows the point cloud data 704 acquired by the three-dimensional camera 462, along with the luggage 72, the luggage 74, and the luggage 76 arranged inside the container 22. In FIG. 7, small circles 710 indicate the points that make up the point cloud. As shown in FIG. 7, there is an area where a single 3D camera cannot acquire a point cloud due to the unevenness of a plurality of packages.

  FIG. 8 shows the point cloud data 802 acquired by the three-dimensional camera 466 together with the luggage 72, the luggage 74 and the luggage 76 arranged inside the container 22, and the manipulator 220. Similarly, FIG. 7 shows the point cloud data 804 acquired by the three-dimensional camera 462 together with the luggage 72, the luggage 74 and the luggage 76 arranged inside the container 22, and the manipulator 220. As shown in FIG. 8, since the manipulator 220 enters the angle of view of the three-dimensional camera, there is a region where a single three-dimensional camera cannot acquire a point cloud.

  FIG. 9 illustrates that the three-dimensional camera 466 acquires the point cloud data 702 and the three-dimensional camera 462 acquires the point cloud data 704 at time t1. Similarly, FIG. 9 shows that the three-dimensional camera 466 acquires the point cloud data 802 and the three-dimensional camera 462 acquires the point cloud data 804 at time t2. In FIG. 9, a thick line shows a point cloud. In FIG. 9, a region surrounded by a dotted line shows a region in which the presence or absence of a point group corresponding to a part of the surface of the luggage 20 is unknown.

  FIG. 10 schematically shows an example of information processing in the synthesis map generation unit 630. As described above, the synthetic map generation unit 630 acquires the point cloud data acquired by the three-dimensional camera 462 from the three-dimensional camera 462. Similarly, the synthetic map generation unit 630 acquires the point cloud data acquired by the three-dimensional camera 466 from the three-dimensional camera 466.

  The synthetic map generation unit 630 generates the synthetic depth map 1000 based on the point cloud data acquired by the three-dimensional camera 462 and the point cloud data acquired by the three-dimensional camera 466. The synthetic map generation unit 630 generates the synthetic depth map 1000 based on the point cloud data acquired by the three-dimensional camera 462 at one or more time points and the point cloud data acquired by the three-dimensional camera 466 at one or more time points. You may. In the combined depth map 1000, the surface shapes 1010 of the plurality of packages 20 are obtained by connecting the plurality of voxels in which the occupation information is stored.

  FIG. 11 schematically shows an example of the data table 1100. In the present embodiment, the data table 1100 may be an example of the first depth map, the second depth map, and the combined depth map. In the present embodiment, the data table 1100 stores a voxel ID 1120 for identifying each voxel, information 1140 regarding a point group associated with each voxel, and information 1160 indicating a synthesis result in association with each other.

  In the present embodiment, the information 1140 regarding the point group related to each voxel includes a camera ID for identifying the three-dimensional camera that acquired the point group, information indicating the time when the point group was acquired, and the point. Information indicating the evaluation of voxels defined based on the group may be included. Examples of the information indicating the voxel evaluation include occupancy information, blank information, and unknown information.

  The data table of the combined depth map may store the voxel ID 1120 and the information 1160 indicating the combined result in association with each other. The data tables of the first depth map and the second depth map may store the voxel ID 1120 and the point cloud information 1140 in association with each other.

  FIG. 12 schematically shows an example of the internal configuration of the management server 160. In this embodiment, the management server 160 includes a communication control unit 1210 and a package information management unit 1220.

  In the present embodiment, the communication control unit 1210 controls communication between the management server 160 and at least one of the vanning system 120 and the devanning system 140. The communication control unit 1210 may be a communication interface.

  In the present embodiment, the package information management unit 1220 manages information regarding each of the plurality of packages 20 loaded on the truck 12. For example, the luggage information management unit 1220 acquires information regarding the size of each of the plurality of luggage 20 loaded on the truck 12 from the vanning system 120. The baggage information management unit 1220 may acquire, from the vanning system 120, information regarding each characteristic point of the plurality of packages 20 loaded on the truck 12. The luggage information management unit 1220 may acquire information regarding the mass of each of the plurality of luggage 20 loaded on the truck 12 from the vanning system 120. The luggage information management unit 1220 may store various types of information regarding the cargo of the truck 12 in the storage device. The luggage information management unit 1220 may transmit information regarding the load of the truck 12 to the devanning system 140, for example, in response to a request from the devanning system 140.

  FIG. 13 schematically shows an example of the data table 1300. In the present embodiment, the data table 1300 may be an example of information regarding the load of the truck 12. In the present embodiment, the data table 1300 includes a container ID 1320 for identifying a container in which each package is loaded, a package ID 1330 for identifying each package, information 1340 regarding the size of each package, and a shape of each package. The information 1350 relating to each package and the information 1360 relating to the characteristic points of each package are stored in association with each other. The information 1360 regarding the feature points of each package may be a URI of electronic data indicating the feature points of each package.

  Although the present invention has been described using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It is apparent to those skilled in the art that various modifications and improvements can be added to the above-described embodiment. Further, the matters described in the specific embodiment can be applied to other embodiments within a technically consistent range. Further, each component may have the same characteristics as other components having the same name but different reference signs. It is apparent from the scope of the claims that the embodiments added with such modifications or improvements can be included in the technical scope of the present invention.

  The execution order of each process such as operation, procedure, step, and step in the device, system, program, and method shown in the claims, the specification, and the drawings is, in particular, “before” or “prior to”. It should be noted that the output of the previous process can be realized in any order unless the output of the previous process is used in the subsequent process. Even if the operation flow in the claims, the specification, and the drawings is described by using “first,” “next,” and the like for convenience, it means that it is essential to carry out in this order. Not a thing.

  10 communication network, 12 trucks, 20 packages, 22 containers, 72 packages, 74 packages, 76 packages, 100 logistics management system, 120 vanning system, 122 loading / unloading system, 140 devanning system, 142 loading / unloading system, 160 management server, 210 System control unit, 220 manipulator, 222 robot hand, 224 robot arm, 230 pedestal, 232 rail, 240 drive unit, 250 drive control unit, 260 3D camera, 270 conveyor, 280 3D camera, 430 mount, 462 3D camera 464 3D camera, 466 3D camera, 468 3D camera, 620 point cloud information acquisition unit, 630 composite map generation unit, 640 extraction work determination unit, 650 robot control unit, 60 storage unit, 702 point cloud data, 704 point cloud data, 710 circle, 802 point cloud data, 804 point cloud data, 1000 composite depth map, 1010 surface shape, 1100 data table, 1120 voxel ID, 1140 information, 1160 information, 1210 communication control unit, 1220 package information management unit, 1300 data table, 1320 container ID, 1330 package ID, 1340 information, 1350 information, 1360 information

Claims (14)

A first information acquisition unit for acquiring three-dimensional information of a first area on the surface of the plurality of articles, which is obtained by imaging or scanning the plurality of stacked articles from a first point;
A second information acquisition unit for acquiring three-dimensional information of a second region on the surface of the plurality of articles, which is obtained by imaging or scanning the plurality of articles from a second point;
At least one of the surfaces of the plurality of articles is based on the three-dimensional information of the first area acquired by the first information acquisition unit and the three-dimensional information of the second area acquired by the second information acquisition unit. A combining unit that generates information indicating the three-dimensional shape of the unit;
Equipped with
The positions of the first point and the second point are different from each other,
Shape information generation device. The synthesizing unit,
Based on the three-dimensional information of the first area acquired by the first information acquisition unit, for each of the plurality of voxels virtually set in the space in which the plurality of articles are arranged, (a) in the voxel Occupancy information indicating that a part of the surfaces of the plurality of articles exist, (b) blank information indicating that the surfaces of the plurality of articles do not exist in the voxel, or (c) in the voxel Generating the first map information, which is associated with unknown information indicating that presence or absence of a part of the surfaces of the plurality of articles is unknown,
Based on the three-dimensional information of the second area acquired by the second information acquisition unit, (a) the occupancy information, (b) the blank information, or (c) the unknown for each of the plurality of voxels. Generate the second map information to which the information is associated,
By combining the first map information and the second map information, (a) the occupation information, (b) the blank information, or (c) the unknown information is associated with each of the plurality of voxels. The combined map information is generated as information indicating the three-dimensional shape of at least a part of the surfaces of the plurality of articles,
The shape information generation device according to claim 1. The synthesizing unit generates the first map information based on the three-dimensional information of the first region at a plurality of time points acquired by the first information acquiring unit,
The shape information generation device according to claim 2. The synthesizing unit generates the second map information based on the three-dimensional information of the second region at a plurality of time points acquired by the second information acquiring unit,
The shape information generation device according to claim 2 or 3. The synthesizing unit,
Generating the first map information based on the three-dimensional information of the first region at one or more time points acquired by the first information acquisition unit,
Generating the second map information based on the three-dimensional information of the second region at one or more time points acquired by the second information acquisition unit,
When the first map information and the second map information are different for the same voxel,
(I) generating the composite map information by using the information which is newer in time, or
(Ii) generating the composite map information by using information closer to the first point or the second point,
The shape information generation device according to any one of claims 2 to 4. A control device for controlling a robot for taking out at least one article from a plurality of stacked articles using a manipulator,
The shape information generating device according to any one of claims 1 to 5 is provided.
Control device. The synthesizing unit,
From the three-dimensional information of the first area acquired by the first information acquisition unit, information regarding the surface of the manipulator is removed, and from the three-dimensional information of the second area acquired by the second information acquisition unit, the manipulator Removes information about the surface,
One of the three-dimensional information of the first region and the three-dimensional information of the second region, from which the information about the surface of the manipulator has been removed, is complemented by the other, and at least a part of the surfaces of the plurality of articles 3 Generates information indicating the dimensional shape,
The control device according to claim 6. The manipulator further comprises a take-out target determining unit that determines an article to be taken out by the take-out work,
The control device according to claim 6 or 7. Further comprising a take-out direction decision unit that decides a direction in which the article to be taken out is decided by the take-out target decision unit,
The control device according to claim 8. The control device according to any one of claims 6 to 9,
The manipulator,
Loading and unloading device. The control device according to any one of claims 6 to 9,
The manipulator,
A carrying device for carrying at least one of the plurality of articles;
A logistics system equipped with.   A program for causing a computer to function as the shape information generation device according to any one of claims 1 to 5.   A program for causing a computer to function as the control device according to any one of claims 6 to 9. A first information acquisition step of acquiring three-dimensional information of a first region of the surface of the plurality of articles, which is obtained by imaging or scanning the plurality of stacked articles from a first point;
A second information acquisition step of acquiring three-dimensional information of a second region on the surface of the plurality of articles, which is obtained by imaging or scanning the plurality of articles from a second point;
Based on the three-dimensional information of the first area acquired in the first information acquisition step and the three-dimensional information of the second area acquired in the second information acquisition step, A synthesizing step of generating information indicating at least a part of the three-dimensional shape;
Have
The positions of the first point and the second point are different from each other,
Control method.