JP6974925B2 - Unmanned forklift and cargo handling system - Google Patents

Description

The present invention relates to an unmanned forklift and a cargo handling system that takes a load from a truck bed.

It is known that a forklift is used to perform a loading operation (unloading operation) for taking a load from a truck bed (see, for example, Patent Document 1).

An unmanned forklift is known as a type of forklift (see, for example, Patent Document 2). Generally, an unmanned forklift is programmed to raise and stop the fork to a predetermined fixed loading height (fork insertion height) when the fork is unloaded.

Japanese Unexamined Patent Publication No. 2010-235223 Japanese Unexamined Patent Publication No. 11-21098

However, if the ground clearance changes according to the load applied to the truck bed, the fork of the unmanned forklift cannot be inserted into the load loaded on the truck, that is, the unmanned forklift lifts the load from the truck bed. There was a risk that it could not be done.

The present invention has been made in view of the above circumstances, and provides an unmanned forklift and a cargo handling system capable of taking a load from the loading platform even when the ground clearance of the loading platform of a truck changes. Make it an issue.

In order to solve the above problems, the unmanned forklift according to claim 1 is an unmanned forklift that picks up a load loaded on a truck bed, and includes a fork that can be raised and lowered to take the load and the above. An imaging unit that captures an image of a truck, an estimation unit that estimates the ground clearance of the loading platform or the load based on the imaging result of the truck, and the estimation unit that estimates the ground clearance when the fork picks up the load. A control unit that controls the ascent and descent of the fork, and a ground clearance estimation model generated by machine learning using the teacher image data of the truck and the ground clearance at the time when the teacher image data is obtained as teacher data. The estimation unit includes a storage unit for storing, and the estimation unit includes the ground clearance estimation model stored in the storage unit and the ground clearance based on the current image data of the truck generated by the imaging unit. Is characterized by estimating.
The "current time" here does not mean a limited moment, and does not mean the time when the image data for the teacher is obtained (that is, the time when the ground clearance estimation model is generated), but the ground clearance. It means an arbitrary time (timing) when estimating.

Further, in order to solve the above-mentioned problem, the cargo handling system according to claim 2 takes a truck provided with a loading platform, an estimation device for estimating the ground clearance of the loading platform or the load, and the load. It is a cargo handling system including an unmanned forklift, and the estimation device includes an image pickup unit that images the truck, an estimation unit that estimates the ground clearance based on the image pickup result of the truck, and the ground clearance as ground clearance information. The unmanned forklift is provided with an information transmission unit for transmitting the estimation result of the above, a fork provided so as to be able to move up and down to take the load, an information receiving unit for receiving the ground clearance information, and the fork to load the load. The estimation device is provided with a control unit that controls the raising and lowering of the fork based on the ground clearance information, and the estimation device is provided at a position away from the truck and the unmanned forklift .

The cargo handling system according to claim 3 is the cargo handling system according to claim 2 , wherein the ground clearance information is delivered from the information transmitting unit to the information receiving unit provided in each of the plurality of unmanned forklifts. It is a feature.

The cargo handling system according to claim 4 is the cargo handling system according to claim 2 or 3 , wherein the estimation device is the above-ground height at the time when the teacher image data of the truck and the teacher image data are obtained. Further, a storage unit for storing the ground height estimation model generated by machine learning using the above as teacher data is further provided, and the estimation unit is generated by the ground height estimation model stored in the storage unit and the image pickup unit. It is characterized in that the ground height is estimated based on the image data of the truck at the present time.
The "current time" here does not mean a limited moment, and does not mean the time when the image data for the teacher is obtained (that is, the time when the ground clearance estimation model is generated), but the ground clearance. It means an arbitrary time (timing) when estimating.

According to the present invention, it is possible to provide an unmanned forklift and a cargo handling system capable of taking a load from the loading platform even when the ground clearance of the loading platform of a truck changes.

It is a schematic diagram of the cargo handling system which concerns on 1st Embodiment of this invention. It is a block diagram which shows the schematic structure of the unmanned forklift which concerns on the same embodiment. It is a schematic diagram of the cargo handling system which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the schematic structure of the estimation device and the unmanned forklift which concerns on the same embodiment. (A) is a block diagram showing a schematic configuration of an unmanned forklift according to a modified example, and (B) is a block diagram showing a schematic configuration of an estimation device according to a modified example.

(First Embodiment)
The cargo handling system 1 according to the first embodiment of the present invention will be described with reference to FIGS. 1 and 2.
As shown in FIG. 1, the cargo handling system 1 of the present embodiment includes a truck 1A loaded with one or more loads L, and one or more unmanned forklifts 1B for transporting the load L from the truck 1A to a predetermined location. It is composed of. The load L is a heavy object including the pallet P and the article G placed on the pallet P, and the pallet P has an insertion port P1 into which the fork 22 described later is inserted.

The truck 1A includes a loading platform 11 on which a load L is loaded, tires 12 and the like that support a load applied to the loading platform 11. As the tire 12 is deformed according to the weight of the load L loaded on the loading platform 11, the ground clearance of the loading platform 11 (that is, the distance from the ground surface to the loading platform 11) changes. Specifically, when the total weight of the load L loaded on the loading platform 11 is small, the loading platform 11 is high, and when the total weight of the loading platform 11 is large, the loading platform 11 is low.

The unmanned forklift 1B includes a vehicle body 21 traveling on the ground, a fork 22 provided so as to be able to move up and down to take a load L, and an image pickup unit 23 for imaging the truck 1A. The image pickup unit 23 is composed of a digital camera provided with an image pickup element.

FIG. 2 is a block diagram of the unmanned forklift 1B.
As shown in FIG. 2, the unmanned forklift 1B includes an elevating unit 24, a storage unit 25, an estimation unit 26, and a control unit 27.

The elevating part 24 is composed of an elevating device for elevating and lowering the fork 22. Specifically, the elevating portion 24 is composed of, for example, a lift bracket provided so as to be able to move up and down along the mast, a lift cylinder for raising and lowering the lift bracket, a hydraulic circuit for expanding and contracting the lift cylinder, and the like. ing.

The storage unit 25 stores in advance a ground clearance estimation model for estimating the ground clearance of the cargo bed 11 or the load L from the image data of the truck 1A. The ground height estimation model is generated by machine learning using the teacher image data of a large number of trucks 1A and the ground height of the loading platform 11 or the load L at the time when the teacher image data is obtained as teacher data. be.

The estimation unit 26 determines the ground clearance of the truck 1A based on the ground clearance estimation model stored in the storage unit 25 and the image pickup result of the truck 1A by the image pickup unit 23 (that is, the image data of the truck 1A). , The ground clearance of the load L loaded on the loading platform 11 is estimated. Specifically, for example, the estimation unit 26 analyzes the image of the truck 1A to detect the feature amount of the truck 1A, and based on the feature amount and the ground clearance estimation model, the distance from the ground surface to the loading platform 11. The ground clearance of the loading platform 11 is estimated. Further, for example, the estimation unit 26 analyzes the image of the truck 1A to detect the feature amount of the truck 1A and the load L, and based on the feature amount and the ground clearance estimation model, the height of the insertion port P1 ( The distance from the ground surface) is estimated as the ground clearance of the load L.

When the load L is taken by the fork 22, the control unit 27 controls the elevating and lowering of the fork 22 based on the estimation result of the ground clearance by the estimation unit 26. Specifically, for example, when the ground height of the loading platform 11 is estimated by the estimation unit 26, the control unit 27 is loaded on the loading platform 11 by adding a predetermined amount of height to the ground height of the loading platform 11. The height of the insertion port P1 of the pallet P is calculated, and the elevating part 24 is controlled so as to raise the fork 22 to the calculated height of the insertion port P1 and stop the fork 22. Further, for example, when the height of the insertion port P1 is estimated as the ground clearance of the load L by the estimation unit 26, the fork 22 is raised and lowered to the estimated height of the insertion port P1 and stopped. The unit 24 is controlled. That is, the control unit 27 raises the fork 22 to the height (loading height) of the insertion port P1 estimated indirectly or directly, and stops the fork 22.

An example of the operation of the unmanned forklift 1B when taking the load L will be described.
First, in the unmanned forklift 1B, the image pickup unit 23 takes a picture of the truck 1A at a position away from the truck 1A, and the estimation unit 26 estimates the ground clearance of the loading platform 11 or the load L. Next, the unmanned forklift 1B approaches the loading platform 11 of the truck 1A, raises the fork 22 to the height of the insertion port P1 based on the estimation result of the ground clearance of the loading platform 11 or the load L, and then stops the forklift. The fork 22 is inserted into the insertion port P1. Then, the unmanned forklift 1B lifts the load L with the fork 22 (that is, takes the load L) and conveys the load L to a predetermined place.

According to this embodiment, the following effects can be obtained.
(1) The ground clearance of the loading platform 11 or the load L is estimated based on the image pickup result of the truck 1A, and the raising and lowering of the fork 22 when taking the load L is controlled based on this estimation result. Therefore, even if the ground clearance of the loading platform 11 changes according to the load applied to the loading platform 11, the fork 22 is set to an appropriate loading height so that the fork 22 can be inserted into the insertion port P1. It can be raised and stopped, and the load L can be taken from the loading platform 11.

(Second Embodiment)
The cargo handling system 2 according to the second embodiment of the present invention will be described with reference to FIGS. 3 and 4. The description of the same configuration as that of the first embodiment will be omitted.

As shown in FIG. 3, the cargo handling system 2 of the present embodiment includes a truck 1A, one or more unmanned forklift trucks 1B, and at least one estimation device 1C. The unmanned forklift 1B of the present embodiment does not include the image pickup unit 23 described in the first embodiment.

The estimation device 1C is provided at a predetermined position away from the track 1A, and includes an image pickup unit 31 for photographing the track 1A. The image pickup unit 31 is composed of a digital camera provided with an image pickup element.

FIG. 4 is a block diagram of the unmanned forklift 1B and the estimation device 1C.
As shown in FIG. 4, the unmanned forklift 1B of the present embodiment includes an information receiving unit 28 for receiving ground clearance information in place of the storage unit 25 and the estimation unit 26 (see FIG. 2).

The information receiving unit 28 is composed of a communication device that directly or indirectly communicates with the estimation device 1C. The information receiving unit 28 receives the estimation result of the ground clearance of the loading platform 11 or the load L as the ground clearance information. The control unit 27 of the present embodiment controls the raising and lowering of the fork 22 based on the estimation result of the ground clearance received by the information receiving unit 28.

Further, as shown in FIG. 4, the estimation device 1C includes a storage unit 32, an estimation unit 33, and an information transmission unit 34. The estimation device 1C functions as a server that distributes ground clearance information by transmitting ground clearance information to a plurality of unmanned forklift trucks 1B.

Similar to the storage unit 25, the storage unit 32 stores in advance a ground clearance estimation model for estimating the ground clearance of the loading platform 11 or the load L from the image data of the truck 1A.

Similar to the estimation unit 26, the estimation unit 33 determines the ground clearance of the truck 1A based on the image pickup result of the truck 1A by the image pickup unit 31 and the ground clearance estimation model stored in the storage unit 32. The ground clearance of the load L loaded on the loading platform 11 is estimated.

The information transmission unit 34 is composed of a communication device that directly or indirectly communicates with the unmanned forklift 1B. The information transmission unit 34 transmits the estimation result of the ground clearance estimated by the estimation unit 33 as the ground clearance information.

The operation of the estimation device 1C and the unmanned forklift 1B will be described.
Before the unmanned forklift 1B takes the load L, the estimation device 1C photographs the truck 1A by the image pickup unit 31 and estimates the ground clearance of the loading platform 11 or the load L by the estimation unit 33. Next, the estimation device 1C transmits the ground clearance estimation result of the loading platform 11 or the load L as the ground clearance information, and the unmanned forklift 1B receives the ground clearance estimation result. Next, the unmanned forklift 1B raises the fork 22 to the height of the insertion port P1 and stops it based on the estimation result of the ground clearance of the loading platform 11 or the load L, and then inserts the fork 22 into the insertion port P1. .. Then, the unmanned forklift 1B lifts the load L with the fork 22 and conveys the load L.

According to this embodiment, the same effect as (1) above and the following effects can be obtained.
(2) The ground clearance of the loading platform 11 or the load L is estimated by the estimation device 1C, and this estimation result is received as the ground clearance information by the unmanned forklift 1B, and the fork 22 moves up and down when the load L is taken based on the ground clearance information. Is controlled. Therefore, it is not necessary to provide the image pickup unit 23 and the estimation unit 26 on the unmanned forklift 1B.

(3) Since the ground clearance information is distributed from the information transmission unit 34 to the information reception units 28 of the plurality of unmanned forklift trucks 1B, the ground clearance information used for controlling the elevation of the fork 22 is centrally distributed by the estimation device 1C. Can be managed.

The present invention is not limited to the above embodiment, and the above configuration can be changed. For example, the following changes can be made and implemented, or the following changes can be combined and implemented.

• As shown in FIG. 5 (A), the unmanned forklift 1B of the first embodiment may include a model generation unit 41, and similarly, as shown in FIG. 5 (B), estimation of the second embodiment. The device 1C may include a model generation unit 41. The model generation unit 41 generates a ground height estimation model by machine learning using the teacher image data of the track 1A and the ground height of the loading platform 11 or the load L at the time when the teacher image data is obtained as teacher data. , The storage units 25 and 32 store the ground height estimation model generated by the model generation unit 41. In this way, the ground clearance of the loading platform 11 or the load L may be estimated based on the ground clearance estimation model generated by the model generation unit 41. As the teacher image data constituting the teacher data, the image data of the track 1A captured by the imaging units 23 and 31 may be used, and the fork 22 is used as the ground height of the loading platform 11 or the loading L constituting the teacher data. The calculation result of the ground height of the loading platform 11 or the loading L based on the amount of elevation of the load may be used.

According to the above configuration, the ground clearance estimation model can be easily updated by the model generation unit 41, and the ground clearance estimation accuracy is improved by updating the ground clearance estimation model (that is, by developing machine learning). be able to.

-In the second embodiment, even if the cargo handling system 2 further includes a server that functions as a communication relay unit that relays communication between the information transmission unit 34 of the estimation device 1C and the information reception unit 28 of the plurality of unmanned forklift trucks 1B. good. According to this configuration, it is not necessary to provide the estimation device 1C with a function as a server.

-The article G constituting the load L does not have to be placed on the pallet P. That is, the insertion port into which the fork 22 is inserted may be formed in the article G, and the load L may not be composed of the pallet P.

1, 2, Cargo handling system 1A Truck 1B Unmanned forklift 1C Estimator 11 Loading platform 22 Fork 23 Imaging unit 25 Storage unit 26 Estimating unit 27 Control unit 28 Information receiving unit 31 Imaging unit 32 Storage unit 33 Estimating unit 34 Information transmission unit G Article L Load P pallet P1 outlet

Claims (4)

An unmanned forklift that picks up the load on the truck bed.
With a fork provided so that it can be raised and lowered to take the load,
An imaging unit that captures the track and
An estimation unit that estimates the ground height of the loading platform or the load based on the image pickup result of the truck, and
A control unit that controls the ascent and descent of the fork based on the estimation result of the ground clearance when taking a load with the fork .
It is provided with a storage unit for storing a ground height estimation model generated by machine learning using the teacher image data of the track and the ground height at the time when the teacher image data is obtained as teacher data.
The estimation unit is characterized in that the ground clearance is estimated based on the ground clearance estimation model stored in the storage unit and the current image data of the truck generated by the imaging unit. Unmanned forklift. A truck with a loading platform and a truck
An estimation device that estimates the above-ground height of the loading platform or the loading, and
A cargo handling system equipped with the unmanned forklift that takes the load.
The estimation device is
An imaging unit that captures the track and
An estimation unit that estimates the ground height based on the image pickup result of the truck, and an estimation unit.
It is equipped with an information transmission unit that transmits the estimation result of the ground clearance as ground clearance information.
The unmanned forklift is
With a fork provided so that it can be raised and lowered to take the load,
The information receiving unit that receives the ground clearance information,
A control unit that controls the raising and lowering of the fork based on the ground clearance information when the load is taken by the fork is provided .
The estimation device is a cargo handling system characterized in that it is provided at a position away from the truck and the unmanned forklift. The cargo handling system according to claim 2 , wherein the ground clearance information is delivered from the information transmitting unit to the information receiving unit included in each of the plurality of unmanned forklifts. The estimation device stores a storage unit that stores a ground height estimation model generated by machine learning using the teacher image data of the track and the ground height at the time when the teacher image data is obtained as teacher data. Further prepare,
The estimation unit is characterized in that the ground clearance is estimated based on the ground clearance estimation model stored in the storage unit and the image data of the truck at the present time generated by the imaging unit. The cargo handling system according to claim 2 or 3.

Images