Safety device and picking lift equipped with it
The safety device uses an imaging unit and machine learning to accurately determine belt attachment, preventing falls by ensuring the operator wears the belt and safely operates the lift.
Patent Information
- Application Number
- JP2024007157
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2024-01-22
- Publication Date
- 2026-11-11
- Estimated Expiration
- 2044-01-22
AI Technical Summary
Existing safety devices for picking lifts struggle to accurately determine whether an operator is wearing a safety belt, leading to potential falls due to erroneous detection or failure to detect the belt's position.
A safety device equipped with an imaging unit, a fitting determination unit that uses machine learning to analyze images of the operator and belt, and a decision unit to warn or control the lift's operation based on the belt's state, including a learning model to predict belt attachment and a warning unit to ensure proper belt usage.
Accurately determines the state of the operator and belt, preventing falls by ensuring the belt is worn and the lift is operated only when the belt is correctly attached, reducing accidents and enhancing safety.
Smart Images

Figure 0007948738000001 
Figure 0007948738000002 
Figure 0007948738000003
Abstract
Description
Technical Field
[0001] The present invention relates to a safety device that prevents an operator boarding the cab of a picking lift from falling off the cab, and a picking lift equipped with the same.
Background Art
[0002] The picking lift includes a cab that can move up and down along a mast, and a pair of left and right forks protruding rearward from the cab. In cargo handling work using a picking lift, a pallet is inserted into the forks, the cab carrying the operator moves to a high position, and the operator can load and unload cargo between the shelf and the pallet.
[0003] For example, as described in Patent Document 1, the picking lift is provided with a safety device that includes a belt to be worn by the operator and a connecting portion that connects the belt and the picking lift, so as to prevent the operator from falling from the cab to the floor surface.
[0004] By the way, the operator may find the wearing work troublesome or dislike being restrained, and may perform cargo handling work on the high-positioned cab without wearing the belt on the body. Since this work is extremely dangerous, in Patent Document 1, the safety device is provided with a wearing determination unit that determines whether the operator is wearing the belt, and based on the determination result of the wearing determination unit, the operator is warned to wear the belt, so as to prevent the operator from falling accidents.
[0005] The wearing determination unit of the safety device determines whether the operator is wearing the belt by detecting a mark attached to the belt with an image reading device. However, depending on the position of the belt worn by the operator, there has been a problem that it is difficult for the image reading device to detect the mark, or the mark is erroneously detected.
Prior Art Documents
Patent Documents
[0006] [Patent Document 1] Japanese Patent Publication No. 2015-224114 [Overview of the project] [Problems that the invention aims to solve]
[0007] Therefore, the problem that the present invention aims to solve is to provide a safety device that can more accurately determine the condition of the operator and the belt, such as whether the operator is wearing the belt, and a picking lift equipped with the same. [Means for solving the problem]
[0008] To solve the above problems, the safety device according to the present invention is A safety device to prevent an operator from falling from the driver's cab of a picking lift, The belt worn by the operator, An imaging unit that images the driver's cab, It includes a fitting determination unit that determines whether the operator is wearing the belt, The fitting determination unit is, A collection unit that collects images of the operator and the belt's condition, A learning model generation unit that generates and stores learning models using machine learning, An acquisition unit acquires an image of the current state of the operator and the belt using an imaging unit, The learning model generation unit generates a learning model, and the current image acquired from the acquisition unit is input to the prediction unit, which predicts the state of the operator and the belt. The system includes a decision unit that determines whether to warn the operator based on the prediction unit.
[0009] Preferably, the state of the operator and the belt includes the state in which the operator is in the process of putting on the belt.
[0010] Furthermore, the picking lift according to the present invention is The above safety device and, A warning unit that warns the operator to wear the belt, and a vehicle control unit that controls the warning unit.
[0011] Preferably, the vehicle control unit operates the warning unit when the wearing determination unit determines that the belt is not worn by the operator, then continues to operate the warning unit until a predetermined set time elapses when it is determined that the operator is in the middle of wearing the belt, and when the set time elapses, controls to stop the warning unit and disable the operation of the picking lift. [Advantageous Effects of Invention]
[0012] In the safety device according to the present invention and the picking lift equipped therewith, it is possible to more accurately determine whether the operator is wearing the belt. [Brief Description of Drawings]
[0013] [Figure 1] A side view showing a picking lift equipped with the safety device according to the present invention. [Figure 2] A block diagram showing the configuration of the picking lift. [Figure 3] A side view showing the first to third states of the operator and the belt. [Figure 4] A side view showing the fourth and fifth states of the operator and the belt. [Figure 5] A block diagram showing the wearing determination unit. [Figure 6] A flowchart showing the control method of the safety device. [Mode for Carrying Out the Invention]
[0014] Hereinafter, an embodiment of the safety device according to the present invention and the picking lift equipped therewith will be described based on the drawings.
[0015] As shown in Figure 1 and Figure 2, the order picking lift 1 includes a vehicle main body 2 that accommodates a traveling unit 21, a hydraulic unit 22, and the like, a mast 3 erected at the rear of the vehicle main body 2, an operator cab 4 that moves up and down along the mast 3, and a pair of left and right forks 8 protruding rearward from the operator cab 4. The traveling unit 21 includes a traveling motor for rotating wheels, a steering motor for turning the wheels, and the like. The hydraulic unit 22 includes a hydraulic tank, a hydraulic pump, and the like for lifting and lowering the operator cab 4.
[0016] The operator cab 4 has a floor frame 5, a front frame 6, a head guard 7, and an operation panel 9 provided on the front frame 6. The operation panel 9 is provided with a key switch for starting the order picking lift 1, as well as a handle and levers for instructing traveling of the order picking lift 1 and lifting / lowering of the operator cab 4.
[0017] The order picking lift 1 is provided with a safety device 10, and the safety device 10 includes a belt 11, a connecting part 12, and a winding part 13. The belt 11 is worn around the waist of an operator O who boards the operator cab 4. The connecting part 12 has one end attached to the belt 11 and the other end attached to the winding part 13.
[0018] Further, the winding part 13 is fixed to the head guard 7 of the operator cab 4. A portion up to 1 m from one end of the connecting part 12 is always pulled out from the winding part 13, and when tension is applied, the remaining portion is pulled out according to the tension. When a large tension is applied to the connecting part 12, the winding part 13 restrains the connecting part 12 from being pulled out, so that even if the operator O falls from the operator cab 4 at a high place, the operator O is suspended near the operator cab 4, and thus does not collide with the floor surface.
[0019] In the cargo handling work by the order picking lift 1, a pallet P is inserted into the forks 8, the operator cab 4 on which the operator O rides moves to a high place, and the operator O can perform the cargo handling work of loading and unloading a high-place cargo B between the shelf and the pallet P.
[0020] The safety device 10 includes an imaging unit 14 consisting of a CCD camera or the like. The imaging unit 14 is located below the control panel 9 and is positioned to capture the lower body of the operator O who is seated in the driver's cab 4, allowing for the determination of the condition of the operator O and the belt 11. The imaging unit 14 acquires image data at predetermined intervals, for example, 30 frames per second.
[0021] The safety device 10 includes a belt detection unit 16. The belt detection unit 16 is equipped with a locking part, such as a hook, on the upper part of the front frame 6 of the driver's cab 4, so as to be able to hold the belt 11. The belt detection unit 16 is also equipped with a weight sensor, a vibration sensor, etc., on the locking part such as the hook, so as to be able to detect the belt 11 held by the locking part. When operator O removes the belt 11 from his body and hooks the belt 11 onto the belt detection unit 16, the belt detection unit 16 can detect the belt 11 and detect that the belt 11 is not attached to operator O.
[0022] Furthermore, the safety device 10 includes an attachment determination unit 15. The attachment determination unit 15 consists of a microprocessor (MPU) and is configured to determine the state of the operator O and the belt 11, such as whether the belt 11 is attached to the operator O, based on the image captured by the imaging unit 14, and to determine the control of the picking lift 1, such as whether to warn the operator O.
[0023] The picking lift 1 is equipped with a warning unit 25. The warning unit 25 is configured to warn operator O to put on the belt 11 when the belt 11 is not attached to operator O. The warning unit 25 consists of a lamp or buzzer, etc., provided on the front frame 6 of the operator's cab 4, and warns operator O to put on the belt 11 by lighting up or sounding.
[0024] The picking lift 1 is equipped with a vehicle control unit 20. The vehicle control unit 20 consists of a microprocessor (MPU) and may be integrated with the mounting determination unit 15. The vehicle control unit 20 is configured to control the travel unit 21, the hydraulic unit 22, and the warning unit 25 based on instructions from the operator O from the control panel 9, the detection results of the belt detection unit 16, and the determination results of the mounting determination unit 15.
[0025] As shown in Figures 3 and 4, there are five possible states for the relationship between operator O and belt 11. (1) As shown in Figure 3(A), the state in which operator O is not in the driver's cab 4 and the belt 11 is held by the belt detection unit 16 is defined as the "first state". (2) As shown in Figure 3(B), the state in which operator O is not in the driver's cab 4 and the belt 11 is not held by the belt detection unit 16 is defined as the "second state". (3) As shown in Figure 3(C), the state in which operator O is seated in the driver's cab 4 and the belt 11 is not attached to operator O is defined as the "third state". (4) As shown in Figure 4(D), the state in which operator O is seated in the driver's cab 4 and the belt 11 is attached to operator O is referred to as the "fourth state". (5) As shown in Figure 4(E), the state in which operator O is seated in the driver's cab 4 and the belt 11 is in the process of being attached to operator O is referred to as the "fifth state".
[0026] As shown in Figure 5, the fitting determination unit 15 includes a collection unit 40 that collects image feature quantities of the state (1st to 5th states) of the operator O and the belt 11 as input data for training data 46. The feature quantities are extracted from the image using a general method, which quantifies the features of the object in the image.
[0027] The fitting determination unit 15 includes a learning model generation unit 41 that performs machine learning on training data 46 collected by the collection unit 40 and generates and stores a learning model through machine learning. In this embodiment, the learning model generation unit 41 performs supervised learning with training data 46 that associates the state of operator O and belt 11 (first to fifth states) with image features. The learning model uses the training data 46 generated by the learning model generation unit 41 to output the state of operator O and belt 11 (first to fifth states) as output data according to the image features.
[0028] The learning model generation unit 41 uses a general machine learning algorithm such as a neural network. The learning model generation unit 41 generates a learning model that estimates output from input, that is, a learning model that takes input data as input and outputs output data.
[0029] The fitting determination unit 15 includes an acquisition unit 45 that acquires an input data ID, which is a feature quantity based on the current image captured by the imaging unit 14, at predetermined intervals.
[0030] The fitting determination unit 15 includes a prediction unit 42 that predicts the state of operator O and belt 11 (first state to fifth state) as output data OD by applying the learning model generated by the learning model generation unit 41 to the current input data ID acquired by the acquisition unit 45.
[0031] The mounting determination unit 15 includes a decision unit 43. Based on the output data OD predicted by the prediction unit 42, the decision unit 43 recognizes the current state of the operator O and the belt 11 (first state to fifth state) and decides to warn the operator O to mount the belt 11, etc. The vehicle control unit 20 is configured to control the warning unit 25, etc., based on the decision of the decision unit 43.
[0032] As shown in Figure 6, the picking lift 1 equipped with the safety device 10 performs the following control.
[0033] Operator O turns on the key switch to start the picking lift 1 (step S1). Then, the imaging unit 14 images the driver's cab 4, and the attachment determination unit 15 acquires the images from the imaging unit 14 at predetermined intervals.
[0034] The mounting determination unit 15 determines whether the belt 11 is visible in the image from the imaging unit 14 (step S2). If the belt 11 is visible in the image, the mounting determination unit 15 determines whether the belt 11 is mounted on the operator O (step S3). Then, in the fourth state where the belt 11 is mounted on the operator O (Figure 4(D)), the vehicle control unit 20 controls the vehicle so that the driving unit 21 and the hydraulic unit 22 can be operated (step S4).
[0035] When the attachment determination unit 15 determines that the belt 11 is not visible in the image (step S2), or that the belt 11 is visible in the image but is not attached to operator O (step S3), it determines whether the belt 11 is being held by the belt detection unit 16 (step S5).
[0036] Step S5 determines whether the belt detection unit 16 is holding the belt 11, and when the belt 11 is in the first state (Figure 3(A)) where it is being held by the belt detection unit 16, the system controls the operation of the running unit 21 and the hydraulic unit 22 to be disabled (Step S11).
[0037] Furthermore, the vehicle control unit 20 determines whether the belt detection unit 16 is holding the belt 11 (step S5), and if it determines that the belt 11 is not being held by the belt detection unit 16 in a second state (Figure 3(B)), a third state (Figure 3(C)), or a fifth state (Figure 4(E)), the vehicle control unit 20 controls the warning unit 25 to operate (step S6).
[0038] Then, after the warning unit 25 is activated, the mounting determination unit 15 determines whether the belt 11 is in the fifth state (Figure 4(E)) in the process of being mounted on the operator O (step S7). When the mounting determination unit 15 determines that the belt 11 is in the fifth state (Figure 4(E)) in the process of being mounted on the operator O, the vehicle control unit 20 controls the warning unit 25 to stop (step S8).
[0039] Furthermore, when the mounting determination unit 15 determines that the belt 11 is not in the process of being mounted on the operator O, which is a third state (Figure 3(C)) (step S7), the vehicle control unit 20 starts counting that state (step S9), returns to step S6, and the operation of the warning unit 25 continues. If the counted time is less than a predetermined set time, the operations of steps S6, S7, and S9 are repeated.
[0040] Then, when the counting time has elapsed to a predetermined set time (step S9), the vehicle control unit 20 stops the operation of the warning unit 25 (step S10), disables the operation of the travel unit 21 and the hydraulic unit 22 (step S11), and controls the key switch to turn off (step S12). As a result, operator O is unable to operate the picking lift 1.
[0041] Although preferred embodiments of the present invention have been described above, the configuration of the present invention is not limited to these embodiments. For example, it can be modified as follows.
[0042] The attachment determination unit 15 does not need to make a determination based solely on whether it corresponds to the first to fifth states, but should be able to determine whether the belt 11 is not attached to the operator O or whether the belt 11 is attached to the operator O.
[0043] Alternatively, instead of the belt detection unit 16 determining whether it is holding the belt 11, the attachment determination unit 15 may determine whether the belt 11 is being held by a locking mechanism such as a hook, based on the image from the imaging unit 14.
[0044] Furthermore, the attachment determination unit 15 may be configured to map to the first to fifth states by means of unsupervised learning or semi-supervised learning, in addition to supervised learning as in the above embodiment.
[0045] Furthermore, the warning unit 25 may be configured to display warning marks or the like on the display unit of the control panel 9, in addition to a lamp or buzzer as in the above embodiment.
[0046] The effects of the present invention will be explained.
[0047] The safety device 10 according to the present invention is a safety device 10 for preventing an operator O who is on the driver's cab 4 of a picking lift 1 from falling, and comprises a belt 11 worn by the operator O, an imaging unit 14 that images the driver's cab 4, and a fitting determination unit 15 that determines whether the operator O is wearing the belt 11. The fitting determination unit 15 comprises an imaging unit 40 that collects images of the state of the operator O and the belt 11, a learning model generation unit 41 that generates and stores a learning model by machine learning, an acquisition unit 45 that acquires an image of the current state of the operator O and the belt 11 by the imaging unit 14, a prediction unit 42 that predicts the state of the operator O and the belt 11 by inputting the current image acquired from the acquisition unit 45 into the learning model generated by the learning model generation unit 41, and a determination unit 43 that decides whether to warn the operator O based on the prediction unit 42.
[0048] In this way, the attachment determination unit 15 of the safety device 10 predicts the state of operator O and belt 11 based on the image captured by the imaging unit 14 and the learned model. Therefore, regardless of the position of the belt 11 attached to operator O, it is possible to more accurately determine the state of operator O and belt 11, such as whether operator O is wearing the belt 11.
[0049] Preferably, the state of the operator O and the belt 11 includes the state in which the operator O is in the process of putting on the belt 11.
[0050] As a result, when the attachment determination unit 15 determines that operator O is in the process of attaching the belt 11, the belt 11 is not yet attached to operator O, but operator O is in the process of attaching the belt 11. Based on this state, the picking lift 1 can be operated more appropriately.
[0051] Furthermore, the picking lift 1 according to the present invention includes the safety device 10 described above, a warning unit 25 that warns the operator O to attach the belt 11, and a vehicle control unit 20 that controls the warning unit 25.
[0052] In this way, the attachment determination unit 15 of the safety device 10 determines the state of the operator O and the belt 11, so it can more accurately determine the state of the operator O and the belt 11 regardless of the position of the belt 11 attached to the operator O. Therefore, based on this determination result, the warning unit 25 can appropriately warn the operator O to attach the belt 11.
[0053] Preferably, when the mounting determination unit 15 determines that the belt 11 is not mounted on operator O, the vehicle control unit 20 activates the warning unit 25. Subsequently, when it determines that operator O is in the process of mounting the belt 11, the warning unit 25 continues to operate until a predetermined set time has elapsed. Once the set time has elapsed, the warning unit 25 is deactivated, and the vehicle control unit 20 controls the system to disable the operation of the picking lift 1.
[0054] As a result, when the attachment determination unit 15 determines that the belt 11 is not attached to operator O, it can appropriately warn operator O to attach the belt 11. When it determines that operator O is in the process of attaching the belt 11, the warning unit 25 stops issuing a warning because operator O is trying to attach the belt 11, thus not interfering with operator O's work. The vehicle control unit 20 then counts the time it takes for operator O to attach the belt 11, and once the set time has elapsed, it determines that operator O is unable to attach the belt 11, thus disabling the operation of the picking lift 1 and preventing operator O from falling. [Explanation of symbols]
[0055] 1 Picking lift 4 Driver's cab 10 Safety equipment 11 belts 14 Imaging Unit 15. Fitting determination unit 20 Vehicle Control Unit 25 Warning section 40 Collection Department 41 Learning Model Generation Unit 42 Prediction Section 45 Acquisition Department 43 Decision Section O Operator
Claims
[Claim 1] A safety device to prevent the operator from falling from the driver's cab of the picking lift, A warning unit that warns the operator to put on the belt, A picking lift comprising a vehicle control unit that controls the warning unit, The aforementioned safety device is A belt attached to the operator, An imaging unit for imaging the driver's cab, The system includes an attachment determination unit that determines whether the operator is wearing the belt, The aforementioned mounting determination unit is A collection unit that collects images of the state of the operator and the belt, A learning model generation unit that generates and stores learning models using machine learning, An acquisition unit that acquires an image of the current state of the operator and the belt using the imaging unit, A prediction unit predicts the state of the operator and the belt by inputting the current image acquired from the acquisition unit into the learning model generated by the learning model generation unit, The system includes a decision unit that determines whether to warn the operator based on the prediction unit, The vehicle control unit, When the attachment determination unit determines that the belt is not attached to the operator, the warning unit is activated. Subsequently, when the unit determines that the operator is in the process of attaching the belt, the warning unit is activated until a predetermined set time has elapsed. Once the set time has elapsed, the warning unit is deactivated, and the system is controlled to disable the operation of the picking lift. A picking lift characterized by the following features.
Citation Information
Patent Citations
Logistics forklift safety protection device based on electronic safety belt
CN219326579U
Vehicle occupant restraint protection device
JP2000127897A
Safety device of picking fork lift truck
JP2014144865A
Safety device of picking lift
JP2015224114A
Determination device and determination method
JP2020021323A
