JP2022015592A - Cargo handling system - Google Patents

Abstract

To provide a cargo handling system capable of preventing the occurrence of cargo collapsing even when load stacking is not appropriate in a pallet.SOLUTION: This cargo handling system comprises: a data collection unit 30 that collects load data relating to the appearance feature of a load, stop-time load inclination data relating to inclination of the load during a traveling stop, and traveling-time load inclination data relating to inclination of the load during traveling; a model generation unit 31 that generates an estimation model through machine learning carried out with the load data, the stop-time load inclination data and the traveling-time load inclination data which have been collected as teacher data; a data acquisition unit 33 that acquires the load data and the stop-time load inclination data which become input data; an estimation unit 34 that estimates the inclination of the load during the traveling by inputting the acquired input data to the estimation model; and a control unit 35 that performs predetermined load collapsing prevention control when the estimation result shows the high risk of load collapsing.SELECTED DRAWING: Figure 1

Description

The present invention relates to a cargo handling system including a forklift that conveys a pallet loaded with cargo according to an operator's operation.

Conventionally, a forklift has been used to transport a load loaded on a pallet in a facility such as a warehouse. The forklift is configured to perform traveling and cargo handling operations according to the operator's operation. Specifically, the forklift consists of a vehicle body having drive wheels, a pair of left and right straddle legs protruding forward from the vehicle body, a mast that moves back and forth along the straddle legs, and a pair of left and right that moves up and down along the mast. (See, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2001-354394

By the way, a plurality of packages of the same type or different types may be loaded on a pallet. A skilled operator can determine whether the stacking method is appropriate or inappropriate, and if it is inappropriate, take measures such as making the running operation slower than usual. However, new operators cannot determine whether the loading method is appropriate or inappropriate, and as a result, the improperly loaded load may be unloaded by the usual driving operation.

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 preventing the occurrence of cargo collapse even if the loading method of cargo on the pallet is inappropriate.

In order to solve the above problems, the cargo handling system according to the present invention is a cargo handling system provided with a fork lift that supports and transports a pallet loaded with luggage with a fork according to an operation of an operator, and is an appearance of the luggage. Data collection unit that collects luggage data related to features, load bias data during stoppage regarding the bias of the weight of luggage acting on the fork when traveling, and load bias data during traveling regarding the weight bias of luggage acting on the fork during traveling. A model generator that generates an estimation model by machine learning using the collected baggage data, stop load bias data, and running load bias data as teacher data, and load data and stop load bias data as input data. The data acquisition unit to be acquired, the estimation unit that estimates the bias of the weight of the luggage during traveling by inputting the acquired input data to the estimation model, and the estimation result by the estimation unit have a high risk of load collapse. It is provided with a control unit that performs predetermined load collapse prevention control when the above is indicated.

The data collection unit of the cargo handling system includes, for example, a camera for collecting luggage data and a plurality of load sensors for collecting stop load bias data and running load bias data, and includes a right side fork and a left side fork. It is possible to take the configuration that one load sensor is provided near each of the root and the tip of each of the above.

The data acquisition unit of the cargo handling system includes, for example, a camera for acquiring baggage data and a plurality of load sensors for acquiring stop load bias data, and is located near the roots of the right and left forks, respectively. It can be configured that one load sensor is provided near the tip.

The cargo data of the cargo handling system may include the contour shape and height of the cargo, and may further include the total number of cargoes loaded.

The forklift of the cargo handling system may include a speaker capable of outputting a warning sound to the operator. In this case, the control unit can operate the speaker as a load collapse prevention control.

The forklift of the cargo handling system may include a drive system for driving a traveling motor or a traveling engine. In this case, the control unit can give a command to the drive system so that the acceleration of traveling is limited as a load collapse prevention control.

According to the present invention, it is possible to provide a cargo handling system capable of preventing the occurrence of cargo collapse even if the loading method of cargo on the pallet is inappropriate.

It is a block diagram of the cargo handling system which concerns on 1st Embodiment of this invention. It is (A) plan view and (B) side view of the forklift provided in the cargo handling system which concerns on 1st Embodiment. It is a perspective view of a pallet loaded with luggage. It is a figure which shows the example of the baggage data collected or acquired in the cargo handling system which concerns on 1st Embodiment. It is a figure which shows the example of the load bias data at the time of stop collected or acquired in the cargo handling system which concerns on 1st Embodiment. It is a figure which shows the example of the load deviation data at the time of running collected in the cargo handling system which concerns on 1st Embodiment. It is a block diagram of the cargo handling system which concerns on 2nd Embodiment of this invention. It is (A) plan view and (B) side view of the forklift provided in the cargo handling system which concerns on 2nd Embodiment.

Hereinafter, the first embodiment and the second embodiment of the cargo handling system according to the present invention will be described with reference to the accompanying drawings.

[First Example]
FIG. 1 shows a cargo handling system according to the first embodiment of the present invention. As shown in the figure, the cargo handling system according to the present embodiment includes a plurality of forklifts 10a, 10b, 10c ... Having the same configuration, and an arithmetic unit configured to be able to exchange data with these. It includes a model generation unit 31. Data may be exchanged by wireless / wired communication, or may be performed via a recording medium such as a memory card.

The forklift 10a includes a data collection unit 30, a model storage unit 32, a data acquisition unit 33, an estimation unit 34, a control unit 35, and a warning unit 36.

Further, as shown in FIG. 2, the forklift 10a has a vehicle body 11 having a driver's cab 12, a pair of left and right straddle legs 13R, 13L protruding forward from the vehicle body 11, and front and rear along the straddle legs 13R, 13L. It further includes a mast 14 that moves to, a lift bracket 15 that moves up and down along the mast 14, a pair of left and right forks 16R, 16L, and a head guard 17 provided above the vehicle body 11. A drive wheel (rear wheel) 26 is provided at the lower part of the vehicle body 11, and a driven wheel (front wheel) 27R / 27L is provided at the lower part of the straddle legs 13R and 13L. Note that FIG. 2A omits the illustration of the head guard 17.

FIG. 3 shows an example of the pallet 100 and the cargo 200 carried by the forklift 10a. As shown in the figure, the pallet 100 includes an upper plate 101 and a lower plate 102, and a right side support member 103, a central support member 105, and a left side support member 104 provided between them. As a result, a space 106R for inserting the fork 16R is formed between the support members 103 and 105, and a space 106L for inserting the fork 16L is formed between the support members 104 and 105.

As shown in FIG. 2, the forklift 10a includes a camera 20 provided at the upper end of the lift bracket 15, a root load sensor 21R, 21L provided near the roots of the forks 16R, 16L, and a fork. One tip load sensor 22R, 22L provided near the tip of each of 16R, 16L, a speaker 23, a traveling motor 24 connected to a drive wheel 26, and a drive system for driving the traveling motor 24. It also has 25.

The image pickup surface of the camera 20 is directed forward. The camera 20 constitutes the above-mentioned data collection unit 30 and data acquisition unit 33, and collects and acquires images of the luggage 200 and the pallet 100 to be transported. In this embodiment, the data obtained from this image is used as the teacher data TDa or the input data IDa.

The load sensor 21R constitutes the data acquisition unit 30 and the data acquisition unit 33 described above, and is configured to detect the weight of the load 200 (and the pallet 100) acting near the root of the fork 16R. Similarly, the other load sensors 21L, 22R, 22L are also configured to detect the weight of the load 200 (and pallet 100) acting on a particular point on the forks 16R, 16L. By integrating the detection results of the four load sensors 21R, 21L, 22R, 22L, it is possible to generate data indicating the load bias, and in this embodiment, this data is used as the teacher data TDa or the input data IDa. do.

The speaker 23 is provided in the vicinity of the head guard 17. The speaker 23 constitutes the above-mentioned warning unit 36, and is configured to output a warning sound to the operator boarding the driver's cab 12.

The traveling motor 24 and the drive system 25 are housed in the vehicle body 11 together with the model storage unit 32, the estimation unit 34, and the control unit 35 described above. The model storage unit 32, the estimation unit 34, and the control unit 35 are composed of a microprocessor (MPU, Micro Processing Unit) and a memory device associated therewith.

As described above, the forklifts 10b, 10c ... Have the same configuration as the forklift 10a.

The cargo handling system according to this embodiment includes (1) a first step in which the data collecting unit 30 of each forklift 10a, 10b, 10c ... Collects teacher data TDa, TDb, TDc ..., And (2) a model. The second step in which the generation unit 31 generates an estimation model by machine learning using the teacher data TDa, TDb, TDc ..., And (3) the data acquisition unit 33 of each forklift 10a, 10b, 10c ... Are input. The third step of acquiring the data IDa, IDb, IDc ..., And (4) the estimation unit 34 of each forklift 10a, 10b, 10c ... The fourth step of estimating the risk of load collapse based on the risk, and (5) the control unit 35 of each forklift 10a, 10b, 10c ... Performs predetermined load collapse prevention control based on the estimated risk. It is configured to prevent the load from collapsing by executing the fifth step. Hereinafter, each step will be described in detail.

(First step)
In the first step, the data collection unit 30 of the forklift 10a collects the teacher data TDa. The teacher data TDa includes luggage data, stop load bias data, and running load bias data.

The baggage data includes the contour shape S and the height H of the entire baggage 200 analytically obtained from the image generated by the camera 20 (see FIG. 4). The absolute value of the height H can be determined by comparison with the known height of the pallet 100. The luggage data is preferably collected when the luggage 200 to be transported comes in front of the camera 20 (that is, immediately before inserting the forks 16R and 16L into the pallet 100 on which the luggage 200 to be transported is loaded).

The load bias data at the time of stopping is the load sensors 21R, 21L, 22R immediately after the pallet 100 loaded with the load 200 to be transported is supported by the forks 16R, 16L (that is, when the forklift 10a is stopped). , 22L outputs W1R, W1L, W2R, W2L are included (see FIG. 5). In each figure of FIG. 5, the size of the circle indicates the magnitude of the detected load.

When one or more luggage 200s are loaded on the pallet 100 in a well-balanced manner, the outputs W1R, W1L, W2R, W2L of the load sensors 21R, 21L, 22R, 22L become equal as shown in FIG. 5 (A). .. On the other hand, when the balance is not good, as shown in FIGS. 5B and 5C, there is a difference in the outputs W1R, W1L, W2R and W2L of the load sensors 21R, 21L, 22R and 22L. In other words, if the balance is not good, the load acting on the forks 16R and 16L will be biased. For example, in the state shown in FIG. 5B (W1R = W1L <W2R = W2L), the luggage 200 is more likely to collapse forward than in the state shown in FIG. 5A (W1R = W1L = W2R = W2L). I can say.

The running load bias data is the output W1R, W1L of the load sensors 21R, 21L, 22R, 22L when the forklift 10a is run while the pallet 100 loaded with the load 200 to be transported is supported by the forks 16R, 16L. , W2R, W2L are included (see FIG. 6). In each figure of FIG. 6, the size of the circle indicates the magnitude of the detected load.

The load bias data during running changes from moment to moment during running. For example, even when the luggage 200 is loaded in a well-balanced manner (that is, when the load bias data at the time of stopping as shown in FIG. 5A is obtained), when the forklift 10a starts to move forward, the forklift 10a is on the root side due to inertia. The outputs W1R and W1L are relatively large (see FIG. 6A). Further, even when the cargo 200 is loaded in a well-balanced manner, if the forklift 10a decelerates while moving forward, the outputs W2R and W2L on the tip side become relatively large due to inertia (see FIG. 6B). Similarly, when the forklift 10a turns to the left while moving forward, the outputs W1R and W2R on the right side, which are outside the turn due to inertia, become relatively large (see FIG. 6C).

The running load bias data can be collected, for example, at a predetermined frequency (for example, once / second) during a test run in which the vehicle travels on a predetermined route at a predetermined speed. It is preferable that the test run imitates the actual transportation of the luggage 200.

The data collection unit 30 collects baggage data, stop load bias data, and running load bias data for the same baggage 200 in association with each other. For example, the data collecting unit 30 collects the luggage data of the loaded luggage 200 as shown in FIG. 4A in association with the stopped load bias data and the traveling load bias data of the luggage 200.

In this step, like the forklift 10a, the data collecting unit 30 of the other forklifts 10b, 10c ... Also collects the teacher data TDb, TDc ... The teacher data TDb, TDc ... Also include baggage data, load bias data during stop, and load bias data during running.

(Second step)
In the second step, the model generation unit 31 generates an estimation model by machine learning using the teacher data TDa, TDb, TDc ... Collected in each forklift 10a, 10b, 10c ... The model generation unit 31 may use all the collected teacher data TDa, TDb, TDc ..., Or may use only a part (for example, teacher data TDa), but as much teacher data as possible may be used. It is preferable to use. The estimated model generated by the model generation unit 31 is stored in the model storage unit 32 of each forklift 10a, 10b, 10c ....

(Third step)
In the third step, the data acquisition unit 33 of the forklift 10a acquires the input data IDa. The input data IDa includes baggage data and load bias data when stopped. The data acquisition unit 33 acquires the package data of the package 200 to be transported and the load bias data at the time of stop in the same manner as the data collection unit 30.

In this step, like the forklift 10a, the data acquisition unit 33 of the other forklifts 10b, 10c ... Also acquires the input data IDb, IDc ... The input data IDb, IDc ... Also include the baggage data and the load bias data at the time of stopping.

(4th step)
In the fourth step, the estimation unit 34 of each forklift 10a, 10b, 10c ... (Representatively, the forklift 10a will be described below. The same applies hereinafter) stores the input data IDa acquired by the data acquisition unit 30 as a model storage unit. By inputting to the estimation model stored in 32, the bias of the load that will occur in the transportation of the load 200 to be performed, that is, the risk of load collapse is estimated.

As described above, the input data IDa is composed of the luggage data and the load bias data at the time of stopping, and does not include the load bias data at the time of traveling. This means that the load bias (risk level) during running can be estimated before the running.

(Fifth step)
In the fifth step, the control unit 35 of the forklift 10a operates the speaker 23 as the warning unit 36 based on the degree of danger estimated by the estimation unit 34. For example, the control unit 35 does not output a warning sound from the speaker 23 when the risk level does not reach the predetermined first threshold value, and the risk level is the first threshold value and the predetermined second threshold value (however, however). If it is between the second threshold value and the first threshold value, a relatively low volume warning sound is output from the speaker 23, and if the risk level exceeds the second threshold value, a relatively loud warning sound is output from the speaker 23. Is output. This allows the operator to know how inappropriate the loading of the luggage 200 is.

[Second Example]
FIG. 7 shows a cargo handling system according to the second embodiment of the present invention. As shown in the figure, the cargo handling system according to the present embodiment is different from the first embodiment in that it is composed of one forklift 10d and that the forklift 10d is provided with the model generation unit 31. ing.

As shown in FIG. 8, the model generation unit 31 is housed in the vehicle body 11 together with the model storage unit 32, the estimation unit 34, and the control unit 35. The model generation unit 31, the model storage unit 32, the estimation unit 34, and the control unit 35 are composed of a microprocessor and a memory device associated therewith.

In the cargo handling system according to this embodiment, (1) the first step in which the data acquisition unit 30 of the forklift 10d collects the teacher data TDd, and (2) the model generation unit 31 of the forklift 10d uses the teacher data TDd for machine learning. The second step of generating the estimation model by The fourth step of estimating the risk of load collapse based on the above, and (5) the fifth step of performing predetermined load collapse prevention control based on the estimated risk by the control unit 35 of the forklift 10d. By doing so, it is configured to prevent the load from collapsing.

Like the teacher data TDa, the teacher data TDd includes baggage data, stop load bias data, and running load bias data. Further, the input data IDd includes baggage data and load bias data at the time of stop, similarly to the input data IDa and the like.

[Modification example]
Although the first embodiment and the second embodiment of the cargo handling system according to the present invention have been described above, the configuration of the present invention is not limited thereto.

For example, the baggage data to be the teacher data and the input data may further include the stacking number of the baggage 200 loaded on the pallet 100. This is expected to enable more accurate estimation of the risk of cargo collapse. The number of stacked luggage 200 in FIG. 3 is 2, and the number of stacked luggage 200 in FIG. 4 is 4.

Further, the drive system 25 for driving the traveling motor 24 may be included in the warning unit 36. In this case, the control unit 35 prevents the load from collapsing by giving a command to the drive system 25 so that the traveling acceleration of the forklifts 10a, 10b, 10c, and 10d is limited according to the estimated risk level. Can be done.

Further, the warning unit 36 may be a device that gives a visual warning. As such a device, for example, a warning light whose lighting state changes under the control of the control unit 35, a display whose display content changes under the control of the control unit 35, and the like can be considered.

Further, the forklifts 10a, 10b, 10c and 10d may be engine type forklifts. That is, in the forklifts 10a, 10b, 10c, 10d, a traveling engine provided in place of the traveling motor 24 may be driven by the drive system 25.

10a, 10b, 10c, 10d Forklift 11 Vehicle body 12 Driver's cab 13R, 13L Straddle leg 14 Mast 15 Lift bracket 16R, 16L Fork 17 Head guard 20 Camera 21R, 21L Root load sensor 22R, 22L Tip load sensor 23 Speaker 24 Traveling motor 25 Drive system 26 Drive wheels 27R, 27L Driven wheels 30 Data collection unit 31 Model generation unit 32 Model storage unit 33 Data acquisition unit 34 Estimating unit 35 Control unit 36 Warning unit 100 Pallet 200 Luggage

Claims (7)

It is a cargo handling system equipped with a forklift that supports and transports pallets loaded with cargo with forks according to the operation of the operator.
Luggage data relating to the appearance characteristics of the luggage, load bias data at the time of stopping regarding the bias of the weight of the luggage acting on the fork when traveling is stopped, and load bias data regarding the weight bias of the luggage acting on the fork during traveling. A data collection unit that collects load bias data,
A model generation unit that generates an estimation model by machine learning using the collected baggage data, the stopped load bias data, and the running load bias data as teacher data.
A data acquisition unit that acquires the baggage data and the load bias data at the time of stopping, which are input data,
An estimation unit that estimates the bias of the weight of the cargo during traveling by inputting the acquired input data into the estimation model, and an estimation unit.
A control unit that performs predetermined load collapse prevention control when the estimation result by the estimation unit indicates that the risk of load collapse is high.
A cargo handling system characterized by being equipped with. The data collecting unit includes a camera for collecting the baggage data, and a plurality of load sensors for collecting the stopped load bias data and the traveling load bias data.
The fork consists of a right fork and a left fork.
The cargo handling system according to claim 1, wherein one load sensor is provided near the root and near the tip of the right fork and the left fork, respectively. The data acquisition unit includes a camera for acquiring the baggage data and a plurality of load sensors for acquiring the load bias data at the time of stopping.
The fork consists of a right fork and a left fork.
The cargo handling system according to claim 1 or 2, wherein one load sensor is provided near the root and near the tip of the right fork and the left fork, respectively. The cargo handling system according to claim 2 or 3, wherein the baggage data includes the contour shape and height of the baggage. The cargo handling system according to claim 4, wherein the cargo data further includes the number of stacked packages. The forklift includes a speaker capable of outputting a warning sound to the operator.
The cargo handling system according to any one of claims 1 to 5, wherein the control unit operates the speaker as the load collapse prevention control. The forklift includes a drive train that drives a drive motor or a drive engine.
The cargo handling according to any one of claims 1 to 5, wherein the control unit gives a command to the drive system so as to limit the acceleration of traveling as the load collapse prevention control. system.

Images