Field management system

Abstract

To provide a field management system that achieves high-precision self-position estimation for robots moving in the field by a technique readily applicable in agricultural settings.SOLUTION: A field management system includes a track and positional information identification means provided along the track. The robot includes movement means that moves along the track and also includes an environmental sensor that collects environmental information on the field. Self-position generation means generates self-position information of the robot on the basis of at least information exchanged between the robot and the positional information identification means. An information management device manages the filed on the basis of the self-position information and the environmental information.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a farm land management system that realizes highly accurate self-location estimation of a robot moving in a farm land. [Background technology]

[0002] In recent years, the use of robots and IoT has led to more sophisticated management of agricultural fields.

[0003] Conventionally, in farm field operations using robots, the position of the robot is estimated by techniques such as GPS, SLAM, odometry, etc., and the farm field is managed. In addition, as in Non-Patent Document 1, it is being considered to identify the positions of individual fruits, etc., create a map of them, and use them in farm field management. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Development of an automatic tomato fruit harvesting robot for greenhouse horticulture (Journal of the Robotics Society of Japan Vol.39 No.10 , pp921-925 , 2021) Summary of the Invention [Problem to be solved by the invention]

[0005] However, in relatively small-scale fields such as vinyl greenhouses, it is difficult to adopt expensive self-location estimation technologies such as GPS and SLAM, and their operation and management are also not easy. In addition, although odometry is a simple method and is widely used, in a field environment, tire slippage and skidding occur, making it difficult to obtain sufficient accuracy with self-location estimation.

[0006] Therefore, an object of the present invention is to provide a farm field management system that realizes highly accurate self-location estimation of a robot moving in a farm field, using a method that is easy to adopt in agricultural fields. [Means for solving the problem]

[0007] The present invention provides the following solutions.

[0008] (1) A farm field management system comprising: a robot having a track set in a farm field, a position information identification means set along the track, and a moving means for moving along the track and equipped with an environmental sensor for collecting environmental information regarding the farm field, a self-position generation means for generating self-position information of the robot based on at least information exchanged between the robot and the position information identification means, and an information management device for managing the farm field based on the self-position information and the environmental information.

[0009] According to the above configuration, the robot moving in the field generates its own location information by sending and receiving information to and from the location information identification means provided along the trajectory of the robot. This makes it possible to estimate the robot's own location using a simple method that is easily adoptable in agricultural fields without introducing an expensive system.

[0010] (2) The farmland management system described in (1) above, wherein the information management device controls the behavior of the robot based on the self-location information and the environmental information.

[0011] According to the above configuration, the robot is controlled based on its own position information and environmental information, making it possible to realize more advanced control based on various information related to the field.

[0012] (3) A farm field management system as described in (1) or (2) above, characterized in that the self-position generation means generates self-position information of the robot based on information exchanged between the robot and the position information identification means and the amount of movement obtained from the movement means.

[0013] According to the above configuration, the self-location information of the robot is generated based on the information exchanged with the location information specifying means and the amount of movement obtained from the moving means.

[0014] This makes it possible to generate more accurate robot self-location information using a simple method that can be easily adopted in agricultural fields, without the need for expensive systems.

[0015] (4) A field management system as described in (1) or (2) above, characterized in that a three-dimensional position of the harvest is identified based on the robot's self-position generated by the self-position generating means, the position of the environmental sensor in the robot, and the position of the harvest relative to the environmental sensor acquired by the environmental sensor, and the field is managed based on the three-dimensional position of the harvest.

[0016] According to the above configuration, the three-dimensional position of the harvest is identified based on the robot's own position, the position of the environmental sensor in the robot, and the position of the harvest relative to the environmental sensor.

[0017] This makes it possible to grasp the three-dimensional position of each harvested crop using a simple method that can be easily adopted in agricultural fields, without the need for expensive systems.

[0018] (5) The farm field management system described in (1) or (2) above, characterized in that the track is a wire rope stretched in the air along the ridges of a farm field, and the robot moves along the track while being suspended from the wire rope.

[0019] According to the above configuration, the robot moves along a wire rope stretched in the air without running on the ground, so that the robot can move accurately along a trajectory, and can generate self-location information with higher accuracy without being affected by uneven and slippery ground.

[0020] (6) A farm field management system as described in (1) or (2) above, characterized in that the environmental information is collected at a number of different points in time, and the farm field is managed based on changes in the environmental information over time.

[0021] According to the above configuration, since the farm field is managed based on temporal changes in the environmental information, it becomes possible to manage the farm field based on more accurate future predictions.

[0022] (7) The farm field management system described in (6) above, characterized in that the environmental information includes growth information of the harvested crop, and the optimum harvest time of the harvested crop is estimated based on changes in the growth information over time.

[0023] According to the above configuration, the optimum harvest time is estimated based on temporal changes in the growth information of the harvested crop, making it possible to manage the farm field based on more accurate information on the optimum harvest time.

[0024] (8) The farm land management system according to (1) or (2) above, wherein the location information identification means transmits and receives information using short-range wireless communication.

[0025] According to the above configuration, the location information identifying means and the robot transmit and receive information using short-range wireless communication, making it possible to generate advanced self-location information using a cheaper and simpler method. Effect of the Invention

[0026] According to the present invention, it is possible to estimate the self-position of a robot moving in a farm field with high accuracy using a simple and inexpensive method that can be easily adopted in agricultural fields. [Brief description of the drawings]

[0027] [Figure 1] FIG. 1 is a schematic diagram illustrating a first embodiment according to the present invention. [Diagram 2] FIG. 2 is a schematic diagram illustrating a modified example of the first embodiment according to the present invention. [Diagram 3] FIG. 3 is a schematic diagram illustrating a second embodiment according to the present invention. [Figure 4] Figure 4 is an image of a robot that can be placed in a farm field. [Diagram 5]FIG. 5 is a diagram showing a flow for identifying the three-dimensional position of a harvested product. [Figure 6] Figure 6 is a 3D map visualizing the three-dimensional location of the harvested crops. [Figure 7] FIG. 7 is a diagram showing a flow of updating a 3D map. [Figure 8] FIG. 8 is a diagram showing a utilization pattern of time-series data. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0028] Hereinafter, an example of an embodiment of the present invention will be described with reference to the drawings. Note that this is merely an example, and the technical scope of the present invention is not limited to this example. In the following description of the embodiment, bell peppers will be used as an example of a crop to be harvested, but the present invention is not limited to bell peppers and can be applied to harvesting various vegetables and fruits, etc. For example, the present invention can be used to harvest vegetables and fruits such as cucumbers, tomatoes, mandarin oranges, strawberries, eggplants, and apples.

[0029] 1. First embodiment 1 is a schematic diagram illustrating a first embodiment of the present invention. A track 10 is provided in a farm field, such as a vinyl greenhouse, and a position information identification means 11 is provided along the track 10. A robot 12 moves along the track 10 and can collect various information via an environmental sensor 14.

[0030] The track 10 may be a wire rope or rail stretched in the air, or a rail or magnetic marker installed on the ground. The position information identification means 11 is a means capable of exchanging information with the robot 12, and enables contactless information exchange by transmitting and receiving radio waves, such as RFID (Radio Frequency Identifier). Specifically, it is preferable to use short-range wireless communication NFC (Near Field Communication) as RFID. An NFC tag may be installed as the position information identification means 11 in a location close to the track 10, and information may be exchanged when the robot 12 passes in front of the NFC tag.

[0031] A robot 12 moving along a track 10 is equipped with a self-position generating means 15. The robot 12 exchanges information when it passes in front of an NFC tag, which is a position information identifying means 11, and the self-position generating means 15 generates accurate self-position information of the robot by using the position information associated with the NFC tag.

[0032] The means of information exchange between the robot 12 and the location information identification means 11 does not need to be limited to communication via radio waves, and other methods such as placing a two-dimensional barcode on one side and a camera on the other side to read it, or various switches that are activated when the robot 12 passes in front of them, can be used.

[0033] The position information specifying means 11 may be provided at an end or an intermediate portion of the track 10. The self-position generating means 15 can generate more accurate self-position information of the robot 12 by acquiring not only information exchanged between the self-position generating means 15 and the position information specifying means 11, but also the amount of movement of the robot 12 along the track 10 from the moving means 13 and performing calculations based on this information.

[0034] When the position information identifying means 11 is provided at the end of the trajectory 10, the robot 12 generates its own position information at the end of the trajectory 10, obtains the amount of movement from that position from the moving means 13, and updates its own position information by taking this into consideration. Alternatively, it is also possible to set the end of the trajectory 10 as the zero point, update the self-position information based on the amount of movement obtained from the moving means 13, and obtain position information linked to the position information identifying means 11 at the timing when the robot 12 passes in front of the position information identifying means 11 located midway along the trajectory 10, thereby correcting the self-position of the robot 12, thereby updating the self-position information.

[0035] When the trajectory 10 is long, errors are likely to occur in the amount of movement of the robot 12 obtained from the moving means 13, so it is desirable to provide position information identification means 11 at predetermined intervals along the trajectory, obtain position information when the robot 12 passes in front of each position information identification means 11, and correct the robot's own position.

[0036] Calibration is required to generate the self-location information of the robot 12. Possible methods include running the robot along the entire track when it is introduced and having it learn the information necessary to generate the self-location information, such as the positions of the ends of the track 10 and the position of the location information identification means 11, or by identifying the initial position of the robot (such as the end of the track) as the zero point and calibrating every time it runs.

[0037] The robot 12 is equipped with an environmental sensor 14, which collects environmental information around the robot. The environmental information may include various information such as images including the fruit to be harvested, illuminance, temperature, humidity, and CO2 concentration. While moving along the track 10 in the farm field, the robot 12 collects various information via the environmental sensor 14, and is capable of linking this information to the robot's own position information and storing it.

[0038] The robot 12 is equipped with an information management device 16, which makes it possible to control the behavior of the robot based on the robot's own position information and environmental information. By performing such control, it is possible to realize more efficient and sophisticated work such as harvesting, fertilizing, irrigation, and spraying pesticides.

[0039] 2, the information management device 16 does not necessarily have to be mounted inside the robot 12, and can be installed outside the robot 12 while ensuring communication with the robot 12. In this case, the information management device 16 may be configured to communicate with not only the robot 12 but also the position information identification means 11.

[0040] Furthermore, the self-position generating means 15 does not necessarily need to be mounted inside the robot 12, and can be provided outside the robot 12 like the information management device 16 in Fig. 2. For example, the robot 12 can obtain the ID of the position information identifying means 11 and the time information when it passed in front of it, and transmit the ID and the time information to the self-position generating means 15, thereby generating the self-position information of the robot 12 outside the robot 12.

[0041] 2. Second embodiment 3, the robot 12 can be equipped with a harvesting hand 17. When the robot 12 is equipped with the harvesting hand 17, it becomes possible for the robot 12 to perform more optimized harvesting actions in response to commands from the information management device 16.

[0042] While moving along the track 10, the robot 12 collects data on the surrounding environment with the environmental sensor 14. The collected data is processed by the information management device 16 to identify the location of the next crop to be harvested as a priority, and the robot 12 can move to that location and perform harvesting work. The behavior of the robot 12 can be controlled to optimize the amount of crop to be harvested, the time, and the location in the field.

[0043] An image of a robot harvesting peppers is shown in Figure 4. A wire rope 10 is stretched along the furrows in which the peppers are grown. A moving means 13 is provided on the ceiling of the robot 12, and the moving means 13 is equipped with multiple pulleys through which the wire rope is passed, allowing the robot 12 to be suspended movably.

[0044] Wire rope 10 stretched across the field is supported by pillars and beams (not shown) at predetermined intervals to keep the amount of sagging of wire rope 10 below a certain value. Position information identifying means 11 are installed on the pillars and beams at predetermined intervals, and when robot 12 passes in front of position information identifying means 11, the distance between them is within 10 cm, preferably 1 to 2 cm.

[0045] The environmental sensor 14 provided on the robot 12 may employ a camera that captures images of the harvested peppers. Figure 5 illustrates the process of identifying the three-dimensional location of each of the harvested peppers from the images captured by the camera 14 to create a 3D map of the peppers.

[0046] First, camera 14 in Fig. 5 acquires an image including green peppers, and image processing device 18 identifies the position of the green peppers to be harvested from the image. The position of the green peppers that can be obtained here is the relative coordinates of the green peppers with respect to camera 14. Information management device 16 acquires position information 161 of the harvested product with respect to camera 14 from image processing device 18. Information management device 16 also acquires position information 162 of the camera with respect to the robot from control device 19 that controls the entire robot. Here, the relative coordinates of camera 14 with respect to the robot are variable values ​​that change in accordance with the movement of harvesting hand 17 when camera 14 is attached to harvesting hand 17, and are fixed values ​​when camera 14 is attached to the robot body.

[0047] The information management device 16 is able to identify the three-dimensional position 160 of the harvest based on the above-mentioned harvest position information 161 relative to the camera, the camera position information 162 relative to the robot, and the robot's self-position information 163 generated by the self-position generating means 15.

[0048] If the three-dimensional position of each harvested pepper can be identified, it becomes possible to create a 3D map of the peppers in the field, as shown in Fig. 6. The 3D map can display not only the three-dimensional position of the pepper's fruit, but also the position information of the flowers before they turn into fruit, and information on pests and diseases obtained from image information acquired by the camera 14. It is also possible to link all data related to the growth status, such as the length and width of the pepper and its growth rate, in order to identify the peppers to be harvested.

[0049] By adding prediction information about the optimum harvesting period for bell peppers (described later) to this 3D map, the information management device 16 can generate an optimal harvesting plan for the robot 12, and by instructing this harvesting plan to the robot 12, the harvesting behavior can be optimized. Furthermore, in addition to the harvesting behavior of the robot, the 3D map can also be used to optimize the harvesting behavior of workers.

[0050] The 3D map of the peppers needs to be updated from time to time as the peppers grow. The 3D map update process is shown in FIG. 7. First, in step 20, the robot or camera is at a predetermined position, and the 3D position of the peppers visible from that position is identified. Next, in step 21, the robot moves. If the camera is attached to the robot's arm or the like, the robot may not move and only the position of the camera may be moved. Then, in step 22, the 3D position of the peppers visible from the position after the movement is identified. In step 23, the 3D positions of the peppers obtained in steps 20 and 22 are compared, and the difference peppers are identified. Finally, in step 24, the information of the newly recognized peppers among the difference peppers identified in step 23 is integrated into the 3D map, thereby updating the 3D map.

[0051] Among the difference peppers identified in step 23, the ones that have disappeared (become invisible) are those that have gone out of the camera's field of view or are hidden by leaves or other objects and cannot be recognized, but because they are still fruits that exist but have simply become invisible, they will continue to be managed. However, after the harvesting process, they may be removed from the list of managed fruits as they have been harvested.

[0052] Although the fruit of a bell pepper may not be recognized by the camera 14 due to the presence of leaves between the camera 14 and the bell pepper, it is possible to generate a more accurate 3D map by identifying the three-dimensional position of the bell pepper from different locations using the procedure shown in Figure 7 and integrating this information.

[0053] Once a 3D map has been created, it will vary depending on the environment and circumstances, but it is a good idea to integrate the entire map once a day or once every few days. In this integration, the 3D map created once is compared with the 3D map created just before (the day before or a few days before), and each pepper is associated with the other. The 3D coordinates and growth information of each pepper are updated while taking into account positional deviations caused by harvesting behavior, plant growth, natural falling, etc. By repeating this overall integration, it is possible to grasp chronological information about each pepper.

[0054] In this way, it is beneficial to collect the above-mentioned information about bell peppers not only in three dimensions, but also in a time series. Figure 8 shows how to use data collected in time series. With time series data, it is possible to grasp the growth status of the harvested crop in a timely manner and predict the optimum harvest time. By combining the 3D map of bell peppers (31) with predicted information on the harvest volume and optimum harvest time (32), it is possible to optimize the harvesting behavior of the robot (33).

[0055] In addition, time-series image data can be organized as AI training data (34), and machine learning can be used to generate a trained model (35). This can then be used to quickly and accurately identify abnormalities in the harvest and notify people or suggest actions to take. (36) It is also possible to organize time-series information for each plant or area, rather than for each individual pepper (37), making it easier to determine whether the growing environment is good or bad (38), and this can be reflected in environmental control such as fertilization and irrigation (39).

[0056] Although the embodiments and modifications of the present invention have been described above, the present invention is not limited to these embodiments. Furthermore, the effects described in the embodiments and modifications of the present invention are merely a list of the most preferable effects resulting from the present invention, and the effects of the present invention are not limited to those described in the embodiments and modifications of the present invention. [Explanation of symbols]

[0057] 1. Field management system 10 orbits 11 Location information identification means 12. Robot 13. Transportation 14 Environmental Sensors 15 Self-position generation means 16 Information management device 160 3D position of crops 161 Location of crop relative to camera 162 Position information of the camera relative to the robot 163 Robot self-location information 17 Harvest Hand 18 Image Processing Device 19 Control device 2. Flow for updating 3D maps 3. Use patterns of time series data

Claims

1. a track installed in the field; a position information specifying means provided along the track; a robot having a moving means for moving along the track and having an environmental sensor for collecting environmental information about the field; a self-position generating means for generating self-position information of the robot based on information exchanged at least between the robot and the position information identifying means; an information management device that manages the field based on the self-location information and the environmental information; A field management system equipped with the above.

2. The farmland management system according to claim 1 , wherein the information management device controls the behavior of the robot based on the self-location information and the environmental information.

3. The farmland management system according to claim 1 or 2, characterized in that the self-position generation means generates the self-position information of the robot based on information exchanged between the robot and the position information identification means and the amount of movement obtained from the movement means.

4. a self-position of the robot generated by the self-position generating means; the location of the environmental sensor on the robot; Identifying a three-dimensional position of the harvest based on the position of the harvest relative to the environmental sensor acquired by the environmental sensor; 3. The farmland management system according to claim 1, wherein the farmland is managed based on the three-dimensional positions of the harvested products.

5. The track is a wire rope or rail stretched in the air along the furrows of the field, or a rail or magnetic marker installed on the ground. The farmland management system according to claim 1 or 2, characterized in that

6. 3. The farmland management system according to claim 1, wherein the environmental information is collected at a plurality of different points in time, and the farmland is managed based on changes in the environmental information over time.

7. the environmental information includes growth information of the harvested product, 7. The farmland management system according to claim 6, wherein the optimum harvest time for the crop is estimated based on the change over time in the growth information.

8. 3. The farm land management system according to claim 1, wherein the location information specifying means transmits and receives information using short-range wireless communication or a two-dimensional barcode.

Images