JPH03252707A - Unmanned forklift with vision - Google Patents

Abstract

PURPOSE:To easily change a travel route by generating data of range where a carframe can travel with a first analysis means based on an image signal from a first image pickup means. CONSTITUTION:A television camera 51 for travel is provided with a function to photograph a white line plotted on the travel route, the edge of the travel route, and the type of facility in a factory arranged in the neighborhood of the travel route, and a photographed image is stored in an image memory 52, and is inputted to an image processing means 53. The image processing means 53 finds the range where an unmanned forklift can travel i.e. a travel area where the unmanned forklift 1 can travel by evading obstruction based on inputted image data. Thereby, it is possible to dispense with the installation of a specific mark to recognize the travel route by the unmanned forklift and to easily perform the change of the travel route.

Description

[Detailed description of the invention]

[Industrial Application Field] The present invention relates to an unmanned forklift 1 used in a production factory, etc.
- In particular, it relates to unmanned forklifts that can travel without the use of guide lines or anti-fouling tape laid on the road surface. Those with guided wires or pasted on the driving road surface.

[Some methods are known, including one using a foul tape that has been pulled and one using a gyroscope. However, in recent years, technology has been developed that allows an automated guided vehicle to travel along a travel route determined by anodicizing signs placed on the road surface and processing the pixel signals in the image. known (JP-A-1-188909, JP-A-1-188910, JP-A-1-18)
Publication No. 8911). The image-based automatic guided vehicle disclosed in this publication images discretely arranged marks, processes the pixel signals in the image to recognize the marks, and based on this, determines the deviation of the travel line. I am trying to determine the steering angle. [Problems to be Solved by the Invention] However, in conventional automatic guided vehicles, guide wires, reflective tape, etc. must be laid on the traveling road surface as a guiding means, and automatic guided vehicles equipped with imaging devices Even in this case, it is necessary to install special markings to recognize the driving route. Therefore, there is a problem in that easily changing the traveling route involves turning, and it is not possible to respond quickly to changes in the production method. By the way, the above-mentioned technology is a technology for automatic guided vehicles that simply loads cargo onto a loading platform, and in the case of an unmanned forklift that receives cargo placed at a predetermined location on a ramp, there are even more difficult problems regarding cargo handling. exist. In other words, if the cargo is placed out of alignment, if you do not take the misalignment into consideration when handling the cargo, the cargo will not be loaded correctly on the fork, and the cargo may collapse while traveling. , leading to a fall onto the road surface. Furthermore, if cargo is handled while the cargo is not in its normal position, the cargo may be transported in a slumped state, causing it to fall onto the road as described above. In this way, in the conventional unmanned forklift 1~,
If the cargo is not handled properly because it is not possible to identify misalignment or poor posture of the cargo on the forklift platform, problems are likely to occur during the transportation of the cargo, which hinders the improvement of logistics efficiency. The present invention has focused on the above-mentioned problems, and an object of the present invention is to provide an unmanned forklift with visual function that can easily change the travel route and can reliably perform normal cargo handling operations. [Means for Solving the Problems] The unmanned forklift with vision according to the present invention that meets this objective can take the following aspects. (1) A first Ik3 image means, which is installed on the side of the machine that can travel, and can show the white line drawn on the travel route, the road shoulder of the travel route, and the form of factory equipment placed near the travel route. a first movable means that holds the first Vii image means and changes the field of view of the first image means; and a range within which the machine base can travel based on the image signal from the first image means. a first analysis means that creates data on the machine; and a second analysis means that captures the form of the cargo transported by the machine.
an image means, a second movable means for holding the second eleven means and changing the field of view of the second image means, and determining the attitude of the baggage based on the image signal from the second image means, An unmanned forklift with vision, comprising: second analysis means for correcting the trajectory of the approach of the machine to the cargo; and an unmanned forklift with vision. (2) White lines installed on the side of the machine that can run and drawn on the running route, road shoulders of the running route, form of factory equipment placed near the running route, and cargo carried by the machine. a movable means for holding the image means and changing the field of view of the YRI image means; and data on a range within which the machine can travel based on the image signal from the image means. An unmanned forklift with a visual system, characterized in that the unmanned forklift with a visual system is equipped with the following: , the first imaging means carries the white line drawn on the driving route, the shoulder of the driving route, and the form of factory equipment placed near the driving route.In this case, the first imaging means Since the field of view is changed by the first movable means, the first imaging means can dance around the driving route over a wide range.When each form is captured by the first imaging means, the first imaging means By the analysis means of
[Cliff] ~ data on the range in which the machine can travel is created. That is, based on the image signal from the first image capturing means,
Image processing is performed to find a travel area in which the machine can avoid obstacles. Further, the second imaging means captures the form of the cargo being transported by the machine, and the second analysis means analyzes the image. In this case as well, since the field of view of the second imaging means is changed by the second movable means, the second imaging means can capture the state of something over a wide range. By the second leaping means, picture 1! When I98 Tanuki is performed, the attitude of the cargo is recognized and the approach trajectory of the aircraft is determined. In other words, the locus of movement of the machine to the cargo is determined so that the cargo is loaded in a normal state throughout the day, and cargo handling is performed based on this. Therefore, even if the cargo is deviated from its normal position or its posture is not normal, it is possible to handle the cargo in a way that corrects and covers it, and the cargo is loaded in a normal state by nodding. Note that even if the first image-jumping means and the second image-jumping means are shared, it is possible to recognize the form near the traveling route f=l and the form of the baggage, so the configuration of the device can be simplified. It is possible to perform operations similar to those described above while changing the [Embodiments] Hereinafter, preferred embodiments of the visual unmanned forklift according to the present invention will be described with reference to the drawings. First Embodiment FIGS. 1 to 5 show a first embodiment of the present invention. In the figure, reference numeral 1 indicates an unmanned vehicle lift 1 to 1, and the unmanned forklift 1 has a machine platform (vehicle body) 3 that travels along a traveling road surface 2. The machine base 3 includes a front wheel 5, a rear wheel 6, a running motor 7, a steering motor 8,
It consists of masts 21, 22, etc. The machine base 3 has a running section frame 3a extending along the running road surface 2, and the running section frame 3a has 21! ! J's front wheels 5 and 1
rear wheels 6 are attached. The front wheel 5 is a driven wheel, and is placed at the front end of the traveling section frame 3a. rear wheel 6
is a steering/driving wheel, which is arranged at the rear of the traveling section frame 3a. The rear wheels 5 are rotationally driven by a traveling motor 7 located above, and are steered by a steering motor 8. A case 9 is provided above the travel section frame 3a, and a general control device 20 as a transportation and cargo handling control processing device is housed in the case 9. A battery 14 is housed in the case 9, and the battery 14 drives the travel motor 7 and other electrical equipment described later. At the front end of the running section frame 3a, there is a space extending upward.
- 21 and 22 are arranged, and the upper ends of both masts 21 and 22 are connected via a connecting member 23. A hydraulic cylinder 24.25 as a descending means is arranged inside each mass (~21.22).The lower end of each hydraulic cylinder 24.25 is fixed to the mast 21.22, and the hydraulic cylinder The lot is 'Nikata Enhiru J
, is becoming a sea urchin. At the tip of each lot, there is a bowl of soup, and each sprocket is threaded with a roller rubber fixed to one side or square 1. Outside of 21.22, these squares 1 to 21.
A guide portion IJ27 that is slidable along 22 is provided. The guide member 27 has a 1-shaped) A-ri 33.3.
4 or each is taken.'' (Pulled. Fork 33.3
4 is for lifting loads, as is well known, and its maiden part is flat. The hydraulic cylinders 24 and 25 are driven by oil pumped from hydraulic units 1 to 50 mounted on the travel section frame 4a. Oil direction switching control is performed by electromagnetic valves 51, 52 which are activated by signals from the general control device 20. In other words, the forks 33, 34 are raised and lowered by the expansion and contraction of the rods of the hydraulic cylinders 24, 25, and the amount of lift of the forks 33, 34 is detected by the lift height sensor 1-35. The masts 21.22 can be tilted forward and backward using the lower end as a fulcrum.At the bottom of each mast 21.22,
Hydraulic cylinders 28.29 for tilting, one of which is fixed to the traveling section frame 3a side, are connected, and the mast 21.22 is tilted in the front-rear direction by the expansion and contraction of this hydraulic cylinder 28.29. . In other words, when the unmanned forklift 1 is running, the dill I
・The mast 21.22 is tilted backwards so that the hydraulic cylinders 28.29 are in a retracted state and the tip of the lever 33.34 is higher than the base. During the cargo handling work of T-Clear 1 to 1, the hydraulic cylinders 28 and 29 for tilting are in an extended state, so that the maiden part of the lift 33 and 34 is horizontal. Squares 1-21.22 are placed in an upright position. Each tilt hydraulic cylinder is driven by oil pumped from a hydraulic unit 50. In this case, the oil direction switching control is performed by controlling the solenoid valve 53 which is activated by a signal from the central control unit 1 device 20. Mast 2 is located near the bottom of squares 1 to 21.
An aisle I/angle detection sensor 71 is arranged to check the tilt state of the vehicle 1, and the aisle I/angle detection sensor 71 is connected to the general control device 20. At the bottom end of 21 to square 1, as shown in FIG.
The hereach cylinder 72 is triple-coated, and the first
21 is a cylinder 72 for this mass 1 to reach during cargo handling.
It is pushed forward by. The hydraulic cylinder 72 is connected to an electromagnetic valve 73 controlled by the overall control device 20. The amount of movement of the mast 21 is detected by a reach sensor 36. A signal from the reach sensor 36 is input to the overall control device 20. In addition, lift 1 load sensors 37 are provided in the spaces 1 to 21. The lift load sensors 37 are required to detect the weight ω of the cargo loaded on the machine platform 3. "(,
) 4-33.34 ℃-2 sabot gain for driving, traveling motor 7, sdering and 1 of 8
By changing the no-vo gain, it is possible to achieve optimal operation with less loss. This riff 1~ta ■ hand sen 1 no 37
A signal from the controller is input to the overall control device 20. A load position detection sensor 38 is provided above the suspension root of the beam A-33. There are 11 pieces of luggage! It has a function of detecting whether a load is loaded and correcting the override locus of the machine 3 based on information from a leg sensor 44, which will be described later. The loading device detection sensor 1/38 and leg sensor 44 (No. 9) are also input to the overall control device 20. The left and right front wheels 5 have encoders 40a for position detection.
, 40b are connected to each other. Each encoder 4
0a and 40b have a function of opening the rotation amount of the front wheels 5, and the straightness is detected from the mileage d3 and the difference between the left and right rotation amounts. An encoder 41 is installed on the rear wheel 6.
or connected. The encoder 41 has the function of detecting the steering angle of the rear wheels 6. At the hanging root part of the rim A-ri 33.34, there is a rim 33.
A false insertion detection sensor 1-42 detects whether the cargo 45 is completely inserted into the pallet 46 on which the cargo 45 is placed.
is set up. Also, there is a false insertion detection sensor +)
For the near 1 power of 42, there is a palette that detects palettes 1 to 46]
- A contact detection sensor 43 is provided. A leg sensor 44 is installed at the front end of the machine base 3 to detect the lateral displacement of the pallet 46. It consists of a cutting-type two-dimensional sensor 1), with one each near the left and right front wheels (diameter width) 5.
They are placed one by one. This leg sensor 44 detects the amount of positional deviation of the pallet 46 when the machine 3 approaches the target object vJVIJJA 5 by 1 to 5 m using the prefectural image means described later, and detects the amount of positional deviation of the pallet 46. The present invention provides a function to accurately measure the position and posture of the loaded baggage 45. Based on the data from this leg sensor 1 no 44, the luggage 45 of aircraft 3
The ab [l-ch locus has been corrected, making it possible to perform accurate cargo handling operations. At the tip of the fork 33.34,
5 is marked R'At, and this fork sensor 48 is composed of a light-cutting two-dimensional sensor, similar to the small one above. The fork sensor 48 has a function of detecting the fork insertion openings of the pallets 1 to 46 and detecting displacement of the pallets 1 to 46 in the vertical direction. The signal from the fork sensor 48 is input to the overall control til11 device 20. This makes it possible to accommodate variations in working height at the cargo handling position. A traveling television camera (CCD camera) 51 is arranged at the upper end of the machine stand 3 as a first imaging means. The traveling television camera 51 is fixed to the machine base 3 via a pan head 54 as a first movable means. The pan head 54 has a function of swinging the traveling television camera 51 vertically and horizontally, thereby causing the traveling television camera 51 to swing vertically and horizontally.
field of view has been greatly expanded. The driving TV camera 51 is connected to the general system 1 via an image mail 52 and an image processing means 53 as a first analysis means.
The operation of one device 20 is connected to the vJ role control processing device 20a. The traveling television camera 51 has the function of capturing the white lines drawn on the traveling route, the shoulders of the traveling route, and the form of factory equipment placed near the traveling route. In this example, a white line is drawn on the driving route in order to raise the level of the pixel signal, but if the road surface is 18e, the color is not limited to a white line, and as long as it is a bright color, the function will be sufficient. J
ffl will be displayed. The image camera 52 displays an image of the driving route (J) near the camera 51 for driving.
It is converted into a value (replaced with a binary number) and stored. The image processing means 53 is configured to create and restore data for a range in which the machine 3 can travel, based on the image data stored in the image memory 52. A cargo handling television camera (CCD camera) 61 is arranged at the top of the mast 21 as a second imaging means. 21. Pan head 64
has a function of swinging the cargo handling television camera 61 vertically and horizontally, thereby greatly expanding the field of view of the cargo handling television camera 61. The inspection television camera 61 is connected to the operation and cargo handling control processing device 20a of the general control device 20 via an image memory 62 and an image processing device 63 as a second analysis device. The television camera 61 for the load 19 has the function of being able to change the shape of the load 45 being conveyed by the machine 3. The image memory 62 is captured by the cargo handling television camera 61! lj
: Activates the function of binarizing and storing the image of the shadowed baggage 45. Further, the image processing means 63 includes an image memory 62
Based on the image data stored in the image data, it is determined whether there is a plurality of pieces of luggage 45. The image processing means 63 searches for the upper left corner, lower left corner, and lower right corner of the stacked luggage based on the image data stored in the image memory 62, for example. By searching each of these points, the aspect ratio is calculated from the length (width) of the bottom side of the package and the height of the stacked package. The image processing means 63 stores in advance the aspect 1-ratio of only one cargo 45 d3, and by overlapping the aspect I ratio of the whole cargo and the aspect 1-ratio of one cargo on the ratio axis, The number of stacked stages p is set to 1. In this way, the image processing means 63 as the second analysis means
measures the object, the baggage 45, three-dimensionally, and
The function to create and cover data on the position and orientation of the baggage 45 is performed. Then, based on the data on the position and orientation of the baggage 45, the abroadge locus of the machine platform 3 to the baggage 45 is corrected. Next, we will explain the effect of returning to the visual field 14 unmanned screen 4-clino 1-. The unmanned forklift i-1 travels along the traveling road surface 2 when the destination ([1 marker station]) is specified. When the vehicle is traveling, the traveling distance is subtracted by the encoders 40a and 40b, and the straightness is detected from the difference in rotation between the two encoders. When traveling, the unmanned forklift A-cliffs 1 to 1 travel while determining whether or not they are at a predetermined stage, and the forward field of view of the unmanned forklift 1 in the traveling direction is seen by the traveling television camera 51. In this case, the traveling television camera 51 is operated by the pan head 54.
Since it can be swung vertically and horizontally, the field of view in the direction of travel is greatly expanded. Traveling television camera 51
As mentioned above, it has the function of changing the shape of the white line drawn on the driving route, the road shoulder of the driving route, and the factory equipment placed near the driving route, and is recorded by the driving television camera 51. The obtained image is stored in the image memory 52. This image is input by the image processing means 53. The image processing means 53 creates data on the travelable range of the unmanned forklift 1 based on the input image data. In other words, the image processing means 53 determines a driving area in which the unmanned forklift 1-1 can avoid the roadblocks. Therefore, it is no longer necessary to lay a special mark for the unmanned forklift truck 1 to recognize the travel route, as in the past, and the travel route can be easily changed. When the unmanned forklifts 1 to 1 reach their target stations, the unmanned forklifts 1 to 1 gradually approach the cargo 45. In this case, the cargo 45 is tracked by the cargo handling television camera 61. The TV camera 61 for cargo 19 can be swung vertically and horizontally relative to the stand 64, so that multiple cargoes 45 (even if they are piled up, the entire cargo is left unattended) can be moved by A-crit 1. It can be decoyed while standing still. Images captured by the cargo handling television camera 61 are stored in an image memory 62. The image of the luggage 45 stored in the image memory 62 is three-dimensionally measured by the image (3) processing means 63, and its size and posture are analyzed. That is, here the entire load is coordinated,
The aspect ratio of the entire load and the aspect ratio of a single load are compared, and the stacked height of the load is determined, as well as the actual positional deviation of the load 45 with respect to its normal posture. Once the position, height, and attitude of the load 45 are determined, the approach locus of the unmanned forklift [~1] to the load 45 is corrected based on this data. That is, the unmanned forklift 1 stores in advance a traveling pattern in which it advances toward the cargo 45, and this traveling pattern is corrected based on the actual positional deviation of the cargo 45, etc. When the unmanned forklift 1 approaches the cargo 45, the machine 3
A leg sensor 44 installed at the front end of the container detects the lateral displacement of the pallets 1 to 46 on which the cargo 45 is placed. Therefore, the position and attitude of the cargo 45 can be measured more accurately than the data obtained by the cargo handling television camera 61. Based on the data measured by this leg sensor 1-44, the cargo 4 of the unmanned cargo A-cliff 1~1
The approach trajectory to No. 5 is modified to be even more accurate. Therefore, the unmanned cliffs 1 to 1 can approach the cargo 45 from a direction corresponding to the positional shift of the cargo 45. When the unmanned FL-Cliff I~1 approaches the cargo 45 to a predetermined distance, the traveling of the unmanned FL-Cliff]~1 is stopped. When unmanned forklift l-1 comes to a complete stop,
The rod of the hydraulic cylinder 72 for the mass mill reach is extended to ensure that the flanges 33, 34 are inserted into the pallet 46. A confirmation sensor 42 confirms whether or not the leaves 33, 34 are completely inserted into the pallet 46. When the fliers 33 and 34 are inserted into the pallet 46,
The rods of the hydraulic cylinders 61, 62 for the aisles 1-1 are retracted, and the pallets 46 and cargo 45 are tilted toward the platform 3 of the unmanned forklifts 1-1. When the pallet 46 and the cargo 45 are moved slightly toward the machine base 31, the ends of the hydraulic cylinders 24.25 extend, and the pallet (~46 and the cargo 45) is moved above the delta 33.34. It is lifted off the floor by its back. In this feeling of being alone, F-A-33.
It means that the pallets 1 to 46 are loaded on the 340th block] - From the contact detection sensor 43 to the general control device 2
Output to 0. When the pallet 46 is lifted onto the lift 33.34, the pallet 46 and the cargo 4
The load No. 5 is detected by the refmill cart sensor 37. Riff 1-Load Sen 1 No. 37 J: Tsutepare h 4
6J-; J: When the 'vJ weight of the luggage 45 is detected,
Based on this load, the drive gain of the travel motor 7 and steering motor 8 is changed to realize optimal operation with less loss. In addition, when the pallets 1 to 46 and the cargo 45 are lifted up, the cargo position detection sensor 38 detects the pallets 1 to 46 and the cargo 45.
46 is loaded on the flywheel 33, 34 is detected. This makes it possible to determine if the pallet 46 or the forks 33, 34 are not loaded in the appropriate position. Therefore, if the pallet 46 is not loaded in an appropriate position, the unmanned forklift 1 will temporarily retreat from the pallet 46 and return to the platform 33.34.
The insertion operation is performed. That is, here, the approach trajectory is corrected based on information from the 1tPJ load position detection sensor 38 and the leg sensor 44 mentioned above, and based on this, the slide A-33, 34 is inserted into the correct position of the pallets I-46. It will be done. When the pallet 1-46 and the cargo 45 are lifted by the fork 33.34, the height of the fork 33.34 is detected by the lifting height sensor 35, and the fork 33.34 is stopped at a preset position. In this state, 111
(Person) A-kuri 71-1 travels to the next destination and unloads pallet 1-46 and baggage 45 at the destination. In this way, even if the pallet 1-46 cargo 45 is not set accurately in the normal position, it is possible to handle the cargo accordingly, and the pallet 1-46 can be moved by the fork 33.
．． The receiver is stopped at the most suitable position of 34. Therefore, the situation in which the load 45 slides forward or falls onto the road surface while the load 45 is being conveyed is avoided, and efficient conveyance is achieved. Second Embodiment FIG. 6 shows a second embodiment of the present invention. The second embodiment differs from the first embodiment only in the configuration and mounting position of the imaging means; other parts are the same as in the first embodiment. By adding the numeral ``,'' the explanation of the corresponding parts will be omitted and only the different parts will be explained.In Fig. 6, 101 indicates a traveling and load saving camera as a layer image means.First implementation In the example,
Although the first imaging means and the second M imaging means were respectively installed in the unmanned A-clients 7I to 1, in this embodiment,
There is only one historical means. That is, in this embodiment, the imaging means 101 has the function of imitating the white line drawn on the travel route, the shoulder of the travel route, the form of factory equipment placed near the travel route, and the form of the cargo 3 to be transported. have. The path means 101 is provided at the top of the mast 21 via a pan head 102 as a movable means. The pan head 102 has the function of swinging the television camera 101 for traveling and VJ up and down and left and right, thereby greatly expanding the field of view of the television camera 201 for traveling and VJ. The pan head 102 of this embodiment can rotate in the entire circumferential direction (360° C.). In the second embodiment constructed as described above, the J3 has the functions of one television camera or a traveling television camera and a cargo handling television camera, so it is more convenient than the first embodiment. The configuration of the device is simplified, and the device size can be reduced to 1. [Effects of the Invention] As explained above, when using the visual unmanned forklift truck 1~ according to the present invention, the white line drawn on the traveling route, the road shoulder of the traveling route, and the intra-factory transport station located near the traveling route are provided. The shape is captured by the first imaging means held by the first movable means, and based on the image signal from the first imaging means, the data of the range in which the platform can travel is analyzed in the first analysis. Since I created J: for the means, it becomes possible for an unmanned forklift to run without track. Therefore, it is not necessary to install special marks or the like to recognize the driving route as in the past, and the driving route can be easily changed or covered. This makes it possible to quickly respond to changes in production methods. Further, the configuration of the cargo to be transported is captured by the second layer image means held by the second movable means, the posture of the cargo is determined based on the image signal from the second My image means, and the cargo on the machine base is By using the second analysis means to correct and cover the approach trajectory, the load can be loaded in a normal state against the slope, and the front side of the load can be minimized when traveling. It is possible to reliably prevent the product from falling onto the road surface. In addition, by using imaging means that can capture the white lines drawn on the driving route, the road shoulders of the driving route, the form of factory equipment placed near the driving route, and the form of the cargo being transported, two The mud image means can be combined into one decoy image means, and the configuration of the apparatus can be simplified.

[Brief explanation of drawings]

FIG. 1 is a side view of an unmanned forklift with visual function according to the first embodiment of the present invention, FIG.
The figure is a front view of Figure 2, Figure 4 is a rear view of Figure 2, Figure 5 is a block diagram showing the control system of the Senjin forklift of Figure 1, and Figure 6 is a block diagram of the control system of the Senjin forklift of Figure 1. Vision (=I unattended) A according to the second embodiment
-A side view of Chryso1-. 1......Unmanned forklift 1~2...Travel road surface 3...Machine base 20...General control device 33.34...Fork 37...Ref1 load sensor 38...Loading position detection sensor 1 No.44...
・Legsen 1no 45... Luggage 46... Baren 1-51... First road means (driving camera) 53... First Analysis means (image processing means) 54
. . . First movable means 61 . . . Second roadway means (evening/revision camera) 63 . . . Second analysis means (image processing means) 64
. . . Second movable means 101 . . . Roadway means (J-Rebi camera for traveling and loading one load) 102 . . . Movable means

Claims (1)

[Scope of Claims] 1. A first camera installed on the side of the machine that can travel and capable of photographing the white line drawn on the travel route, the road shoulder of the travel route, and the form of factory equipment placed near the travel route. an imaging means; a first movable means for holding the first imaging means and changing a field of view of the first imaging means; and a movable means for moving the machine based on an image signal from the first imaging means. a first analysis means for creating range data; and a second analysis means for taking a picture of the form of the cargo being transported by the machine.
imaging means; second movable means for holding the second imaging means and changing the field of view of the second imaging means; determining the attitude of the luggage based on the image signal from the second imaging means; An unmanned forklift with vision, comprising: a second analysis means for correcting the approach locus of the machine to the cargo; 2. The white line drawn on the running route, the road shoulder of the running route, the form of factory equipment placed near the running route, and the cargo carried by the machine. an imaging means for photographing the form; a movable means for holding the imaging means and changing the field of view of the imaging means; creating data on a range in which the machine can travel based on image signals from the imaging means; An unmanned forklift with vision, comprising: an analysis means for correcting the approach locus of the machine toward cargo; and an unmanned forklift with vision.