Picking support system, and picking support method

Abstract

To improve business efficiency in a work area in consideration of a state of an operator.SOLUTION: A robot managing apparatus transmits picking instructions to an operator managing apparatus carried by an operator selected as an operator who carries out the next picking work, and transmits, if there is the possibility that the selected operator cannot take care of the picking instructions, to each of the operator managing apparatus, a request inquiring whether or not it is possible to take care of the picking instructions. Each of the operator managing apparatus calculates, based on the situation obtained upon reception of the picking instructions, a cost value based on at least one of distance costs and time costs required for the picking operation based on the request to transmit it to the robot managing apparatus. The robot managing apparatus selects an operator based on the cost value received from each of the operator managing apparatus to transmit the picking instructions.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a picking support system and a picking support method. [Background technology]

[0002] In warehouses, robots are sometimes used to assist workers in picking products from shelves. In this case, the work efficiency in the warehouse can be improved by having the workers and the robot work together. For example, in a situation where the workers and the robot work together, if the robot instructs the workers on the next picking location according to the operating status of the robot, it is possible to reduce the workers' picking time. Furthermore, for example, Patent Document 1 discloses a system in which a robot transports products picked by a worker to a station. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Special Publication No. 2022-519054 Summary of the Invention [Problem to be solved by the invention]

[0004] However, in situations where workers and robots work together, even if the worker is instructed on the next picking location based on the robot's operating status, situations may arise where it is not necessarily optimal for the worker to move to the location instructed due to unexpected behavior of the worker (going to the wrong location, stopping for some reason, etc.) or personal characteristics (walking slowly, needing frequent rests).

[0005] The present invention has been made in consideration of the above circumstances, and its purpose is to provide a picking support system and a picking support method that can improve work efficiency in the work area while taking into account the situation of the workers. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems, one aspect of the present invention is a picking support system including a robot management device mounted on a robot and managing the robot, and a worker management device attached to each of a plurality of workers performing picking work in a work area, the robot management device including a first transmission unit that transmits a picking instruction to a worker management device carried by a worker selected as the worker performing the next picking work, a determination unit that determines whether or not the selected worker is likely to be unable to respond to the picking instruction after transmitting the picking instruction, a second transmission unit that transmits a request to each of the worker management devices inquiring whether or not the selected worker is able to respond to the picking instruction in response to the determination that the selected worker is likely to be unable to respond, and a third transmission unit that receives the request and transmits a request to each of the worker management devices inquiring whether or not the selected worker is able to respond to the picking instruction. a receiving unit that receives cost values ​​transmitted from each of the plurality of worker management devices, a selection unit that selects a worker based on the respective cost values ​​received from each worker management device, and a third transmission unit that transmits a picking instruction to the worker management device of the selected worker, wherein the worker management device has a receiving unit that receives a request, a cost calculation unit that calculates a cost value based on at least one of a distance cost based on a travel distance required to handle the picking based on the request and a time cost based on the time required to handle the picking based on the request based on the situation at the time the request is received, and a transmission unit that transmits the calculated cost to a robot management device that is the sender of the request.

[0007] Another aspect of the present invention is a picking support method executed in a picking support system including a robot management device mounted on a robot and managing the robot, and a worker management device attached to each of a plurality of workers performing picking work in a work area, the robot management device transmits a picking instruction to a worker management device carried by a worker selected as the worker performing the next picking work, the robot management device determines, after transmitting the picking instruction, whether or not there is a possibility that the selected worker will not be able to respond to the picking instruction, and in response to determining that there is a possibility that the selected worker will not be able to respond, the robot management device transmits a request to each of the worker management devices inquiring whether or not the selected worker is able to respond to the picking instruction, and each of the worker management devices responds to the request. and each of the worker management devices calculates a cost value based on at least one of a distance cost based on a travel distance required to handle the picking based on the request and a time cost based on the time required to handle the picking based on the request based on the situation at the time the request is received, each of the worker management devices transmits the calculated cost to a robot management device that is a sender of the request, the robot management device receives cost values ​​respectively transmitted from a plurality of worker management devices that received the request, the robot management device selects a worker based on the respective cost values ​​received from each worker management device, and the robot management device transmits a picking instruction to the worker management device of the selected worker. Effect of the Invention

[0008] As described above, according to the present invention, it is possible to improve the work efficiency in a work area by taking into consideration the status of workers. [Brief description of the drawings]

[0009] [Figure 1]FIG. 1 is a diagram for explaining an overview of the work in warehouse WH where picking is performed. [Diagram 2] FIG. 2 is a diagram showing the general configuration of a picking system S. [Diagram 3] 1 is a schematic configuration diagram for explaining the functions of a picking support system including a plurality of human agents HA and a plurality of robot agents RA. [Figure 4] 1 is a flow chart for explaining the operation of a picking support system including a plurality of human agents HA and a plurality of robot agents RA. [Diagram 5] 1 is a flow chart for explaining the operation of a picking support system including a plurality of human agents HA and a plurality of robot agents RA. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] A picking support system according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram for explaining an overview of a picking operation performed in a work area. In FIG. 1, the work area is a warehouse WH. A plurality of shelves T (here, T1, T2, T3, T4, T5, T6, T7, T8, T9, and T10) are installed in the warehouse WH. Each shelf T stores a plurality of products. The shelves T are installed at intervals from each other. Therefore, for example, it is possible to pass between shelves T1 and T2 which are installed adjacent to each other. The number of shelves is 10 in FIG. 1, but it may be a number other than 10.

[0011] In the warehouse WH, a plurality of automatic transport robots AMR (here, AMRa, AMRb, AMRc) are used. The automatic transport robot AMR moves to the shelf T on which the product to be picked is stored, stores the product picked by the worker PW, and moves to the shelf T on which the next product to be picked is stored or to the packing space PS. The automatic transport robot AMR is provided with a storage section for storing the picked product. As long as the storage section can store the product, the automatic transport robot AMR itself may be provided with a storage section, or a storage container for storing the product may be mounted. Although three automatic transport robots AMR are shown in FIG. 1, the number may be one or two, or four or more.

[0012] Workers PW (here, PWa, PWb, PWc) move to the shelf T in the warehouse WH where the product to be picked is stored, pick up the product, and store it in the storage section of the target automatic transport robot AMR that stores the product. When picking is completed, the worker moves to the shelf T where the next product to be picked is stored and performs the next picking task. Here, the automatic transport robot AMR that stores the product has moved to the vicinity of the shelf T where the product to be picked is stored, so the worker PW simply stores the product in the automatic transport robot AMR that is in the vicinity of the shelf T where the product to be picked is stored.

[0013] When the automatic transport robot AMR has stored multiple products to be picked, it moves to the packing space PS. A workbench is installed in the packing space PS for packing the picked products. For example, a worker PWc takes the products from the storage section of the automatic transport robot AMR that has arrived at the packing space PS and packs them. The packed products are transported to a later process and shipped to a designated delivery destination or picked up by a collection staff member.

[0014] FIG. 2 is a configuration diagram showing a schematic configuration of the picking system S used in the warehouse WH shown in FIG. The picking system S includes a host system 1, a robot control system RCS, an automatic transport robot AMR (here, AMARa, AMRb), a navigation system NVG (here, NVGa, NVGb), shelves T (here, T1 to Tn), and a packaging space PS.

[0015] The upper system 1 performs data processing to realize various functions such as accepting product orders, managing the inventory of products stored on each shelf T, generating picking lists in response to product orders, transmitting the picking lists, and managing the shipping of packaged products. The robot control system RCS identifies the shelf T on which the items included in the picking list are stored, and transmits the coordinates of the identified shelf T to the automatic transport robot AMR.

[0016] The automatic transfer robot AMR (here, AMARa, AMRb) has a robot agent RA and a traveling device RG. The robot agent RA receives a picking list from the host system 1, queries the robot control system RCS as to which shelf the product to be picked is stored on in the received picking list, and receives data indicating the coordinates. The robot agent RA also regards the received coordinates as target coordinates, obtains a route from the current position to the target coordinates, and instructs the traveling device RG to move along the route. The robot agent RA is, for example, a software agent executed on a computer mounted on the automatic transfer robot AMR. The robot agent RA is mounted on the automatic transfer robot AMR and functions as a robot management device that manages the automatic transfer robot AMR. The running device RG runs within the warehouse HW based on instructions from the robot agent RA. A plurality of such automatic transport robots AMR are used in a warehouse. In Fig. 2, two automatic transport robots AMR are illustrated, but the number of automatic transport robots AMR may be one, or three or more.

[0017] The navigation system NVG (here, NVGa, NVGb) is attached to a worker. A plurality of navigation systems NVG are used in a warehouse. One navigation system NVG is attached to one worker. The navigation system NVG includes a human agent HA and a barcode reader BR. In this embodiment, a case will be described in which a barcode is attached to a product and a worker reads the barcode using the barcode reader BR, but other aspects may be used as long as identification information capable of identifying the product can be obtained from the product. For example, the barcode may be a one-dimensional barcode or a two-dimensional barcode. Also, instead of attaching a barcode to a product, an RFID (Radio Frequency Identification) tag may be attached to the product. In this case, identification information stored in the RFID tag may be used as identification information for identifying the product. Furthermore, when an RFID tag is attached to a product, the navigation system NVG may use a tag reader that wirelessly communicates with the RFID tag instead of the barcode reader BR.

[0018] The human agent HA is, for example, a software agent executed in an electronic device such as a smartphone or a tablet. The electronic device on which the software agent is executed may be a wearable terminal such as smart glasses or a smart watch. The human agent HA functions as a worker management device that is attached to each of multiple workers who perform picking work in a warehouse. The barcode reader BR reads the barcode attached to the product picked by the worker, and transmits the reading result to an electronic device (e.g., a human agent HA) carried by the worker. As a result, the reading result by the barcode reader BR is transmitted from the human agent HA to the host system 1. The host system 1 can perform operations such as matching the picked product with the products on the picking list. Here, the barcode reading result may be transmitted from the barcode reader BR to both the host system 1 and the human agent HA.

[0019] The navigation system NVG may be attached by having the electronic device on which the human agent HA is executed and the barcode reader held in the worker's hand, or at least one of the electronic device and the barcode reader BR may be attached by being worn on any part of the worker's body, such as the wrist, arm, or head.

[0020] Here, a picking support system may be constructed by at least one robot agent RA and a plurality of human agents HA. Next, the operation of the picking support system will be described.

[0021] FIG. 3 is a schematic diagram for explaining the function of a picking support system including a plurality of human agents HA and a plurality of robot agents RA. The human agent HA includes a sensor H11, a self-information storage unit H12, an agent storage unit H13, an agent communication unit H14, a request communication unit H15, and a picking planning unit H16. has.

[0022] The sensor H11 measures the position of its own human agent HA at every fixed time interval, and stores the measurement result in the self-information storage unit H12. The sensor H11 can use, for example, a beacon.

[0023] The self-information storage unit H12 stores various information related to the navigation system NVG in which the human agent HA itself is mounted. More specifically, it stores various information related to the worker carrying the human agent HA itself, and the various information includes, for example, current location information, current status, instructions during picking, next corresponding picking instructions, next destination, etc. That is, the self-information storage unit H12 of the navigation system NVGa stores various information related to the worker carrying the navigation system NVGa, and the self-information storage unit H12 of the navigation system NVGb stores various information related to the worker carrying the navigation system NVGb.

[0024] The current location information indicates the current location of the human agent HA within the warehouse based on the measurement results obtained from the sensor H11. The current state is information indicating whether the human agent HA itself is in a state (status) of picking, moving, or waiting to be picked. "Picking" indicates a state in which the human agent HA is picking an item included in the picking list. "Moving" indicates a state in which the agent is moving to the location of the next shelf for picking. "Waiting to pick" indicates a state in which the agent HA is not performing picking work, is not moving, and is available for the next picking work. The picking instruction indicates the instruction number that identifies the picking instruction that is currently being handled (executed). The picking instruction is information that instructs the picking to be performed by a worker selected from among multiple navigation systems NVG as the target for the picking to be performed. The next corresponding picking instruction indicates an instruction number that identifies the next corresponding (executed) picking instruction. The next destination indicates the destination to which the human agent HA will head next.

[0025] The agent storage unit H13 stores various information related to the navigation system NVG in which the other human agents HA are installed. The agent storage unit H14 stores, for example, an agent ID (IDentification), an agent type, current location information, a current state, a next destination, and instructions during picking. The agent ID is identification information that individually identifies each agent. The agents referred to here include multiple human agents and multiple robot agents. By using the agent ID, multiple human agents and multiple robot agents can be individually identified. The agent type indicates whether the agent is a human agent or a robot agent. By referring to the agent type, it is possible to identify whether the agent is a human agent or a robot agent. The current location information is information that indicates the current location of each agent within the warehouse. The current state is information indicating whether each agent is in the process of picking or moving. The next destination indicates the next destination for each agent. The picking instruction in progress indicates the picking instruction number that each agent is currently handling.

[0026] The agent communication unit H14 is connected to the agent communication units 14 of other human agents HA so as to be able to communicate wirelessly. The agent communication unit H14 reads out and transmits each piece of information stored in the self-information storage unit H12 to the other human agents HA at regular intervals. For example, the agent communication unit H14 transmits various pieces of information (current location information, current status, instructions for picking, next corresponding picking instructions, next destination, etc.) stored in the self-information storage unit H12 to each of the other human agents HA by unicast. In addition, the agent communication unit H14 receives various information transmitted from other human agents HA and stores it in the agent information storage unit H13. This allows the human agent HA to notify other human agents HA of various information about itself, and also allows the agent HA to grasp the status of other human agents HA by acquiring various information about the other human agents HA.

[0027] The request communication unit H15 is wirelessly connected to the robot agent RA so as to be able to communicate with the robot agent RA. The request communication unit H15 includes a receiving unit H151 and a transmitting unit H152. The receiving unit H151 receives a picking request transmitted from a robot agent RA. The picking request is a request to inquire of multiple human agents HA as to whether they can handle the picking. The transmission unit H152 transmits the cost value calculated by the picking planning unit H16 to the robot agent RA that is the sender of the request.

[0028] The picking planning unit H16 performs various processes related to the picking plan. The picking plan unit H16 includes a cost calculation unit H161 and a deletion processing unit H162. The cost calculation unit H161 calculates a cost value based on at least one of a distance cost based on the travel distance required to respond to picking based on a request and a time cost based on the time required to respond to picking based on a request, based on the situation at the time the request is received.

[0029] When a request is received after the picking instruction is stored, the deletion processing unit H162 deletes the picking instruction from the self-information storage unit H12.

[0030] The robot agent RA includes a sensor R11, a self-information storage unit R12, an agent information storage unit R13, an agent communication unit R14, a request communication unit R15, and a picking plan unit R16.

[0031] The sensor R11 measures the position of the automatic transfer robot AMRa equipped with its own robot agent RA at regular intervals and stores the measurement result in the self-information storage unit R12. The sensor R11 stores map data showing, for example, the arrangement of shelves and aisles in a warehouse, and also obtains odometry information and the like from the running device RG showing the amount of rotation of the wheels measured as the running device RG runs, and measures the position of the automatic transfer robot AMRa in the warehouse based on these data.

[0032] The self-information storage unit R12 stores various information related to the automatic transfer robot AMR on which the robot agent RA itself is mounted. For example, the self-information storage unit R12 stores current location information, current state, instructions during picking, next corresponding picking instructions, and next destination. That is, the self-information storage unit R12 of the automatic transfer robot AMRa stores various information related to the automatic transfer robot AMRa, and the self-information storage unit R12 of the automatic transfer robot AMRb stores various information related to the automatic transfer robot AMRb.

[0033] The current position information is information that indicates the current position of the robot agent RA within the warehouse. The current state is information indicating whether the robot agent is picking, moving, or waiting to be picked. The picking instruction in progress indicates the picking instruction number currently being handled (executed). The next corresponding picking instruction indicates the instruction number of the next corresponding (to be executed) picking. The next destination indicates the next destination to which the robot agent RA will head.

[0034] The agent information storage unit R13 stores various information related to the automatic transfer robot AMR on which other robot agents RA are mounted. The agent information storage unit R13 stores, for example, an agent ID, an agent type, current location information, a current state, a next destination, and instructions during picking. The current location information is information that indicates the current location of each agent within the warehouse. The current state is information indicating whether each agent is in the process of picking or moving. The next destination represents the next destination for each agent. The picking instruction in progress indicates the picking instruction number that each agent is currently handling.

[0035] The agent communication unit R14 is wirelessly connected to the agent communication units R14 of other robot agents RA so as to be able to communicate with each other. The agent communication unit R14 reads out and transmits each piece of information stored in the self-information storage unit R12 to the other robot agents RA at regular intervals.

[0036] The request communication unit R15 is wirelessly connected to the human agent HA so as to be able to communicate with the human agent HA. The request communication unit R15 has a first transmission unit R151, a second transmission unit R152, a reception unit R153, and a third transmission unit R154. The first transmission unit R151 transmits a picking instruction to a worker management device carried by the worker selected as the worker who will perform the next picking operation. In response to determining that there is a possibility that the agent cannot respond, the second transmission unit R152 transmits a request to each of the human agents HA to inquire whether the agent can respond to the picking. The receiving unit R153 receives the cost values ​​transmitted from each of the multiple worker management devices that have received the request. The third transmission unit R154 transmits a picking instruction to the worker management device of the selected worker.

[0037] The picking planning unit R16 performs various processes related to picking planning. The picking planning unit R16 has a determination unit R161 and a selection unit R162. After transmitting a picking instruction, the determination unit R161 determines whether or not there is a possibility that the selected worker cannot handle the picking instruction. Furthermore, if a reference time has elapsed since the first transmission unit R151 transmitted the picking instruction, the determination unit R161 determines that the selected worker may not be able to respond to the picking instruction. In addition, after the first transmission unit R151 sends a picking instruction, if the judgment unit R161 receives a response from the worker management device that received the picking instruction indicating that the picking instruction cannot be fulfilled, the judgment unit R161 judges that the selected worker may not be able to fulfill the picking instruction. In addition, the determination unit R161 performs a determination in either case where a picking instruction is transmitted by the first transmission unit R151 or where a picking instruction is transmitted by the third transmission unit R154.

[0038] The selection unit R162 selects a worker based on each cost value received from each worker management device.

[0039] FIG. 4 is a flow chart illustrating the operation of a picking support system including a plurality of human agents HA and a plurality of robot agents RA. When the upper system 1 receives an order, it generates a picking list based on the received order content. When the picking list is generated, the upper system 1 selects an automatic transfer robot AMR (for example, an automatic transfer robot AMRa) to which the picking list is to be sent from among multiple automatic transfer robots AMR present in the warehouse, and transmits the picking list to the selected automatic transfer robot AMRa (step S101).

[0040] When the robot agent RA of the automatic transport robot AMRa receives a picking list from the host system 1, it extracts the products to be picked based on the received picking list and identifies the shelf on which the extracted products are stored (step S201). The shelf on which the product is stored may be specified, for example, by making an inquiry to the host system 1 by specifying the product, and acquiring from the host system 1 identification information of the shelf on which the product is stored. The robot agent RA of the automatic transport robot AMRa identifies the coordinates of the shelf to which the identification information is assigned based on the identified shelf identification information. The robot agent RA may identify the shelf coordinates based on the correspondence between the shelf identification information and the shelf coordinates stored in advance. The robot agent RA may also specify the identification information of the identified shelf and inquire of the robot control system RCS about the coordinates where the shelf is installed. In this case, the robot control system RCS reads out the coordinates in the warehouse where the shelf corresponding to the shelf identification information is installed from the location data based on this inquiry, and transmits it to the automatic transport robot AMR that made the inquiry. This allows the robot agent RA to identify the coordinates of the shelf.

[0041] Next, the robot agent RA of the automatic transfer robot AMRa sets the shelf indicated by the specified shelf coordinates as the destination, and controls the automatic transfer robot AMRa to move to the position indicated by the specified shelf coordinates. As a result, the automatic transfer robot AMRa starts moving from its current position to the destination shelf (step S202). Here, if the automatic transfer robot AMRa is currently performing a picking operation, the robot agent RA starts moving to the destination shelf after the picking operation is completed. Next, the robot agent RA designates the position of the shelf that is the destination as the next picking position, and transmits picking requests to each of the human agents HA requesting them to come to the picking position (step S203). In this case, if two navigation systems NVGa and NVGb exist in the warehouse, the robot agent RA transmits picking requests to the human agent HA of the navigation system NVGa and the human agent HA of the navigation system NVGb, respectively.

[0042] Thereafter, when the automatic transfer robot AMRa arrives at the destination and stops traveling, the robot agent RA detects that the automatic transfer robot AMRa has arrived at the destination (step S204).

[0043] On the other hand, when a picking request is transmitted from the robot agent RA, each human agent HA receives the picking request. For example, when the human agent HA of the navigation system NVGa receives the picking request, it performs cost calculation (step S301), and similarly, the human agent HA of the navigation system NVGb also performs cost calculation (step S401).

[0044] For example, the picking plan unit H16 of the human agent HA of the navigation system NVGa calculates the cost of a worker PWa carrying the navigation system NVGa moving to a location instructed to pick up based on a picking request and information stored in the agent information storage unit H13 and the self information storage unit H12. The calculated cost may be at least one of a distance cost and a time cost. For example, the picking plan unit H16 calculates a distance cost according to the distance traveled along a route from the current position of the human agent HA stored in the self-information storage unit H12 to a position (position of the target to be picked) specified as a destination by the picking request. For example, the distance cost is a value such that the shorter the distance, the smaller the cost value, and the longer the distance, the larger the cost value.

[0045] For example, the picking plan unit H16 calculates a time cost according to the time it takes to move along a route from the current position of the human agent HA stored in the self-information storage unit H12 to a position specified as a destination by the picking request (the position of the object to be picked). For example, the time cost is a value such that the shorter the distance, the smaller the cost value, and the longer the distance, the larger the cost value. In addition, for example, when the current state of the human agent HA stored in the self-information storage unit H12 indicates that it is waiting to be picked, the picking planning unit H16 can quickly respond to the next picking, and therefore calculates the distance cost of moving along the route from the current position of the human agent HA stored in the self-information storage unit H12 to the position specified as the destination by the picking request.

[0046] For example, the picking plan unit H16 calculates a time cost according to the time required to move along a route from the current position of the human agent HA itself stored in the self-information storage unit H12 to a position specified as a destination by the picking request (the position of the target to be picked). For example, the time cost is a value such that the shorter the time, the smaller the cost value, and the longer the time, the larger the cost value. Here, when the current state of the human agent HA stored in the self-information storage unit H12 indicates that it is waiting to be picked, the picking planning unit H16 can quickly respond to the next picking, and therefore can calculate the time cost based on the travel time. In addition, when the current state of the human agent HA stored in the self-information storage unit H12 is picking, the picking planning unit H16 can calculate the time cost based on the time until the current picking is completed and the travel time required to move from the position of the shelf where the current picking work is being performed to the position of the next shelf to be picked, which is included in the picking request. In addition, when the current state of the human agent HA stored in the self-information storage unit H12 is that it is moving, the picking planning unit H16 can calculate the time cost based on the time required to move to the destination to which the human agent HA is currently moving and complete the picking operation, and the travel time required to move from the position of the shelf where the picking operation has been completed to the position of the next shelf to be picked, which is included in the picking request.

[0047] When the cost value is calculated, the picking plan unit H16 transmits the calculated cost value to the robot agent RA that is the transmission source of the picking request via the request communication unit H15 (step S302). In addition, such calculation and transmission of the cost is performed by each of the multiple human agents HA. Therefore, the human agent HA of the navigation system NVGb also calculates the cost and transmits the calculated cost to the robot agent RA that is the sender of the picking request (step S402).

[0048] The robot agent RA (robot agent RA of the automatic transport robot AMRa) that is the sender of the picking request selects one human agent HA to handle the picking request, based on the costs received from each human agent HA (step S205).

[0049] When the robot agent RA selects a human agent HA, if the robot agent RA places a priority on starting picking early, it can select the human agent HA with the smallest time cost among the distance cost and the time cost. On the other hand, if the robot agent RA places a priority on shortening the travel distance of the worker, it can select the human agent HA with the smallest distance cost among the distance cost and the time cost. If the robot agent RA places a priority on the distance cost, it can select the workers so that they are distributed as evenly as possible within the warehouse.

[0050] Also, the robot agent RA may select a human agent HA based on both the distance cost and the time cost, for example, the human agent HA with the smallest sum of the distance cost and the time cost. In this way, when selecting a human agent HA, whether to prioritize distance cost or time cost, or to use both distance cost and time cost, may be determined in advance as an operation rule. In addition, although two types of cost, distance cost and time cost, are described here, a cost other than these two types (such as a cost based on the worker's picking history on the day) may be obtained.

[0051] In addition, the case where one human agent HA is selected based on the cost calculated by each of the human agents HA has been described here, but at the time when the robot agent RA receives the cost from each of the human agents HA, it may have already arrived at the destination based on the picking request, or it may be in the process of moving. If the robot agent RA has already arrived at the destination, it may select one human agent HA based on the costs received from each of the human agents HA.

[0052] In addition, when the robot agent RA is moving, the costs received from each of the human agents HA and the costs (distance cost, time cost) for the automatic transport robot AMRa itself to move to the location of the shelf that is the destination may be calculated, and one human agent HA may be selected based on the calculated cost of the robot agent RA and the costs received from each of the human agents HA. When calculating the cost of the robot agent RA, the human agent HA can be selected taking into consideration the time and distance required for the automated transport robot AMRa itself to travel to the shelf position that is the destination. For example, if the route that the automated transport robot AMRa travels has a narrow passageway in part and it passes through a section where it must slow down, the time cost is calculated taking into account the slowdown in the travel speed. In such a case, the distance cost remains the same, but the time cost increases.

[0053] When a human agent HA is selected based on the cost of the robot agent RA itself and the costs received from each of the human agents HA, if the human agent HA arrives at the destination before the automatic transfer robot AMRa equipped with the robot agent RA arrives, the worker carrying the human agent HA will need to wait at the destination until the automatic transfer robot AMRa arrives. Therefore, a human agent HA that can arrive as close as possible to the timing of the arrival of the automatic transfer robot AMRa or after the arrival of the automatic transfer robot AMRa may be selected. This makes it possible to minimize the waiting time of the worker carrying the human agent HA.

[0054] Next, when a human agent HA is selected, the robot agent RA transmits a picking instruction to the selected human agent HA (step S206). Here, the picking instruction is transmitted to, for example, a human agent HA of the navigation system NVGa among the multiple navigation systems NVGa.

[0055] When the human agent HA of the navigation system NVGa receives the picking instruction, the human agent HA stores the received picking instruction in the self-information storage unit HA and then displays the picking instruction (step S303). By storing the picking instruction in the self-information storage unit HA, the "next destination" stored in the self-information storage unit HA is updated to the destination included in the currently received picking instruction. Then, the human agent HA displays the picking instruction received by the picking planning unit H16 on a liquid crystal display device mounted on the human agent HA. For example, the picking planning unit H16 displays the position of the shelf that is the next destination, the picking contents, etc. This allows the worker carrying the human agent HA to understand the picking instruction, the position of the shelf that is the next destination, the picking contents, etc., and move to the next destination. When displaying information based on the picking instruction, the human agent HA may display a movement route to a position where the shelf that is the next destination is located as route guidance information. When displaying route guidance information, the human agent HA may display the reason for the route planning together with the movement route. For example, the reason for the route planning may be "because the route with the shortest distance is congested" when a route different from the route with the shortest distance is displayed as the movement route. Such a reason for the route planning may be calculated using a technology or method classified as Explainable AI (Artificial Intelligence).

[0056] After confirming the picking instruction, the worker moves to the location of the shelf that is the destination specified by the picking instruction. At this time, the human agent HA detects the current location at regular intervals using the sensor H11, and determines whether the current location is within a certain range from the destination. If the current location is not within the certain range from the destination and the current location has been changed, the human agent HA rewrites the current state in the self-information storage unit H12 to "on the move". Then, when the picking plan unit H16 detects that the current location is within the certain range from the destination, it detects that the worker has arrived at the destination (step S304). Next, when the picking planner H16 detects that the worker has arrived at the destination, it updates the current status in the self-information storage unit H12 by rewriting it from "moving" to "working". When the current status is updated, the agent communication unit H14 is notified of the status "working" to each of the robot agents RA based on the arrival of a predetermined communication timing (e.g., the passage of a certain period of time) (step S305), and also notifies each of the human agents HA of the status "working" (step S306). This allows the robot agent RA and the other human agents HA to recognize that the picking work has been started by the worker carrying the navigation system NVGa.

[0057] Next, the picking plan unit H16 detects that the worker has started picking (step S307). The picking plan unit H16 may detect whether the work has started by, for example, detecting that the barcode of the product for which the picking instruction has been issued has been read by the barcode reader BR. Alternatively, the picking plan unit H16 may detect that the work has started when it detects that an operation button indicating the start of the picking work among the operation buttons provided on the automatic transport robot AMRa has been pressed by the worker and a notification indicating that the work has started is received from the agent communication unit R14 of the robot agent RA by the agent communication unit H14 of the human agent HA of the navigation system NVGa. This notification of the start of the work may be sent between the request communication unit R15 and the request communication unit H15.

[0058] The worker places the product picked from the shelf in the storage unit of the automatic transport robot AMRa. When picking up multiple products from one shelf, the products are taken out of the shelf and stored in the storage unit. When the picking planner H16 of the human agent HA detects that the picking has been completed (step S308), it transmits a notice of the completion of the picking to the robot agent RA (step S309). The picking planner H16 may detect that the picking has been completed when it detects that all of the products to be picked from the shelf at the current position have been read by the barcode reader BR. Alternatively, the picking planner H16 may detect that the work has been completed when it detects that an operation button indicating the completion of the picking work, among the operation buttons provided on the automatic transport robot AMRa, has been pressed by the worker, and a notice indicating the completion of the work has been sent from the robot agent RA to the human agent HA.

[0059] When the robot agent RA receives the picking completion notification, it extracts the next product to be picked from the multiple products included in the picking list, and identifies the shelf on which the extracted product is stored (step S207). Thereafter, the above-mentioned processes from step S to step S308 are repeatedly executed until picking is completed for each of the products included in the picking list. Then, when the robot agent RA detects that picking of each product included in the picking list has been completed (step S208), it moves to the packing space (step S209). As a result, the products stored in the automatic transport robot AMRa are taken out and packed by workers waiting in the packing space.

[0060] According to the embodiment described above, the picking of products included in the picking list can be carried out by workers carrying human agents HA selected according to cost.

[0061] Here, in step S206, the robot agent RA transmits a picking instruction to the human agent HA, and the worker picks the items based on the picking instruction. However, there are cases where the worker cannot respond to the picking instruction or takes time to respond. In such cases, if the worker waits for the arrival of the picking destination, the time until the picking operation is completed will be prolonged. In such a case, the time taken to complete the picking operation can be reduced by performing the following process.

[0062] Fig. 5 is a flow chart for explaining the operation of a picking support system including a plurality of human agents HA and a plurality of robot agents RA. In Fig. 5, steps S101 to S206 in Fig. 4 are executed, and a case will be explained in which a target worker to whom a picking instruction is sent in step S206 does not comply with the picking instruction. The picking plan unit H16 of the robot agent RA judges whether or not the selected worker may not be able to respond to the picking instruction transmitted in step S206. For example, when judging whether or not the selected worker may not be able to respond, the picking plan unit H16 judges whether or not the selected worker may not be able to respond by judging whether or not a notification of the status "working" has been received from the human agent HA to which the picking instruction was transmitted within a predetermined time (step S220). If the picking plan unit H16 does not receive a notification of the status "working" within a predetermined time (step S221), it judges that the selected worker may not be able to respond, and if it receives a notification of the status "working" within the predetermined time, it judges that the selected worker may not be able to respond (response has already started).

[0063] In addition, when the picking plan unit H16 of the automatic transport robot AMRa determines whether the selected worker may not be able to handle the job, it may determine that the selected worker may not be able to handle the job if the picking plan unit H16 has received a status of "working" from the human agent HA to which the picking instruction was sent, but the started picking is based on a picking instruction sent from another automatic transport robot AMR. In other words, if a worker carrying the human agent HA to which the picking instruction was sent starts another picking operation, the status "working" is sent to each robot agent RA, so that it is possible to grasp that a picking operation based on a picking list different from the picking instruction sent this time has started. Therefore, the current picking operation cannot be started until the other picking operation is completed, so that it is determined that the selected worker may not be able to handle the job.

[0064] When the picking plan unit H16 determines that the selected worker may not be able to handle the picking, it transmits the picking request transmitted in step S203 to each of the human agents HA (step S222). This allows each worker in the warehouse to be called again to see whether or not they can handle the picking. When a picking request is sent from the robot agent RA, each of the human agents HA receives the picking request.

[0065] Here, when the human agent HA of the navigation system NVGa receives a picking request, the picking instruction already received in step S206 is stored in the self-information storage unit HA, so the human agent HA deletes the stored picking instruction (step S320). In association with this deletion, the destination stored as the "next destination" is also deleted. This makes it possible to release the state in which the picking instruction must be responded to, and to prevent multiple human agents HA from responding to the picking instruction in duplicate. In addition, here, the received picking request includes an instruction number that identifies the task that is the picking request. In addition, the picking instruction is a picking instruction generated based on the picking request, and although the picking instruction includes an instruction number, the instruction number included in the picking request that generated the picking instruction is used. Therefore, whether or not the picking request and the picking instruction belong to the same task can be determined by referring to the instruction number, and if the instruction numbers match, it can be determined that they belong to the same task. Therefore, if the instruction numbers of the received picking instruction and the picking request match, the human agent HA deletes the picking instruction.

[0066] Furthermore, when the human agent HA of the navigation system NVGa receives a picking request, it performs a cost calculation (step S321) and transmits the calculated cost value to the robot agent RA (step S322). Furthermore, when the human agent HA of the navigation system NVGb receives the picking request, it performs cost calculation (step S420) and transmits the calculated cost value to the robot agent RA (step S421).

[0067] Here, each of the human agents HA of the navigation systems NVGa and NVGb recalculates the cost based on the status at the time of receiving the picking request transmitted in step S222 and the current location of the human agent HA. This allows the navigation systems NVGa and NVGb to calculate and transmit the cost according to the state at the time of receiving the picking request again. For example, if the current location of the worker of the navigation system NVGb has changed, if another task has started, if the picking task that was being performed has ended, or if there is any change in the situation, the cost after these situations are reflected can be collected.

[0068] The robot agent RA of the automatic transport robot AMRa, which is the sender of the picking request, selects one human agent HA to handle the picking request based on the costs received from each human agent HA (step S223). Here, since the recalculated costs can be obtained in steps S321 and S420, it is possible to select a worker who is preferable in the situation at the time when the cost is recalculated, taking into consideration the change in the situation according to the elapsed time since the previous cost calculation.

[0069] After step S224, if the selected worker cannot respond to this picking instruction, the process is repeated from the picking request transmission process in step S222. As a result, even if the selected worker cannot reach the shelf position, the robot agent RA does not continue to wait, but can reselect another worker, taking into account the cost at that time. This makes it possible to always improve the work efficiency in the warehouse by taking into account the worker's situation. In addition, it is possible to instruct the human agent HA on the optimal route at that time, and to optimize the picking work.

[0070] According to the embodiment described above, it is possible to navigate a picking route in real time while taking into account the behavior of the worker. Even if there are multiple workers in one warehouse, it is possible to navigate a picking route in real time while taking into account the behavior of each worker. In addition to planning routes for workers, routes for automatic transport robots may also be planned. As algorithms used for planning routes for both workers and automatic transport robots, for example, algorithms of classical computers such as nearest neighbor method, A*, genetic algorithm, ant colony optimization, and deep reinforcement learning algorithms may be used, or quantum algorithms of quantum computers may be used.

[0071] In the above-described embodiment, the working area is a warehouse WH, but the working area may be an area where picking is performed. For example, in an online supermarket where products ordered by consumers via a network such as the Internet are picked from the product display floor of a real store, the working area may be a store instead of a warehouse.

[0072] The agent communication unit H14, the request communication unit H15, and the picking plan unit H16 of the human agent HA in the above embodiment may be configured by a processing unit such as a CPU (Central Processing Unit) or a dedicated electronic circuit. Furthermore, the agent communication unit R14, the request communication unit R15, and the picking planning unit R16 of the robot agent RA may be configured, for example, by a processing device such as a CPU or a dedicated electronic circuit. The self-information memory unit H12 and agent information memory unit H13 of the human agent HA, and the self-information memory unit R12 and agent information memory unit R13 of the robot agent RA may be configured using a storage medium, such as a hard disk drive (HDD), flash memory, an electrically erasable programmable read only memory (EEPROM), a random access read / write memory (RAM), a read only memory (ROM), or any combination of these storage media. The self information storage unit H12, the agent information storage unit H13, the self information storage unit R12, and the agent information storage unit R13 may be implemented using, for example, a non-volatile memory.

[0073] The human agent HA and the robot agent RA in the above-mentioned embodiment may be realized by a computer. In that case, a program for realizing the function may be recorded in a computer-readable recording medium, and the program recorded in the recording medium may be read into a computer system and executed to realize the function. The term "computer system" as used herein includes hardware such as an OS and peripheral devices. The term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into a computer system. The term "computer-readable recording medium" may also include a medium that dynamically holds a program for a short period of time, such as a communication line when a program is transmitted via a network such as the Internet or a communication line such as a telephone line, and a medium that holds a program for a certain period of time, such as a volatile memory inside a computer system that is a server or client in that case. The above-mentioned program may be a program for realizing a part of the above-mentioned function, or may be a program that can realize the above-mentioned function in combination with a program already recorded in the computer system, or may be a program that is realized using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0074] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and designs that do not deviate from the gist of the present invention are also included. [Explanation of symbols]

[0075] 1. Host system H11…Sensor H12…Self information storage unit H13…Agent information storage section H14…Agent Communication Department H15…Request Communication Department H16…Picking Planning Department R11…Sensor R12…Self information storage unit R13: Agent information storage section R14…Agent Communication Department R15: Request communication section R16…Picking Planning Department AMR: Automatic transport robot BR: Barcode reader HA...Human Agent NVG: Navigation system PS…Packing space RA...Robot Agent RCS...Robot Control System T…Shelf S... Picking system

Claims

1. a robot management device mounted on the robot and managing the robot; a worker management device provided for each of a plurality of workers performing picking work in a work area; A picking support system including: The robot management device includes: a first transmission unit that transmits a picking instruction to a worker management device carried by the worker selected as the worker performing the next picking operation; a determination unit that determines whether or not the selected worker is likely to be unable to respond to the picking instruction after the picking instruction is transmitted; a second transmission unit that transmits, in response to the determination that the picking instruction may not be able to be handled, a request to each of the worker management devices to inquire whether the picking instruction can be handled, and a receiving unit that receives cost values ​​transmitted from each of the plurality of worker management devices that have received the request; a selection unit that selects a worker based on each of the cost values ​​received from each of the worker management devices; a third transmission unit that transmits a picking instruction to the worker management device of the selected worker; having The worker management device includes: A receiving unit for receiving a request; a cost calculation unit that calculates a cost value based on at least one of a distance cost based on a travel distance required to handle the picking based on the request and a time cost based on a time required to handle the picking based on the request based on a situation at the time the request is received; a transmission unit that transmits the calculated cost to a robot management device that is a sender of the request; A picking support system having the above structure.

2. The determination unit of the robot management device When a reference time has elapsed since the first transmission unit transmitted the picking instruction, it is determined that the selected worker may not be able to respond to the picking instruction. The picking support system according to claim 1 .

3. The determination unit of the robot management device After the first transmission unit transmits the picking instruction, when a response indicating that the picking instruction cannot be handled is received from the worker management device that received the picking instruction, it is determined that the selected worker may not be able to handle the picking instruction. The picking support system according to claim 1 .

4. The determination unit of the robot management device The determination is made in either case of a case where a picking instruction is transmitted by the first transmission unit and a case where a picking instruction is transmitted by the third transmission unit. The picking support system according to any one of claims 1 to 3.

5. The worker management device includes: A storage unit that stores the picking instruction in response to receiving the picking instruction; a deletion processing unit that deletes the picking instruction from the storage unit when the request is received after the picking instruction is stored. The picking support system according to claim 1 .

6. a robot management device mounted on the robot and managing the robot; a worker management device provided for each of a plurality of workers performing picking work in a work area; A picking support method executed in a picking support system including: The robot management device transmits a picking instruction to a worker management device carried by the worker selected to perform the next picking operation, After transmitting the picking instruction, the robot management device determines whether or not there is a possibility that the selected worker cannot respond to the picking instruction, the robot management device transmits a request to each of the worker management devices inquiring whether or not the robot management device can respond to the picking instruction in response to the determination that the robot management device may not be able to respond to the picking instruction; Each of the worker management devices receives the request, each of the worker management devices calculates a cost value based on at least one of a distance cost based on a travel distance required to handle the picking based on the request and a time cost based on a time required to handle the picking based on the request, based on a situation at the time of receiving the request; Each of the worker management devices transmits the calculated cost to the robot management device that is the sender of the request; the robot management device receives cost values ​​transmitted from each of the worker management devices that have received the request; The robot management device selects a worker based on each cost value received from each worker management device; The robot management device transmits a picking instruction to the worker management device of the selected worker. Picking assistance methods.

Images