JP2021076995A - Cargo handling system - Google Patents

Abstract

To provide a cargo handling system that can perform a smooth cargo handling operation even when an unmanned work vehicle cannot travel along a notified travel route.SOLUTION: A cargo handling system 1 comprises a control device 100 and an unmanned work vehicle 200. The control device 100 comprises a communication unit 101 and a travel route determination unit 102 for outputting first route information about a travel route of the unmanned work vehicle 200. The unmanned work vehicle 200 comprises: a recognition unit 201 for acquiring current location information of a vehicle; a detection unit 202 for acquiring obstacle information; a memory unit 203; a learning model unit 204 for performing a machine learning so as to create second route information about a movable route of the vehicle; a processing unit 205 for acquiring the second route information from the learning model unit 204 by inputting the first route information and the obstacle information to the learning model unit 204; and a travel control unit 206 for controlling the travel of the vehicle based on the second route information.SELECTED DRAWING: Figure 1

Description

The present invention relates to a cargo handling system.

In the warehouse, a plurality of unmanned work vehicles (for example, a laser-guided unmanned forklift) run and handle cargo under the control of a management device. The management device determines the travel route of each unmanned work vehicle, and notifies the unmanned work vehicle of the determined travel route. The unmanned work vehicle that receives the notification travels according to the notified travel route while recognizing its own position.

The laser-guided unmanned forklift is provided with a laser scanner (see, for example, Patent Document 1). The laser scanner projects a laser to the surroundings while rotating the laser light source, and detects the reflected light from a plurality of reflectors arranged in the warehouse. The laser-guided unmanned forklift stores the position of the reflector on the map and calculates its own position based on the principle of triangulation. As a result, the laser-guided unmanned forklift truck can travel on the notified travel route while recognizing its own position.

Japanese Unexamined Patent Publication No. 8-161039

However, if there is an obstacle such as luggage on the notified travel route, the unmanned work vehicle cannot travel on the notified travel route, and there is a problem that smooth cargo handling work cannot be performed.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a cargo handling system capable of smooth cargo handling work even if the unmanned work vehicle cannot travel on the notified travel route. There is.

In order to solve the above problems, the cargo handling system according to the present invention
A cargo handling system including a management device and an unmanned work platform that travels and handles cargo under the control of the management device.
The management device is
With the communication unit that communicates with the unmanned work platform,
A travel route determination unit that determines the travel route of the unmanned work platform and outputs first route information related to the travel route via the communication unit.
With
The unmanned work vehicle
With the car body
A recognition unit that recognizes the current location of the vehicle body and acquires current location information related to the current location.
A detection unit that detects an obstacle on the traveling route and acquires obstacle information regarding the position of the obstacle.
A storage unit that stores the obstacle information and
A learning model unit that has been machine-learned to generate second route information regarding the travelable route of the vehicle body when the first route information and the obstacle information are input.
A processing unit that acquires the second route information from the learning model unit by inputting the first route information and the obstacle information stored in the storage unit into the learning model unit.
A traveling control unit that controls the traveling of the vehicle body based on the current location information and the second route information acquired by the processing unit.
It is characterized by having.

In the above cargo handling system
The processing unit notifies the management device of the second route information, and the processing unit notifies the management device.
The management device can be configured to include a display unit that displays the travel route included in the first route information and the travelable route included in the second route information.

In the above cargo handling system
The processing unit compares the travel route included in the first route information with the travelable route included in the second route information, and the matching rate between the travel route and the travelable route is equal to or less than a predetermined value. In the case of, the running stop command is output and
The travel control unit can be configured to stop the travel control based on the travel stop command.

In the above cargo handling system
The management device is
A collecting unit that collects the first route information, the second route information corresponding to the first route information, and the obstacle information to generate learning data.
A parameter update unit that performs machine learning based on the learning data, generates update data for updating the parameters of the machine learning algorithm included in the learning model unit, and transmits the update data to the learning model unit.
With
The learning model unit can be configured to update the parameters based on the update data.

In the above cargo handling system, for example
The unmanned work vehicle is a laser-guided unmanned forklift.
The recognition unit is a laser scanner provided on the upper part of the vehicle body.

According to the present invention, it is possible to provide a cargo handling system capable of smooth cargo handling work even if the unmanned work vehicle cannot travel on the notified travel route.

It is a block diagram of the cargo handling system which concerns on 1st Embodiment. It is a side view of the unmanned forklift of the laser guidance system which concerns on 1st Embodiment. It is a top view of the cargo handling system which concerns on 1st Embodiment. It is a figure which shows the display screen of the management device which concerns on 1st Embodiment, (A) is a figure when the coincidence rate of a travel route and a travelable route is high, (B) is a diagram of a travel route and a travelable route It is a figure when the concordance rate is low. It is a block diagram of the cargo handling system which concerns on 2nd Embodiment.

Hereinafter, embodiments of the cargo handling system according to the present invention will be described with reference to the accompanying drawings.

[First Embodiment]
FIG. 1 shows a block diagram of a cargo handling system 1 according to the first embodiment of the present invention. The cargo handling system 1 includes a management device 100 and at least one laser-guided unmanned forklift truck 200 (corresponding to the “unmanned work vehicle” of the present invention).

The management device 100 includes a communication unit 101, a travel route determination unit 102, and a display unit 103. As shown in FIG. 3, the management device 100 may be provided outside the facility 2 (for example, a warehouse having a plurality of shelves 3) on which the unmanned forklift 200 travels, or may be provided inside the facility 2.

The communication unit 101 is configured to perform wireless communication with the unmanned forklift 200 registered in advance in the management device 100.

The travel route determination unit 102 is configured to determine the travel route of the unmanned forklift 200 and output the first route information regarding the travel route via the communication unit 101. The travel route determination unit 102 is composed of, for example, a microcomputer, and various functions of the travel route determination unit 102 are realized by the CPU of the microcomputer executing a predetermined program or the like.

The display unit 103 is composed of, for example, a liquid crystal display. For example, traveling information and cargo handling work information of the unmanned forklift 200 are displayed on the display unit 103. As shown in FIG. 4, the display unit 103 may display the model diagram M of the facility 2 and display the current location Ma of the unmanned forklift 200, the travel route Ra, and the like as travel information in the model diagram M.

The unmanned forklift 200 is configured to perform traveling and cargo handling operations under the control of the management device 100.

As shown in FIG. 2, the unmanned forklift 200 includes a vehicle body 210, a cargo handling device 211, a laser scanner 212 provided on the upper portion of the vehicle body 210, and an obstacle sensor 213 provided on the front portion of the vehicle body 210. .. The cargo handling device 211 includes a mast provided so as to be horizontally movable in the front and rear of the vehicle body 210, and a fork provided so as to be able to move up and down the mast.

As shown in FIG. 1, the unmanned forklift 200 includes a recognition unit 201, a detection unit 202, a storage unit 203, a learning model unit 204, a processing unit 205, and a travel control unit 206.

The recognition unit 201 is configured to recognize the current location (self-position) of the vehicle body 210 and acquire the current location information regarding the current location of the vehicle body 210. Specifically, the recognition unit 201 includes a laser scanner 212 and a first calculation unit (not shown) provided inside the vehicle body 210.

As shown in FIG. 3, the laser scanner 212 projects the laser L to the surroundings while rotating the laser light source, and detects the reflected light L'from the plurality of reflectors 4 arranged in the facility 2. The first calculation unit stores the position of the reflector 4 on a predetermined map, and calculates the current location of the vehicle body 210 based on the principle of triangulation. In this way, the recognition unit 201 acquires the current location information regarding the current location of the vehicle body 210.

The detection unit 202 is configured to detect an obstacle on the traveling route and acquire obstacle information regarding the position of the obstacle. Specifically, the detection unit 202 includes an obstacle sensor 213 and a second calculation unit (not shown) provided inside the vehicle body 210.

The obstacle sensor 213 detects an obstacle (for example, a luggage placed on the traveling route) existing in front of the vehicle body 210 while the unmanned forklift 200 is traveling. The second calculation unit calculates the position of the obstacle detected by the obstacle sensor 213 based on the current location information acquired by the recognition unit 201. In this way, the detection unit 202 acquires obstacle information regarding the position of the obstacle.

The storage unit 203 is configured to store at least the obstacle information acquired by the detection unit 202. The obstacle information is appropriately updated by the detection unit 202. For example, when the baggage placed on the traveling route is removed, that is, when the baggage is not detected by the detection unit 202, the obstacle information corresponding to the baggage is deleted.

When the first route information and the obstacle information are input, the learning model unit 204 generates the second route information regarding the travelable route of the vehicle body 210 based on a machine learning algorithm such as a neural network having a predetermined parameter. It is a trained model that has been machine-learned.

In machine learning of a trained model, a large amount of teacher data, that is, a set of input data and output data is input by using a machine learning algorithm such as a neural network as described above. The input data includes data on the travel route of the unmanned forklift 200 determined by the management device 100 and data on the positions of obstacles existing on the travel route. The output data includes data on a travelable route for the unmanned forklift 200 to travel around obstacles. It can be inferred that there is a certain relationship such as a correlation between the input data and the output data.

The processing unit 205 acquires the second route information from the learning model unit 204 by inputting the first route information received from the management device 100 and the obstacle information stored in the storage unit 203 into the learning model unit 204. It is configured as. The processing unit 205 may associate the acquired second route information with the obstacle information and notify the management device 100.

When there is no obstacle information corresponding to the first route information, that is, when there is no obstacle on the traveling route included in the first route information, the processing unit 205 outputs the first route information to the traveling control unit 206. To do.

When the processing unit 205 notifies the management device 100 of the second route information and the obstacle information, the display unit 103 of the management device 100 displays the model diagram M of the facility 2 as shown in FIGS. 4A and 4B. May be displayed, and the current location Ma of the unmanned forklift 200, the obstacle Mb, the travel route Ra, and the travelable route Rb may be displayed in the model diagram M.

The travel control unit 206 controls the travel of the vehicle body 210 based on the current location information acquired by the recognition unit 201 and the second route information acquired by the processing unit 205. When the first route information is input from the processing unit 205 (when there is no obstacle on the traveling route), the traveling control unit 206 controls the traveling of the vehicle body 210 based on the current location information and the first route information. Do.

After all, in the cargo handling system 1 according to the present embodiment, when an obstacle exists on the traveling route notified from the management device 100, the learning model unit 204 generates the second route information regarding the traveling route of the vehicle body 210. The travel control unit 206 controls the travel of the vehicle body 210 based on the second route information.

Therefore, according to the cargo handling system 1 according to the present embodiment, the unmanned forklift 200 can travel on a travelable route different from the travelable route even if it cannot travel on the notified travel route, so that smooth cargo handling work can be performed. It will be possible.

[Second Embodiment]
FIG. 5 shows a block diagram of the cargo handling system 11 according to the second embodiment of the present invention. The cargo handling system 11 includes a management device 300 and a plurality of laser-guided unmanned forklifts 200. Since the configuration of the unmanned forklift 200 is the same as that of the first embodiment, the description thereof will be omitted here.

The management device 300 manages the traveling and cargo handling operations of a plurality of unmanned forklifts 200. The management device 300 is common to the management device 100 of the first embodiment except that the collection unit 301 and the parameter update unit 302 are provided.

The collecting unit 301 is configured to collect the first route information from the traveling route determining unit 102 and collect the second route information and obstacle information from the plurality of unmanned forklifts 200.

The collecting unit 301 associates the first route information with the second route information and the obstacle information corresponding to the first route information, and is for learning including the first route information, the second route information, and the obstacle information. Generate data. That is, the collecting unit 301 collects a large amount of data indicating what kind of route (travelable route) the vehicle traveled from each unmanned forklift 200 when an obstacle exists on the traveling route, and is used for learning. Generate data.

The parameter update unit 302 generates update data for updating the parameters of the machine learning algorithm (for example, the weighting parameters of the neural network) included in the learning model unit 204 of the unmanned forklift 200. The parameter update unit 302 performs machine learning based on the learning data by using a machine learning algorithm such as a neural network in order to generate the update data. The parameter update unit 302 transmits the generated update data to a plurality of unmanned forklifts 200 via the communication unit 101.

The learning model unit 204 of the unmanned forklift 200 updates the parameters of the machine learning algorithm based on the received update data. As a result, the learning model unit 204 can generate the second route information regarding the travelable route of the vehicle body 210 based on the machine learning algorithm having the updated parameters.

[Modification example]
Although the embodiment of the cargo handling system according to the present invention has been described above, the present invention is not limited to the above embodiment.

The management device of the present invention comprises a communication unit that communicates with the unmanned work vehicle, and a travel route determination unit that determines the travel route of the unmanned work platform and outputs the first route information regarding the travel route via the communication unit. If provided, the configuration can be changed as appropriate.

The unmanned work vehicle of the present invention has a vehicle body, a recognition unit that recognizes the current location of the vehicle body and acquires current location information regarding the current location, and detects an obstacle on a traveling route to acquire obstacle information regarding the position of the obstacle. A detection unit, a storage unit that stores obstacle information, and a learning model unit that is machine-learned to generate second route information regarding the travelable route of the vehicle body when the first route information and obstacle information are input. , The processing unit that acquires the second route information from the learning model unit by inputting the first route information and the obstacle information stored in the storage unit into the learning model unit, and the second route information acquired by the current location information and the processing unit. If a travel control unit that controls travel of the vehicle body based on route information is provided, the configuration can be appropriately changed.

The unmanned work vehicle of the present invention is not limited to the laser-guided unmanned forklift truck, and may include another type of unmanned forklift truck or automatic guided vehicle.

The processing unit 205 according to the first embodiment may execute a process of comparing the travel route included in the first route information with the travelable route included in the second route information. The comparison result includes, for example, the matching rate between the travel route and the travelable route. The processing unit 205 may output a travel stop command to the travel control unit 206 when the matching rate between the travel route and the travelable route is a predetermined value or less (for example, 50% or less). The travel control unit 206 stops the travel control for at least a certain period of time based on the travel stop command.

As a result, when the matching rate between the traveling route and the traveling route is low, it is possible to prevent the unmanned forklift 200 traveling according to the traveling route from hindering the traveling of another unmanned forklift. Note that FIG. 4A is a display screen when the matching rate between the traveling route Ra and the traveling route Rb is high, and FIG. 4B is a display screen when the matching rate between the traveling route Ra and the travelable route Rb is low. It is a display screen of.

The collecting unit 301 and the parameter updating unit 302 according to the second embodiment may be provided in the unmanned forklift 200 instead of the management device 300. However, from the viewpoint that a large amount of data can be acquired, it is preferable that the management device 300 includes a collection unit 301 and a parameter update unit 302.

1, 11 Cargo handling system 2 Facility 3 Shelf 4 Reflector 100, 300 Management device 101 Communication unit 102 Travel route determination unit 103 Display unit 200 Unmanned forklift 201 Recognition unit 202 Detection unit 203 Storage unit 204 Learning model unit 205 Processing unit 206 Travel control Part 210 Body 211 Cargo handling device 212 Laser scanner 213 Obstacle sensor 301 Collection part 302 Parameter update part

In order to solve the above problems, the cargo handling system according to the present invention
A cargo handling system including a management device and a plurality of unmanned work platforms that travel and carry out cargo handling work under the control of the management device.
The management device is
With the communication unit that communicates with the unmanned work platform,
A travel route determination unit that determines the travel route of the unmanned work platform and outputs first route information related to the travel route via the communication unit.
With
The unmanned work vehicle
With the car body
A recognition unit that recognizes the current location of the vehicle body and acquires current location information related to the current location.
A detection unit that detects an obstacle on the traveling route and acquires obstacle information regarding the position of the obstacle.
A storage unit that stores the obstacle information and
A learning model unit that has been machine-learned to generate second route information regarding the travelable route of the vehicle body when the first route information and the obstacle information are input.
A processing unit that acquires the second route information from the learning model unit by inputting the first route information and the obstacle information stored in the storage unit into the learning model unit.
A traveling control unit that controls the traveling of the vehicle body based on the current location information and the second route information acquired by the processing unit.
Equipped with a,
The processing unit compares the travel route included in the first route information with the travelable route included in the second route information, and the matching rate between the travel route and the travelable route is equal to or less than a predetermined value. In the case of, the running stop command is output and
The travel control unit is characterized in that the travel control is stopped based on the travel stop command.

Claims (5)

A cargo handling system including a management device and an unmanned work platform that travels and handles cargo under the control of the management device.
The management device is
With the communication unit that communicates with the unmanned work platform,
A travel route determination unit that determines the travel route of the unmanned work platform and outputs first route information related to the travel route via the communication unit.
With
The unmanned work vehicle
With the car body
A recognition unit that recognizes the current location of the vehicle body and acquires current location information related to the current location.
A detection unit that detects an obstacle on the traveling route and acquires obstacle information regarding the position of the obstacle.
A storage unit that stores the obstacle information and
A learning model unit that has been machine-learned to generate second route information regarding the travelable route of the vehicle body when the first route information and the obstacle information are input.
A processing unit that acquires the second route information from the learning model unit by inputting the first route information and the obstacle information stored in the storage unit into the learning model unit.
A traveling control unit that controls the traveling of the vehicle body based on the current location information and the second route information acquired by the processing unit.
A cargo handling system characterized by being equipped with. The processing unit notifies the management device of the second route information, and the processing unit notifies the management device.
The cargo handling system according to claim 1, wherein the management device includes a display unit that displays the travel route included in the first route information and the travelable route included in the second route information. .. The processing unit compares the travel route included in the first route information with the travelable route included in the second route information, and the matching rate between the travel route and the travelable route is equal to or less than a predetermined value. In the case of, the running stop command is output and
The cargo handling system according to claim 1 or 2, wherein the travel control unit stops the travel control based on the travel stop command. The management device is
A collecting unit that collects the first route information, the second route information corresponding to the first route information, and the obstacle information to generate learning data.
A parameter update unit that performs machine learning based on the learning data, generates update data for updating the parameters of the machine learning algorithm included in the learning model unit, and transmits the update data to the learning model unit.
With
The cargo handling system according to any one of claims 1 to 3, wherein the learning model unit updates the parameters based on the update data. The unmanned work vehicle is a laser-guided unmanned forklift.
The cargo handling system according to any one of claims 1 to 4, wherein the recognition unit is a laser scanner provided on an upper portion of the vehicle body.

Images