METHOD FOR CULTIVATING A FIELD BY MEANS OF AN AGRICULTURAL MACHINE

Description

[0001] The present invention relates to a method for cultivating a field using an agricultural machine according to the preamble of claim 1.

[0002] Agricultural machinery is becoming increasingly larger. Controlling it is a complex process, requiring the operator to monitor numerous parameters, the surrounding environment, and other factors. In this context, obstacles can easily be seen too late or misjudged. Even stopping in time before an obstacle can necessitate unwanted maneuvering. These obstacles might include well shafts or drain covers, which can be difficult to spot in dense crops.

[0003] Additionally, even in highly automated or autonomous scenarios, environmental perception by the agricultural machinery is necessary to avoid collisions.

[0004] The known prior art (EP 3 529 556 A1) relates to a method for processing a field in which a driving route is planned based on an obstacle map. Using such an obstacle map is generally advantageous for avoiding collisions with obstacles. However, a problem arises because some obstacles are changeable. For example, well shafts can be opened and trees can grow. New obstacles can also appear, or the position or properties of an obstacle may be incorrectly or inaccurately recorded in the obstacle map.

[0005] Improving upon the known state of the art in this area is a challenge.

[0006] Another known prior art (US 2019 / 302783 A1) describes a method for cultivating a field using an agricultural vehicle, wherein the agricultural vehicle has a sensor mounted on it. The agricultural vehicle has an assistance system and accesses an obstacle map. The obstacle map is updated by the sensor during cultivation. Several vehicles can operate in the field simultaneously. The map can be updated by multiple vehicles with one or more sensors as they traverse the field.

[0007] The invention is based on the problem of designing and further developing the known method in such a way that further optimization is achieved with regard to the aforementioned challenge.

[0008] The above problem is solved by the features of the characterizing part of claim 1.

[0009] The fundamental principle is that an obstacle map can be dynamically updated if a suitable sensor is provided on the agricultural machine. This reduces inaccuracies in the obstacle map, leading to fewer false warnings about obstacles or incorrect reactions to them, and consequently, fewer collisions. Simultaneously, the obstacle map is updated for subsequent field operations, offering an advantage over purely sensor-based obstacle detection. For example, other agricultural machines without sensors can then work the field more safely. In this context, it is advantageous according to the invention if the agricultural machine is part of a network of agricultural machines that share the obstacle map.This allows several agricultural machines to benefit from obstacle map updates or to be provided with an obstacle map in general.

[0010] Specifically, it is proposed that the driver assistance system update the obstacle map using the sensor while the field is being worked.

[0011] According to claim 2, it can be provided that the driver is involved in updating the obstacle map. For example, the driver can decide whether an obstacle has been correctly detected or is generally irrelevant.

[0012] Claim 3 relates to the possibility of using pattern recognition to classify obstacles into categories. Of particular interest are obstacles that are not visible or not permanently visible. Unlike, for example, a tree, these are difficult for the driver to notice without assistance.

[0013] Claim 4 relates to the ability to share the updated obstacle map in real time. This allows several of the agricultural machines in the consortium to benefit from the insights gained by the agricultural machine.

[0014] The invention will now be explained in more detail with reference to a drawing that illustrates only one embodiment. The drawing shows Fig. 1 schematically illustrates the proposed method from the perspective of an agricultural machine, and Fig. 2 shows a combination of agricultural machines working a field.

[0015] The proposed solution can be applied to a wide range of agricultural machinery 1, particularly self-propelled agricultural machinery 1. This includes tractors, especially tractors, and harvesting machines, especially combine harvesters, forage harvesters, and the like. Such agricultural machinery 1 can perform a variety of field operations, such as plowing, fertilizing, sowing, and the like. These operations sometimes require consideration of obstacles 2 in the field 3, which may even depend on the type of operation. For example, underground obstacles 2 might be relevant during plowing but not during spraying.

[0016] The embodiment shown in the figures, which is preferred in this respect, relates to a method for cultivating a field 3 using an agricultural machine 1, wherein the agricultural machine 1 has a sensor 4 arranged on the agricultural machine 1, and wherein the agricultural machine 1 has a driver assistance system 5 which includes an obstacle map 6. In principle, the sensor 4 can be, for example, a radar sensor, a camera, or a lidar sensor. Several sensors 4 can also be provided, for example, a camera and a lidar sensor.

[0017] The driver assistance system 5 is preferably located on the agricultural machine 1 and / or in the cloud. It can therefore be a driver assistance system 5 in the narrow sense, supporting the driver 7 in controlling the agricultural machine 1, but also a driver assistance system 5 in the broad sense, which may include planning modules and the like. The driver assistance system 5 comprises hardware and software.

[0018] The obstacles 2 are, and preferably are, static obstacles 2.

[0019] It is now essential that the driver assistance system 5 updates the obstacle map 6 using the sensor 4 while processing field 3.

[0020] Fig. 1This schematically shows how an old obstacle map 8 (top) becomes an updated obstacle map 9 (bottom) after the agricultural machine 1 detects a tree. This example is purely schematic. With trees, a change in their properties is to be expected, but not their spontaneous appearance. Examples of temporary obstacles 2, which are preferably not included in the obstacle map 6, are crop residues, straw bales, and the like.

[0021] In principle, the driver assistance system 5 can be configured to perform a reaction routine when approaching an obstacle 2 on the obstacle map 6, particularly when a distance threshold is undershot, and / or when a collision probability threshold determined based on a direction of travel and / or route is exceeded. The reaction routine can include stopping the agricultural machine 1 and / or performing an evasive maneuver and / or providing information to the driver 7.

[0022] Sensor 4 could be, for example, a lidar sensor, a radar sensor, an ultrasonic sensor, or a camera.

[0023] Updating the obstacle map 6 can involve adding or removing obstacles 2, but also adjusting the position of obstacles 2, their size, type, or the like.

[0024] Furthermore, it is preferably provided that the agricultural machine 1 is part of a group 10 of agricultural machines 1 that jointly work the field 3, that the agricultural machines 1 of the group 10 communicate with each other directly or indirectly, in particular via the Internet, and that the agricultural machines 1 of the group 10 share the obstacle map 6 and / or an updated obstacle map 9 with each other.

[0025] The system 10 can consist of master and slave agricultural machines or of machines with equal rights. It can be stipulated that when a new agricultural machine 1 joins the system 10 and field work begins, the obstacle map 6, in its most up-to-date version, is sent to the new agricultural machine 1. This communication can take place via direct radio or indirectly via the internet. A farm management system can also be integrated into the system.

[0026] Here, and preferably, it is therefore the case that an agricultural machine 1 recognizes a new obstacle 2 or an obstacle 2 that is not correctly placed in the obstacle map 6 and makes this available to at least one other agricultural machine 1.

[0027] If several of the agricultural machines 1 are equipped with sensors 4, it becomes possible to update the obstacle map 6 via a consensus mechanism in order to reduce erroneous updates. For example, an obstacle 2 can only be recorded once it has been detected by two agricultural machines 1. Additionally, it becomes possible to detect or classify the obstacle 2 (to be explained later), or to determine its properties, based on the sensors 4 of several agricultural machines 1. This is particularly advantageous for determining the dimensions of the obstacle 2.

[0028] Additionally or alternatively, it may be provided that the obstacle map 6 is additionally updated based on a driver input, preferably that the driver assistance system 5 detects an obstacle 2 or a property of an obstacle 2 by means of the sensor 4, displays it to the driver 7 for review and updates the obstacle map 6 depending on an input from the driver 7 in response to the display.

[0029] Thus, driver 7 retains control over what is considered obstacle 2 and can react to false positives. Alternatively, another user can make the decision instead of driver 7. This decision does not need to be made immediately. For example, the user can be provided with camera image data. In particular, the decision can be made retrospectively, preferably via a farm management system.

[0030] Furthermore, it is preferably provided that the driver assistance system 5 classifies obstacles 2 detected by the sensor 4 into different categories based on pattern recognition, preferably that the driver assistance system 5 classifies the obstacles 2 at least into invisible obstacles 2 and / or temporarily visible obstacles 2 and / or permanently visible obstacles 2.

[0031] Properties and / or reactions to obstacles 2 can be derived from the category of obstacles 2. For example, some obstacles 2, such as field remnants, can simply be driven over. The driver assistance system 5 can additionally classify obstacles 2 into the categories of permanent and temporary obstacles 2, or it can choose not to recognize temporary obstacles 2 as obstacles 2 that affect or should update the obstacle map 6.

[0032] Invisible obstacles 2 can be, for example, drainage systems, which may be detectable using the radar sensor. It is also conceivable to infer the presence of obstacles 2 indirectly, for example, through missing or altered crops in a field area. Temporarily visible obstacles 2 include, for example, drainage shafts, while permanently visible obstacles 2 include trees. These categories can also, or alternatively, categorize the obstacles 2 already included in obstacle map 6.

[0033] Furthermore, it is preferably provided that in the network 10 updates of the obstacle map 6 are communicated in real time from at least one agricultural machine 1 to at least one other agricultural machine 1, preferably that a driver assistance system 5 of the other agricultural machine 1 reacts to such an updated obstacle 2 by controlling the agricultural machine 1, in particular by changing a driving route or stopping the agricultural machine 1.

[0034] For example, a single agricultural machine 1 with a sensor 4 can also control one or more other agricultural machines 1, for example the one in Fig. 2 Shown loading vehicle 11, protect. Reference symbol list

[0035] 1 agricultural machine 2 obstacle 3 field 4 sensor 5 driver assistance system 6 obstacle map 7 driver 8 old obstacle map 9 updated obstacle map 10 combination 11 loading vehicle

Claims

1. Method for working a field (3) by means of an agricultural work machine (1), wherein the agricultural work machine (1) has a sensor (4) which is arranged on the agricultural work machine (1), wherein the agricultural work machine (1) has a driver assistance system (5) which has an obstacle map (6), wherein the driver assistance system (5) updates the obstacle map (6) by means of the sensor (4) during the working of the field (3), wherein the agricultural work machine (1) is part of a group (10) of agricultural work machines (1) which jointly works the field (3), wherein the agricultural work machines (1) in the group (10) communicate with one another directly or indirectly, in particular via the Internet, wherein the agricultural work machines (1) in the group (10) share the obstacle map (6) and / or an updated obstacle map (9) with one another, wherein at least two agricultural work machines (1) in the group (10) have a sensor (4), and wherein an obstacle (6) is only included in the obstacle map (6) when it has been detected by these two agricultural work machines (1).

2. Method according to Claim 1, characterized in that the obstacle map (6) is additionally updated on the basis of a driver input, preferably in that the driver assistance system (5) determines an obstacle (2) or a property of an obstacle (2) by means of the sensor (4), displays it to the driver (7) for checking, and updates the obstacle map (6) on the basis of an input by the driver (7) in response to the display.

3. Method according to Claim 1 or 2, characterized in that the driver assistance system (5) classifies obstacles (2) detected by means of the sensor (4) into different categories on the basis of pattern recognition, preferably in that the driver assistance system (5) classifies the obstacles (2) at least into invisible obstacles (2) and / or temporarily visible obstacles (2) and / or permanently visible obstacles (2).

4. Method according to one of Claims 1 to 3, characterized in that updates to the obstacle map (6) in the group (10) are communicated in real time from at least one agricultural work machine (1) to at least one further agricultural work machine (1), preferably in that a driver assistance system (5) of the further agricultural work machine (1) reacts to an obstacle (2) updated in this way by controlling the agricultural work machine (1), in particular changing a driving route or stopping the agricultural work machine (1).