JP7493876B2 - Automated Warehouse System - Google Patents

Description

The present invention relates to an automated warehouse system that can operate robots placed in a warehouse autonomously or remotely to pick up items.

In logistics warehouses, racks capable of accommodating a large number of containers, which are article-holding bodies for storing articles, are installed, and various articles are sorted into individual rack containers according to type, size, etc. In such logistics warehouses, various tasks (unpacking, repackaging, picking, inspection, packing, etc.) are performed. For example, in the picking task stage, the rack container is first transported by a stacker crane from its storage position on the rack to a work space provided at the entrance/exit of the rack where a worker is waiting. Next, in the work space, the worker takes out the specified article from the rack container that has been transported out, and transfers it to a sorting container, thus carrying out the picking task. In recent years, with the rapid spread of e-commerce, there has been active expansion and addition of logistics warehouses, and as a result, attempts are being made to automate the tasks performed in logistics warehouses.

To automate this picking task, there is an automatic picking means that includes an imaging means for imaging the inside of a rack container in which items are stored, and a robot arm that actually picks out the specified item by gripping, suction, or other means, and that allows the robot arm to pick out the item from the rack container based on the image captured by the imaging means (see, for example, Patent Document 1).

JP 2018-188308 A (page 6, Figure 1)

Although technology for picking items using automatic picking means is improving day by day through learning control and the like, it has not yet reached the level of human skill, and there are cases where the autonomous operation of the robot arm is unable to successfully pick up an item, or where the target object cannot be properly detected from an image captured by the imaging means. When such a situation occurs, in an automated warehouse system such as that described in Patent Document 1, the robot arm must make multiple attempts until it is able to pick up the item, or it must decide to give up on picking the item and stop the picking work. In either case, the incoming picking orders are interrupted, resulting in a problem that they cannot be completed quickly. Thus, under the current circumstances, it is difficult to rely entirely on autonomous operation to perform picking work.

The present invention was developed to address these problems, and aims to provide an automated warehouse system that can perform picking operations reliably and quickly.

In order to solve the above problems, the automated warehouse system of the present invention comprises:
A robot device that is installed in a warehouse facility and is capable of performing a picking operation in an autonomous operation mode;
a remote terminal that is installed in a remote location and that can perform a picking operation in a remote operation mode using the robot device;
a local management device that is installed in the warehouse facility and controls the robot device in the autonomous operation mode and controls the robot device based on a remote operation command from the remote terminal,
the local management device includes a determination means for determining whether picking operation is possible during the autonomous operation mode, and when the determination means determines that a condition for abandonment is satisfied because picking operation cannot be performed, the local management device transitions from the autonomous operation mode to the remote operation mode;
The local management device is characterized in that it learns the operation of the robot device in the remote operation mode and reflects the learned operation in the control of the picking operation in the autonomous operation mode.
According to this feature, the robot device is controlled by the autonomous operation mode of the local management device arranged in the warehouse facility, and picking work can be performed with high efficiency without being affected by the communication speed of infrastructure such as the Internet, and in the event that a situation occurs in which the picking work of a specified item fails in the autonomous operation mode, the picking work of the specified item can be performed in the remote operation mode instead of the autonomous operation mode. In addition, since the operation of the robot device in the remote operation mode from a remote terminal is also learned by the local management device arranged in the warehouse facility, the learned content can be quickly reflected in the control of the picking work in the autonomous operation mode, and the accuracy of the picking work can be improved, and the work efficiency can be improved.

the local management device comprises: a plurality of controllers with learning capabilities having artificial intelligence data to be used in the autonomous operation mode of the robotic device; and a local learning server having artificial intelligence data to be used in the autonomous operation mode of the robotic device;
The learning controller is connected to a local learning server;
The local learning server is characterized by accepting artificial intelligence data from a plurality of the controllers with learning function, reflecting the data in the artificial intelligence data of the local learning server, and transmitting its own artificial intelligence data to the plurality of the controllers with learning function.
According to this feature, the local learning server can learn from a large amount of data in a local environment by receiving artificial intelligence data from multiple controllers with learning functions arranged in each warehouse, and the artificial intelligence data of the local learning server can grow dramatically and efficiently through learning adapted to the local environment. In addition, the artificial intelligence data of the local learning server is distributed to multiple controllers with learning functions, so that the success rate of picking work in the autonomous operation mode and the efficiency of the picking work can be improved.

the local management device includes a local learning server having artificial intelligence data for use in the autonomous operation mode of the robotic device;
The local learning server is connected to a global learning server installed at a remote location;
The global learning server receives artificial intelligence data from the plurality of local learning servers, reflects the data in the artificial intelligence data of the global learning server, and transmits its own artificial intelligence data to the plurality of local learning servers.
According to this feature, the global learning server can learn from a huge amount of data by receiving artificial intelligence data from multiple local learning servers arranged in each warehouse, and the artificial intelligence data of the global learning server can be improved dramatically. In addition, the artificial intelligence data of the global learning server is distributed to multiple local learning servers, so that the success rate of picking work in the autonomous operation mode and the efficiency of the picking work can be improved in all local management devices.

A robot device that is installed in a warehouse facility and is capable of performing a picking operation in an autonomous operation mode;
a remote terminal that is installed in a remote location and that can perform a picking operation in a remote operation mode using the robot device;
a local management device that is installed in the warehouse facility and controls the robot device in the autonomous operation mode and controls the robot device based on a remote operation command from the remote terminal;
a work score management unit that calculates and manages a work score in a remote operation mode by the remote terminal and a work score in an autonomous operation mode ,
The local management device is characterized in that it learns the operation of the robot device in the remote operation mode and reflects the learned operation in the control of the picking operation in the autonomous operation mode .
According to this feature, businesses that use automated warehouse systems and users of warehouses, etc. can use the work scores for the remote operation mode and the autonomous operation mode managed by the work score management unit to make improvements such as making it easier to remove certain items that are difficult to pick in the autonomous operation mode but are easier to pick in the remote operation mode by strictly organizing the items in the rack, thereby making picking work more efficient.

The work score management unit is characterized in that it calculates and manages the component ratio of the robot device and the remote terminal for a specified period of time from the work score of the autonomous operation mode and the work score of the remote operation mode calculated respectively.
According to this feature, the number of remote operation modes required and the required time for the required number of autonomous operation modes can be determined in a timely manner from the work score, so that the number of remote terminals required to perform picking work reliably and quickly in a given warehouse facility can be determined.

The local management device further comprises an output means for outputting the work score of the autonomous operation mode and the work score of the remote operation mode managed by the work score management unit.
According to this feature, within the local management device, business operators and users of warehouses, etc. that use the automated warehouse system can check the work scores for the remote operation mode and the work scores for the autonomous operation mode managed by the work score management unit, and can independently try to improve the efficiency of picking work.

The output means outputs a component ratio of the robot device and the remote terminal for a predetermined period of time calculated by the work score management unit.
According to this feature, a warehouse user can know the number of remote terminals required to perform picking work reliably and quickly at a particular warehouse facility, thereby making the picking work more efficient.

1 is a conceptual diagram showing a configuration of an automated warehouse system in an embodiment of the present invention. FIG. 2 is a conceptual diagram illustrating a configuration of a local management device. FIG. 1 is a perspective view showing a robot device. 13 is a flowchart showing an autonomous operation mode. 13 is a flowchart showing a remote operation mode. 13 is a display screen displayed on an output means. 13 is a comparison display screen displayed on an output means.

The following describes an embodiment of the automated warehouse system according to the present invention.

The automated warehouse system according to the embodiment will be described with reference to Figures 1 to 7.

As shown in FIG. 1, the automated warehouse system 1 is mainly composed of multiple robot devices 3, 3, ... installed in a warehouse, local management devices 2, 2, ... arranged in each warehouse and controlling these robot devices 3, a remote management server 4 arranged in a remote location, a global learning server 5, and remote terminals 6, 6, .... The local management devices 2, 2, ..., remote management server 4, and global learning server 5 are interconnected via the Internet. The global learning server 5 and the remote terminals 6, 6, ... are interconnected via the Internet.

As shown in FIG. 2, the local management device 2 is configured with a central management device 7, a local learning server 8 having artificial intelligence data used in the autonomous operation mode of the robot device 3, and multiple controllers 9 with learning functions each connected to the central management device 7 and converting the artificial intelligence data into the operation of the robot device 3, all of which are installed in the warehouse.

All of the controllers 9 with learning functions and various devices such as conveyors 11 (see Figure 3) in the warehouse are connected to the central management unit 7 via a LAN. The central management unit 7 continuously updates and manages the shelves in the warehouse, the containers 10 (see Figure 3) stored on the shelves, and the items stored in the containers 10, and issues appropriate instructions to various devices based on received shipping instructions to perform tasks such as shipping preparation, distribution processing, and shipping. If picking work is required, the instructions are sent to the controllers 9 with learning functions.

As shown in FIG. 3, the robot device 3 is placed in a work space where picking work is performed, and is configured with a robot arm 13 and an imaging means 14. The robot arm 13 can have various configurations, but here, for example, it is a manipulator equipped with multiple joints equivalent to a human elbow or shoulder, and members similar to the fingers of a human hand, and is a type of robot arm that can grasp and lift a target item and move it to another location.

In a warehouse, the robot device 3 can be used in various situations to grasp and lift an object and move it to another location. In this embodiment, the robot device 3 is used in preparation for outgoing goods, among the tasks performed in a warehouse, such as receiving goods, putting them on shelves, storing them, preparing for outgoing goods, distribution processing, and outgoing goods. Preparation for outgoing goods refers to the process from removing a specified item from the shelf where it is stored until it is processed for distribution. The robot device 3 detects the specified item from the container 10 that has been moved using the imaging means 14, grasps and lifts the target item from the container 10 where the item is stored using the robot arm 13, moves it to the container 12 for packaging, and places it in the container 12.

The local learning server 8 is a computer equipped with an MPU and a storage medium, etc., and the storage medium stores artificial intelligence data such as operation information such as the operation mode and procedure of the robot arm 13 of the robot device 3, a program for analyzing the image captured by the imaging means 14, detection conditions when detecting an object by analyzing the image, and selection information for selecting operation information suitable for picking the target item from multiple operation information (for example, a combination table of operation information suitable for the category into which the target item is classified and the storage situation in the container 10 of the item), and is capable of learning. The controller with learning function 9 is a computer equipped with an MPU and a storage medium, etc., similar to the local learning server 8, and the storage medium stores artificial intelligence data such as operation information such as the operation mode and procedure of the robot arm 13 of the robot device 3, a program for analyzing the image captured by the imaging means 14, detection conditions when detecting an object by analyzing the image, and selection information for selecting operation information suitable for picking the target item from multiple operation information, and is capable of learning. The controller with learning function 9 also has a function of controlling the operation of the robot device 3, particularly a function of generating an operation command for the robot arm 13 using its own artificial intelligence data and controlling its operation.

When goods are delivered from the warehouse, the central management device 7 first uses a database to identify the container 10 in which the specified goods specified by the delivery command are stored, and then the container 10 is removed from the shelf and moved by the conveyor 11 (see Figure 3) to the work space where the robot device 3 is located.

As shown in FIG. 1, the remote terminal 6 is a terminal that is connected to the robot device 3 under certain conditions so as to be able to remotely control it, and is also connected so as to be able to receive images from the imaging means 14 of the robot device 3. The remote terminal 6 may have any configuration as long as it is equipped with communication means, display means, and operation input means. The remote terminal 6 is a terminal that the operator uses during manual operation, which will be described later, and may be placed in each operator's living space, or multiple remote terminals may be placed at a center where multiple operators work. In any case, it is placed at a remote location separate from the local management device 2.

The controller 9 with learning function of the local management device 2 receives picking work instructions from the central management device 7. Unique identification information that allows each individual controller 9 with learning function to be identified is set, and the central management device 7 uses the identification information to understand the operating status of all robot devices 3 in the warehouse, and designates the robot device 3 to be in charge of the picking work according to predetermined rules.

When the controller 9 with learning function receives an instruction for a picking operation, it uses its own artificial intelligence data to pick the target item in an autonomous operation mode. In the autonomous operation mode, it efficiently uses artificial intelligence data to grasp, move, and place the target item using information such as the category into which the item to be picked is classified, the storage status of the item (such as information on other items stored in the container in which the item is stored), and image information from the imaging means.

The autonomous operation mode will be described with reference to the flowchart in FIG. 4. The learning controller 9 refers to its own artificial intelligence data (step Sa01). Next, it uses the referred artificial intelligence data to generate an operation command for the robot arm 13, and starts the picking operation in the autonomous operation mode (step Sa02). During the picking operation, a determination means, which will be described in detail later, determines whether the picking operation is possible (step Sa03), and if there are no problems, the picking operation is completed (step Sa04). The central management unit 7 records and manages the identification information of the item in question and information indicating that the picking operation has been completed in the autonomous operation mode, in association with each other.

In this way, the autonomous operation mode is adjusted using the artificial intelligence data that the robot device 3 has and picks up the robot using the adjusted autonomous operation mode. This allows the efficiency of the picking operation to be maintained at a high level without being affected by the communication speed of the Internet line, as in cloud computing, where a server having artificial intelligence data is located on the Internet. In addition, as described below, the controller 9 with learning function learns in a local environment, so it can efficiently learn in a way that is suited to the warehouse facility.

The controller 9 with learning function is equipped with a determination means for determining whether picking of an item is possible during the autonomous operation mode, and if it determines that picking cannot be continued, for example because it cannot grasp the item properly, it transmits a signal to the remote management server 4 via the Internet requesting remote operation. In detail, the controller 9 with learning function determines that it cannot pick up the above-mentioned item properly based on the satisfaction of a predetermined abandonment condition, for example, the scheduled process not progressing, and at that point transmits its own identification information, which requires remote operation, to the remote management server 4 and switches to remote operation mode (step Sa05).

Various conditions can be used as the predetermined abandonment condition, but for example, a condition based on image information captured by the imaging means 14 can be that the controller 9 with learning function is unable to stably hold an item based on the image information for a period of time or longer that is set in advance.

An operator who can use each remote terminal 6 performs remote operation work during their own predetermined working hours, or during the hours that they have set. The operator logs in by inputting a unique user ID and password, which are identification information, that have been registered in advance, on the login screen displayed on the display means of the remote terminal 6. If the login is successful, the user ID is sent to the remote management server 4, and the remote management server 4 recognizes the user ID as a remote terminal 6 capable of remotely operating the robot device 3. The remote management server 4 can use the remote operation list to keep track of the status of whether or not support is available for each operator's user ID.

Next, the remote operation mode will be described using the flowchart in FIG. 5. When a signal requesting remote operation is sent from the controller 9 with learning function to the remote management server 4 (step Sb01), the remote management server 4 assigns an operator to remotely operate the robot device 3 for which remote operation has been requested (step Sb02).

In detail, the remote management server 4 transmits to the remote terminal 6 of the user ID that assigned the remote operation, as information about the remote operation, the identification information of the controller with learning function 9 that requested the remote operation, address information such as an IP address and port number that can connect to the controller with learning function 9 from outside, and the details of the picking work to be performed by the robot device 3 whose operation is controlled by the controller with learning function 9.

When the remote terminal 6 that has received the remote control information is connected to the controller with learning function 9 that controls the robot device 3, the controller with learning function 9 displays the image captured by the imaging means 14 of the robot device 3 on the operation screen displayed on the display means of the remote terminal 6. At the same time, the robot device 3 is switched from an autonomous operation mode that uses artificial intelligence data from the local learning server 8 to a remote operation mode that uses operation data from the operation input means of the remote terminal 6.

When the autonomous operation mode is switched to the remote operation mode, the remote terminal 6 starts the picking operation by remote control (step Sb03).

In the remote operation mode, there are cases where an operator assigned by the remote management server 4 to remotely operate the picking work is not skilled enough to perform the picking work, and the operator can transmit a notice to the remote management server 4 to give up the assigned picking work. The remote management server 4 waits in a state in which it can receive information from the remote terminal 6 to give up the picking work (step Sb04), and based on receiving the information from the remote terminal 6 to give up the picking work, the remote management server 4 performs a process to reassign, based on the operator's user ID, an operator with a higher proficiency score than the user ID to perform the picking work instead (step Sb05).

When the operator has not transmitted to the remote management server 4 that he or she will give up on the assigned picking work, i.e., when picking work is in progress in the remote operation mode, the controller 9 with learning function uses the determination means to determine whether the picking work is possible (step Sb06), and if the determination means determines that handover to the autonomous operation mode is possible, the controller 9 switches from the remote operation mode to the autonomous operation mode (step Sb07). For example, in the remote operation mode, the robot device 3 is in charge of operations up to the point of grasping the item, and the autonomous operation mode takes over the movement and placement of the item. If the determination means does not determine that handover to the autonomous operation mode is possible, the picking work is completed in the remote operation mode (step Sb08).

In addition, in steps Sb03 to Sb08, the operation image in the remote operation mode is captured by the imaging means 14 of the robot device 3 and transmitted to the controller with learning function 9, and the control data of the remotely operated robot device 3 is transmitted to the controller with learning function 9. When the picking work is completed in the remote operation mode only, or in a mixture of the remote operation mode and the autonomous operation mode (step Sb08), the controller with learning function 9 sequentially observes and analyzes the operation of the robot device 3 in the picking work using the image data and control data, digitizes the operation of the robot arm 13 in the remote operation mode, and reflects it in its own artificial intelligence data. The artificial intelligence data learned by each controller with learning function 9 is transmitted to the local learning server 8, and the local learning server 8 reflects it in the artificial intelligence data.

The local learning server 8 receives artificial intelligence data from each controller with learning function 9, 9, ... arranged in the warehouse, integrates it in the local learning server 8, reflects the artificial intelligence data, and transmits its own artificial intelligence data to the multiple controllers with learning function 9, 9, ..., which learn the received artificial intelligence data. By receiving artificial intelligence data from multiple controllers with learning function 9, 9, ..., it is possible to learn from a huge amount of data, and the artificial intelligence data of the local learning server 8 can be dramatically improved, and by distributing this artificial intelligence data of the local learning server 8 to the multiple controllers with learning function 9, 9, ..., it is possible to improve the success rate of picking work in the autonomous operation mode and the efficiency of the picking work. In addition, since the local learning server 8 performs learning in a local environment, it is possible to efficiently perform learning suited to the warehouse facility.

In this way, the operation of the robot device 3 in the remote operation mode is learned by the controller 9 with learning function and the local learning server 8 constituting the local management device 2 arranged in the warehouse facility. Therefore, unlike cloud computing, in which a server having artificial intelligence data is arranged on the Internet, the learned content can be quickly reflected in the control of the picking work in the autonomous operation mode without being affected by the communication speed of the Internet line, and the accuracy of the picking work can be improved, and the work efficiency can be improved.

The local learning server 8 is also connected to a global learning server 5 installed in a remote location via the Internet, and transmits artificial intelligence data to the global learning server 5 at a predetermined timing. In this way, by communicating with the global learning server 5 outside of core times when picking work is frequently performed, such as at night, the controller 9 with learning function can concentrate on the picking work and perform calculation processing during core times, and communication with the global learning server 5 does not reduce the work efficiency of the picking work.

The global learning server 5 also receives artificial intelligence data from multiple local learning servers 8, 8, ... arranged in each warehouse, reflects the data in the global learning server 5, and transmits its own artificial intelligence data to the multiple local learning servers 8, 8, ..., which then learn the received artificial intelligence data. By receiving artificial intelligence data from multiple local learning servers 8, 8, ..., learning can be performed from a huge amount of data, and the artificial intelligence data of the global learning server 5 can be dramatically improved. By distributing the artificial intelligence data of the global learning server 5 to the multiple local learning servers 8, 8, ..., the success rate of picking work in the autonomous operation mode and the efficiency of the picking work can be improved in all local management devices 2, 2, ....

The central management device 7 calculates and manages the work score in the autonomous operation mode and the work score in the remote operation mode of the picking work by the robot device 3 for each user ID of the business operator, warehouse, etc. that uses the robot device 3. The work score in the autonomous operation mode is calculated according to a predetermined rule using, for example, the time required for the picking work, the number of picking works in the autonomous operation mode, and the scores of the difficulty of each picking work. The work score in the remote operation mode is calculated according to a predetermined rule using, for example, the time required for the picking work, the number of picking works in the remote operation mode, and the scores of the proficiency of the person who performed the remote operation. As described above, if the picking work is abandoned and another operator is reassigned, the proficiency score of the operator who actually performed the picking work is reflected in the work score in the remote operation mode.

The central management device 7 can output the breakdown of the work score of the autonomous operation mode and the work score of the remote operation mode calculated for each user ID on the display of the computer 17 connected to the central management device 7 (output means). FIG. 6 shows the output screen 20. The output screen 20 displays the user ID, the total number of picking operations in a period (here, October 2020), the total time required for the picking operations, and the percentage of the work score of the autonomous operation mode and the work score of the remote operation mode in the total number of picking operations. In FIG. 6 and FIG. 7, the work score of the autonomous operation mode and the work score of the remote operation mode indicate the percentage of the number of picking operations in which picking operations were performed. Note that, for example, in a picking operation in which a part of the picking operation was performed in the autonomous operation mode, the percentage (for example, 0.3) of the autonomous operation mode in that one picking operation is calculated. The central management device 7 calculates the number of picking operations performed by the autonomous operation mode by adding up the percentages of all the autonomous operation modes and dividing the picking operations performed entirely in the autonomous operation mode by 1.

Accordingly, businesses and users of warehouses, etc. using the automated warehouse system 1 can use the work scores for the remote operation mode and the autonomous operation mode managed by the central management device 7 (work score management unit) to make improvements such as making it easier to remove certain items that are difficult to pick in the autonomous operation mode but are easier to pick in the remote operation mode by strictly organizing the items in the container 10, thereby improving the efficiency of picking work.

As shown in FIG. 6, the output screen 20 has a display 22 that converts and outputs the work scores of the autonomous operation mode and the remote operation mode calculated for a specified period of time as the component ratio of the robot devices 3 and the remote terminals 6 for the specified period of time, in other words, the number of robot devices 3 remotely operated by one remote terminal operating in the remote operation mode. According to this, the greater the number of robot devices 3 that were in charge of picking work by the remote terminal 6 operating in the remote operation mode, the smaller the proportion of the remote operation mode is. Therefore, it is easy to visually grasp the ratio of the work scores of the remote operation mode and the work scores of the autonomous operation mode for a specified period of time.

In addition, as shown in FIG. 7, the central management unit 7 can also output a comparison output screen 21, which allows the ratio of the work score in the remote operation mode to the work score in the autonomous operation mode over a specified period of time to be compared in a chronological order, making it possible to easily check the impact that the improvements have had on the efficiency of the picking work.

Although the embodiments of the present invention have been described above with reference to the drawings, the specific configuration is not limited to these embodiments, and the present invention also includes modifications and additions that do not deviate from the gist of the present invention.

For example, in the above embodiment, the work score for the remote operation mode and the work score for the autonomous operation mode managed by the central management device 7 constituting the local management device 2 are output by the output means within the local management device 2. However, this is not limited to the above. The work score for the remote operation mode and the work score for the autonomous operation mode may be sent to a server of the provider of the automated warehouse system 1 located on the Internet, and may be used by the provider to suggest improvements to users of the automated warehouse system 1. In this case, the work score for the remote operation mode and the work score for the autonomous operation mode do not have to be output as in the output screen described above.

The central management device 7 or the provider's server may also use the work score in the remote operation mode, the work score in the autonomous operation mode, and the type of item for each picking task to determine whether there is a high proportion of remote operations for certain items, and present this to the user using an output means, etc.

The central management device 7 or the provider's server may also assign a difficulty level to the picking task based on the proficiency score of the user ID of the operator who performed the picking task in the remote operation mode, the task time, etc., and present to the user the number of picking tasks of each difficulty level out of the total number of picking tasks in the remote operation mode using an output means, etc. This allows the user to understand which difficulty level of picking tasks has a large impact on the work score in the remote operation mode.

The predetermined abandonment condition set by the learning controller 9 is not limited to a condition based on image information as exemplified in the above embodiment, but may be, for example, a condition that the robot arm 13 is equipped with a pressure sensor and the pressure measured by this pressure sensor, i.e., the gripping force applied when grasping the item, is greater than a preset value. This preset value is set based on the pressure when the item is accurately removed. This allows the item to be accurately removed only with reasonable force, preventing damage to the item. In addition, if the pressure sensor measures a value greater than the preset value, there is a possibility that the robot arm 13 is in contact with other items, containers, etc., and in this case too, it can be determined that the removal operation cannot be performed accurately.

The robot arm 13 is not limited to a configuration with parts similar to a human elbow or shoulder, or the fingers of a human hand, but may be of any configuration as long as it is a type of robot arm that can grasp and lift an object and move it to another location, and may be configured to lift an item by suction, for example.

In the above embodiment, each controller 9 with learning function is assigned unique identification information so that the remote management server 4 can grasp the operating status of each controller. However, this is not limited to the above. For example, if the central management unit 7 to which the robot device 3 belongs has unique identification information, the robot device 3 can be identified based on which central management unit 7 it belongs to. Therefore, as long as each robot device can be identified within the central management unit 7 to which it belongs, it is not necessary for the entire automated warehouse system 1 to be configured to be assigned unique identification information.

In addition, when there are two robot arms, the two robot arms are operated together in the autonomous operation mode and the remote control mode, but this is not limited to the above, and the robot arm used only in the autonomous operation mode and the robot arm used only in the remote control mode may each be configured to operate completely independently.

In addition, the local management device 2 may be configured, for example, to omit the artificial intelligence data and learning function of the controller with learning function 9, and be responsible only for generating operation commands for the robot arm 13 using the artificial intelligence data and controlling its operation, and the autonomous operation mode of the robot device 3 may always be performed by referring to the artificial intelligence data of the local learning server 8, and learning may also be performed by the local learning server 8.

In addition, in the above embodiment, the robot device 3 is described as being used in the preparation for leaving the warehouse, but this is not limiting, and it may be used in other sections, such as the entry work section.

In the above embodiment, the remote terminal 6 is directly connected based on the address information of the controller 9 with learning function, that is, connected in a so-called peer-to-peer manner. However, the present invention is not limited to this. For example, the remote management server 4 or the central management device 7 may be used as a relay to exchange operation signals, etc., and the remote terminal 6 and the robot device 3 may not be directly connected, thereby improving security on the network.

1 Automated warehouse system 2 Local management device 3 Robot device 4 Remote management server 5 Global learning server 6 Remote terminal 7 Central management device (work score management section)
8: Local learning server 9: Controller with learning function 10: Container 11: Conveyor 13: Robot arm 14: Imaging means 17: Computer 20: Output screen 21: Comparison output screen 22: Display

Claims (7)

A robot device that is installed in a warehouse facility and is capable of performing a picking operation in an autonomous operation mode;
a remote terminal that is installed in a remote location and that can perform a picking operation in a remote operation mode using the robot device;
a local management device that is installed in the warehouse facility and controls the robot device in the autonomous operation mode and controls the robot device based on a remote operation command from the remote terminal,
the local management device includes a determination means for determining whether picking operation is possible during the autonomous operation mode, and when the determination means determines that a condition for abandonment is satisfied because picking operation cannot be performed, the local management device transitions from the autonomous operation mode to the remote operation mode;
The automated warehouse system is characterized in that the local management device learns the operation of the robot device in the remote operation mode and reflects the learned operation in the control of the picking work in the autonomous operation mode. the local management device comprises: a plurality of controllers with learning capabilities having artificial intelligence data to be used in the autonomous operation mode of the robotic device; and a local learning server having artificial intelligence data to be used in the autonomous operation mode of the robotic device;
The learning controller is connected to a local learning server;
The automated warehouse system according to claim 1, characterized in that the local learning server receives artificial intelligence data from a plurality of the controllers with learning function, reflects the data in the artificial intelligence data of the local learning server, and transmits its own artificial intelligence data to the plurality of the controllers with learning function. the local management device includes a local learning server having artificial intelligence data for use in the autonomous operation mode of the robotic device;
The local learning server is connected to a global learning server installed at a remote location;
The automated warehouse system described in claim 1 or 2, characterized in that the global learning server accepts artificial intelligence data from a plurality of the local learning servers, reflects the data in the artificial intelligence data of the global learning server, and transmits its own artificial intelligence data to the plurality of the local learning servers. A robot device that is installed in a warehouse facility and is capable of performing a picking operation in an autonomous operation mode;
a remote terminal that is installed in a remote location and that can perform a picking operation in a remote operation mode using the robot device;
a local management device that is installed in the warehouse facility and controls the robot device in the autonomous operation mode and controls the robot device based on a remote operation command from the remote terminal;
a work score management unit that calculates and manages a work score in a remote operation mode by the remote terminal and a work score in an autonomous operation mode,
The automated warehouse system is characterized in that the local management device learns the operation of the robot device in the remote operation mode and reflects the learned operation in the control of the picking work in the autonomous operation mode. The automated warehouse system according to claim 4, characterized in that the work score management unit calculates and manages the composition ratio of the robot device and the remote terminal during a predetermined period from the work score of the autonomous operation mode and the work score of the remote operation mode calculated respectively. The automated warehouse system according to claim 5, characterized in that the local management device further comprises an output means capable of outputting the work score of the autonomous operation mode and the work score of the remote operation mode managed by the work score management unit. The automated warehouse system according to claim 6, characterized in that the output means outputs the composition ratio of the robot device and the remote terminal for a predetermined period calculated by the work score management unit.

Images