A method for generating a cultivation plan and a method to cultivate a piece of farmland

Abstract

The invention pertains to a method for generating a cultivation plan for an autonomous agricultural vehicle in order to process a piece of farmland, the vehicle comprising an implement for performing a row creation process to create multiple rows on the piece of farmland, which cultivation plan comprises multiple distinct paths that extend over the piece of farmland, the method comprising providing a processing unit to generate the cultivation plan, providing input data to the processing unit, on the basis of which input data the processing unit calculates the position and spatial extension of each of the multiple distinct paths on the piece of farmland, the input data comprising farmland data and vehicle data, wherein in the method, in addition to the farmland data and vehicle data, future data that pertain to one or more follow up processes that are due to be performed on the piece of farmland after the said row creating process has been performed, is used to calculate the position and spatial extension of each of the multiple distinct paths on the piece of farmland. The invention also pertains to a method to cultivate a piece of farmland by performing multiple consecutive agricultural processes on the piece of farmland, the first one of these multiple consecutive agricultural processes being a row creation process to create multiple rows in the piece of farmland.

Description

[0001] A METHOD FOR GENERATING A CULTIVATION PLAN AND A METHOD TO CULTIVATE A PIECE OF FARMLAND

[0002] GENERAL FIELD OF THE INVENTION

[0003] The present invention pertains to a method for generating a cultivation plan for an autonomous agricultural vehicle in order to process a piece of farmland, the vehicle comprising an implement for performing a row creation process to create multiple (typical parallel) rows on (i.e. extending on, over or in) the piece of farmland, which cultivation plan comprises multiple distinct paths that extend over the piece of farmland, the method comprising providing a processing unit to generate the cultivation plan, and providing input data to the processing unit, these input data comprising farmland data and vehicle data, on the basis of which input data the processing unit calculates the position and spatial extension of each of the multiple distinct paths on the piece of farmland. The invention also pertains to a method to cultivate a piece of farmland by performing multiple consecutive agricultural processes on the piece of farmland, the first one of these multiple consecutive agricultural processes being a row creation process to create multiple rows in the piece of farmland.

[0004] BACKGROUND ART

[0005] The adoption of technology in agriculture has improved the approaches that farmers use in the farmland nowadays. Modern agriculture has made it easy for farmers to achieve high produce while using less input, in particular less labour. According to the trends in the use of technology in agriculture, there are high concerns whether or not the future of agriculture is bright. For example, mechanization in agriculture has reduced the overuse of manpower in doing some of the farming activities. As a consequence, agricultural machines have become bigger and bigger and more dedicated towards performing one type of cultivation. The introduction of autonomous agricultural vehicles, such as an autonomous tractor that is operatively connected to an agricultural implement such as aplough, is considered a next step into the future of farming and it is expected that using autonomous vehicles there is more freedom to cultivate the land using even less labour.

[0006] Self-driving cars are common these days. Based on the trends in regards to the advancement of technology, it is expected that the technology will also be used on a wide scale for cultivating farmland. At present farmers in advanced countries are giving a tactical approach to how they plant, harvest, as well as maintain their crops. A good example of new tactical approaches is the use of autonomous vehicles in agriculture. The concept of autonomous vehicles (form now on also denoted as autonomous tractors) can be traced back prior to the introduction of the concept of precision farming in the eighties. During these days, farmers used GPS technology as a guide to the tractors driving across the farmland. The aim of such an approach was the reduction of fuel consumption and enhancing the efficiency of the tractors and the farming activities. As such, these initial steps formed the basis for the development of autonomous tractors, following the introduction of technologies that improved communication over wireless devices. Autonomous tractors employ much the same approach as the driverless vehicles, i.e. using advanced control systems and sensors. With the inclusion of auto-steering abilities, such tractors have added control abilities. Evidently, the launch of the autonomous tractors is considerably a manifestation of the extended use of this technology in farming.

[0007] Benefits to farmers are obvious. It is an undeniable fact that farming is not an easy undertaking, it involves working for long hours and the subscription to hard labour in harsh weather conditions. Taking into consideration the common state of farmers, the majority of them have no employees to task them in the farmland and hence, have to do everything all by themselves, the autonomous tractors obviously can be a positive outcome. Next to this, accuracy and precision are important aspects in agriculture in various aspects such as planting. All in all, the use of such tractors may lead to higher return on investment since accuracy is enhanced.

[0008] It is generally recognised that data plays a significant role in determining the farmers’ decisions. Usually, the absence of clear and reliable data can interfere with the decisions farmers make, and subsequently, have adverse impacts on the amount of outcome obtained from the fields. There are diverse sources and types of data that a farmer needs to succeed in their farming activities. For example, data on soil is important in that it helps farmers in determining what crops will do well in a given pieceof land by establishing the moisture content, and the amount of nutrients. The autonomous vehicles are typically be fitted with various sensors that can be used in the collection of data on the conditions of the soil, and hence, offer a platform for improving the outcome of the available crops. The elimination of the human interaction in farming following the use of autonomous vehicles may thus be advantageous. Stressed employees cannot achieve the required efficiency level in the fields. Similarly, it is often hard for humans to manage diverse tasks on the farm especially where a large farmland is involved. Autonomous vehicles have the appropriate sensors to offer the necessary help in the management of a several tasks in the farmland hence reducing stress and the workload in the farm.

[0009] The autonomous agricultural vehicles run on high level technology that can be used in gathering high profile information. For example, some models have automatic steering abilities and GPS or GNSS technologies which enhance the control of the vehicles’ course. The advanced sensors come in handy in the determination of soil moisture level, activities around planting and harvesting, present yield, as well as the amount of fuel needed for a given area of land. Additionally, other models of autonomous vehicles can guide farmers on how to apply fertilizers.

[0010] Autonomous agricultural vehicles allow precise control of work and farm equipment by at least minimising or even ruling out human errors. It makes it possible for farmers to extend their working hours. The sensors fitted in the vehicles can guide it in the right course even in conditions of reduced visibility and at night: work continues even during windy, dusty, and foggy conditions. Additionally, the ability of the vehicles to reduce workload and stress on employees comes in handy in increased working hours in a day since the farmer has a greater flexibility in the management of growing tasks.

[0011] It has thus become a common understanding that the best way for using an autonomous vehicle is to establish a cultivation plan for the piece of farmland, which plan comprises generating multiple paths that spatially extend over the piece of farmland along particular coordinates (such that their position and spatial extension is unambiguously determined), and after the plan has been generated, controlling the autonomous vehicle such that it crosses the farmland by moving over each of the multiple paths autonomously. Also, the way the actual agricultural implement (which may be an integral part of the vehicle or coupled thereto) is operated (for example its driving speed, its height with respect to the land, its angle with respect to the land etc.),may be controlled autonomously, for example using the sensor technology to adapt the predetermined plan to the particular circumstances of the moment in time the land is actually cultivated. Such circumstances can for example be objects that were not present at the piece of land at the time the plan was made, the weather conditions, etc.

[0012] In particular US2023 / 0284548 and US2023 / 0284549 (both assigned to Krone Agricultural and Lemken GmbH & Co, KG), disclose a way to automatically generate a cultivation plan, needing as little input form a human operator as possible, wherein based on particular input data, such as farmland data (also called field data), data regarding the agricultural vehicle (so called vehicle data), or any other data, the processing unit automatically calculates for the multiple distinct paths, a) the positioning of each of the paths on the piece of farmland, b) the direction in which the paths extend over the piece of farmland (i.e. the spatial extension, also described as the orientation of the paths on the land) and c) the order in which all of the multiple distinct paths are to be crossed by the autonomous vehicle.

[0013] This is a convenient and consistent way to determine which route the agricultural vehicle needs to take for cultivating the land, minimising or even excluding human interference and thus minimising operator time needed and potentially even excluding human error. For an operator, a fully automatic determination of the multiple distinct paths, their position, orientation and order in which these paths are to be crossed, thus in particular means operator interference and time is saved, which is the essence of having land cultivated autonomously.

[0014] A particular type of farming is row crop farming, involving the cultivation of crops in rows to facilitate efficient planting, growth, maintenance and harvesting. Common row crops include corn, soybeans, cotton, wheat, berries, potatoes, beets, each requiring specific practices to optimize yield and quality. Row crop farming allows for optimal use of available land. By organizing crops in rows, farmers can maximise planting density while ensuring each plant has adequate space to grow, receive sunlight, and absorb nutrients. One of the key advantages of row crop farming is the ability to use various machinery for various farming activities. Planting, cultivating, weeding, pest management, and harvesting can be mechanised, reducing labour costs and increasing efficiency. This mechanisation also allows for more precise application of inputs like fertilisers and pesticides, enhancing crop health and yield. That is why row crop farming is highly suitable for autonomous machinery, not only for the row creation process itself (i.e.planting, making beds, laying lines, etc), but also for one or more of the follow up agricultural processes, such as irrigation, spraying, pruning, weeding, harvesting etc.

[0015] Since the exact layout and positioning of the rows is known, this can be taken into account when (semi-) autonomously cultivating the farmland after the row creation process. This is described in US patent 11,212,954 (assigned to Deere & Company). The patent discloses methods, apparatus, systems and articles of manufacture for field (i.e. farmland) operations based on historical field operation data. An example apparatus disclosed therein includes a guidance line generator to generate a guidance line (i.e. a cultivation path) for operation of a vehicle during a second operation on a field, the guidance line based on (1) farmland data, i.e. a field map generated from location data collected during a first operation in the field, the field map including a plurality of crop rows and (2) vehicle data, in partiuclar the implement of the vehicle which performs the second operation on the field. The example apparatus further includes a drive commander to cause the vehicle to (semi-) autonomously traverse the field along the guidance line, and an implement commander to cause the implement to perform the second operation as the vehicle traverses the field along the guidance line. This way, the exat position and spatial extension of the rows is taken inot account when performing a follow up process, this in order to minimise damage to the row, e.g. damage to the crop, bed, irrigatng line, etc.

[0016] OBJECT OF THE INVENTION

[0017] It is an object of the invention to devise an improved method for generating a cultivation plan for an autonomous agricultural vehicle in order to process a piece of farmland, in particular for a vehicle comprising an implement for performing a row creation process to create multiple rows on the piece of farmland. It is a further object to devise an improved method to cultivate a piece of farmland by performing multiple consecutive agricultural processes on the piece of farmland, the first one of these multiple consecutive agricultural processes being a row creation process to create multiple rows in the piece of farmland.SUMMARY OF THE INVENTION

[0018] In order to meet the object of the invention, a method as described here above in the General Field of the Invention section has been devised, i.e. a method for generating a cultivation plan for an autonomous agricultural vehicle in order to process a piece of farmland, the vehicle comprising an implement for performing a row creation process to create multiple (typical parallel) rows on (which can also be denoted as “over”, “extending over”, or even “in”) the piece of farmland, which cultivation plan comprises multiple distinct paths that extend over the piece of farmland, the method comprising (as known from the prior art) providing a processing unit to generate the cultivation plan, and providing input data to the processing unit, on the basis of which input data the processing unit calculates the position and spatial extension of each of the multiple distinct paths on the piece of farmland, the input data comprising farmland data and vehicle data, the method being improved in that in the method, in addition to the farmland data and the vehicle data, future data that pertain to one or more follow up processes that are due to be performed on the piece of farmland after the said row creating process has been performed, is used to calculate the position and spatial extension of each of the multiple distinct paths on the piece of farmland.

[0019] The invention is based on the recognition that while the prior art method as known from US 11 ,212,954 is simple and adequate for row crop cultivation of a piece of farmland, there is room for substantial improvement. In the known method, for any follow up process, historical data is taken into account: what has been done in the past, is taken into account for the present processing of the land. However, it may be that there are constraints that cannot be matched, i.e. past events (typically: where the rows are situated) that lead to a situation that cannot be adequately matched with a present event (e.g. a vehicle and implement with a certain wheel base and working width). To illustrate: if the rows have a distance of 100 cm, and in a follow up process a particular implement is intended to be used that has a standard working width of 270 cm, having three adjacent spraying lines at a distance of 90 cm, the farmland cannot be worked upon using the full working width of the implement without accepting a less efficient process, or even (some) damage to the rows. The solution could be to use only one of the spraying lines of the implement (typically the middle line), to search for a different implement with the right working width, or to (temporarily) adapt the implement by mechanic alteration. However, whatever solution is chosen, there is a loss of efficiency. This loss can increase substantially, knowing that during a growing season, often manycultivation processes need to be carried out, and there may be a mismatch with all or a substantial part of the corresponding implements with the position and spatial extension of the rows.

[0020] In the current method, such loss can be prevented or at least decreased substantially by using future data when creating the rows (instead of using historical data, inherently accepting that the position and extension of the rows is carved in stone). When assessing the future processes, the position and spatial extension of the rows when created, can be matched to these processes in order to make the best possible use of the corresponding implements in the future. This also makes the transition to full autonomous cultivation (thus including autonomous cultivation for the follow up agricultural processes) easier since it is known beforehand that the position and spatial extension of all rows can be matched with the cultivation paths. It may be that this is at the loss of some crop density, but this loss can be compensated by the more efficient operations during the growing season. What type of future data is taken into account is not essential to the invention in its broadest sense, the gist is that such data is simply taken into account when creating the rows, such that the position and spatial extension of the rows matches the one or more follow up processes. Which data is relevant for the position and spatial extension of the rows depends on the type of vehicle and type of implement. For example, if the vehicle for a future process is a tractor with an adjustable track width, and the broadest width is even larger than the largest potential distance between the rows, the track width of this vehicle does not need to be taken into account at all for calculating the row distance. However, if the track width of this vehicle is fixed, the data regarding this width are highly relevant for the calculation. For a skilled person it is clear which data regarding a follow up process are relevant for calculating the row position and spatial extension (and therewith the row distance).

[0021] It is noted that US 2023 / 0371417 (assigned to Deere & Company) recognises the same problem, namely that of subsequent vehicles traveling over the filed, may damage planted crop. However, the solution proposed differs from the current solution: in the ‘417 patent application it is proposed to determine an exclusion zone for the planter to prohibit planting, such that this zone (a path free of plants) can be used by any subsequent vehicle operation to travel across, and thus, prevent that plants are damaged. Such a zone may be a headland zone, and / or an unplanted field path.

[0022] US 2024 / 0172581 (assigned to Yanmar Holdings Co), describes a method wherein atarget route of a first work vehicle can be made in consideration of a work range (i.e., an operation width) of a second work vehicle, it may be that the first target route lies outside of the actual working field. There is no disclosure of the actual path planning of a row creation operation, thus the calculation of the position and spatial extension of the the paths needed for the creation of the rows on the piece of farmland, based on 1) the farmland data, 2) the vehicle data, and 3) future data that pertain to one or more follow up processes that are due to be performed on the piece of farmland after the said row creating process has been performed.

[0023] The invention is also embodied in a method to cultivate a piece of farmland by performing multiple consecutive agricultural processes on the piece of farmland, the first one of these multiple consecutive agricultural processes being a row creation process to create multiple rows in the piece of farmland, the method comprising providing an autonomous agricultural vehicle which comprises an implement that is able to perform the row creation process, providing a processing unit to generate a cultivation plan for performing the row creation process on the piece of land, which plan comprises multiple distinct paths that spatially extend over the piece of land, providing input data to the processing unit, on the basis of which input data the processing unit calculates the position and spatial extension of each of the multiple distinct paths on the piece of farmland, the input data comprising farmland data and vehicle data, controlling the autonomous agricultural vehicle such that it autonomously crosses the piece of farmland land by moving over the said multiple distinct paths, while performing the row creation process, wherein in the method, in addition to the farmland data and the vehicle data, future data that pertain to one or more follow up processes of the multiple consecutive agricultural processes, which one or more follow up processes are due to be performed on the piece of farmland after the said row creating process has been performed, is used to calculate the position and spatial extension of each of the multiple distinct paths on the piece of farmland.

[0024] DEFINITIONS

[0025] An agricultural vehicle is a vehicle that is used cultivate land, typically composed of a tractor and a coupled agricultural implement, pulled, pushed or carried by the tractor which is used to generate the actual energy to propel the complete vehicle and cultivatethe land. However, the implement and tractor may also be constituted as one single unit, such as a combine harvester. The vehicle typically has a gasoline or electric engine and large (rear) wheels or endless belt tracks (so called caterpillar tracks).

[0026] An autonomous vehicle is a vehicle that can move over a piece of land according to a predetermined cultivation plan without a human operator controlling (i.e. actively running) its instant movement. Such a vehicle is typically able to automatically perceive its environment, make decisions based on what it perceives and recognizes, and then actuate a movement or manipulation within that environment. These decision-based actions may include, but are not limited to, starting, stopping, and manoeuvring around obstacles that are in its way. Such a vehicle can cross land without needing continuous control of a human operator, and thus is able to autonomously cultivate the land. It is not excluded however, that the movement of the autonomous vehicle is monitored by a human operator, for example in order to meet local safety legislation. Such monitoring can for example take place by an operator taking place on the vehicle, or monitoring the vehicle at the land from an observation post, or from a remote location via one or more cameras.

[0027] An implement is the machine or tool used to carry out agricultural processes on a piece of farmland. An implement can be a self supporting machine pushed forward, carried or trailed behind an agricultural vehicle such as a tractor, or it can be an integral part of such a tractor (such as for example a combine harvester, a self-propelled sprayer, a self-propelled planter etc).

[0028] A row in a piece of farmland is an elongated discontinuity in the piece of farmland, which typically (but not necessarily) extends from one end of the piece of farmland to another, and is an essential part in the whole process of cultivating. Typical rows are plant rows (i.e. a line of plants that together form a row of crops, or (semi-) permanent farmland assets such as beds, drainage pipes, irrigation lines, growing lines, growing racks, etc.

[0029] Farmland data (also called field data) is data that represents properties of a piece of farmland, such as for example its dimensions, its contour lines, obstacles in the land (such as for example a tree, a poles, a well etc.), the local quality of the ground, the local wetness, the presence of crops etc.

[0030] Vehicle data is data that represents properties of a particular vehicle, such as forexample its outer dimensions, weight, type of wheels, wheel base, working width, implement type, implement width, type of engine, power of that engine etc.

[0031] A path of a cultivation plan is a line along which a vehicle crosses a piece of farmland from one end to the other (not excluding that the path does not start or end adjacent an actual boundary of the land). Such a path is typically denoted as a wayline in the art of cultivation, and typically is the route to be taken for a single crossing of the piece of land. Such a path may be straight, but can also be (partly) curved, depending mainly on the shape of the piece of land and the most optimal way of crossing the complete land. Neighbouring (or adjacent) paths are typically connected by turning paths.

[0032] Spatial extension is the geographic area over which something extends.

[0033] For lines to extend in parallel, means that the lines along their length in essence keep the same distance to each other. Parallel lines are not necessarily straight lines, they may be curved or include curves.

[0034] When an event is due, this means that it is expected to happen in the future, which does not exclude that in practice it will not happen at all.

[0035] A growing season of a crop is the period of the year when the crop grows successfully from seed or seedling until harvest. The length of a growing season varies from place to place.

[0036] When two processes are consecutive, this means the processes follow one after another, which may be without any waiting period in between, or with a period of interruption.

[0037] A processing unit is the part of a computing system (which may be a local system or a distributed system) that performs logical and arithmetical operations on data as specified in the instructions for this unit. A processing unit is generally composed of hardware (one or more processors) and instructions programmed therein.

[0038] To calculate means to determine or ascertain by mathematical methods such as by using a computer. To establish a value by calculation may be as simple as determining by a processing unit what an input value is and using that value as such for furtherprocessing.

[0039] To manually determine means that a human person by acts of its own makes a determination.

[0040] An option is one thing that can be chosen from a set of possibilities.

[0041] Farmland is land that is used for or suitable for farming. A piece of farmland is a part or the totality of a farmland plot.

[0042] A human operator of a machine or device is a real-life person that has the skills to control this machine or device.

[0043] Cultivating is the act of preparing land for improving its properties, in particular for growing something, improving the growth, or harvesting thereof, especially crops. Typical cultivation acts include tillage, seeding, fertilising, spraying, harvesting etc. but also simply moving from one point to another point on the land when this is part of the said act.

[0044] A cultivation plan for a vehicle to cultivate a piece of land, is a plan which defines at least the position, direction and speed of the corresponding agricultural vehicle when crossing the land such that the land in essence can be cultivated completely.

[0045] Automatically means without the need of (human) operator intervention. The term automatically does not exclude that something is operator initiated or operator stopped as long the process can be completed without needing operator intervention.

[0046] A user interface (Ul) is the space where interactions between a human operator and a machine occurs. The goal of this interaction is to allow effective operation and control of the machine from the human end, while the machine simultaneously feeds back information that aids the operators' decision-making process. A Ul typically includes hardware such as for example a display screen, a keyboard, a hand held controller, a mouse, a smart phone and the appearance of a desktop. A user interface interacts with one or more human senses, typically via touch (a tactile Ul), sight (a visual Ul) and sound (an auditory Ul), but also smell (an olfactory Ul), balance, (an equilibria Ul), and taste (a gustatory Ul) are options for a Ul.EMBODIMENTS OF THE INVENTION

[0047] In a first further embodiment of the method according to the invention, the future data comprise data of a vehicle that is used to perform the one or more follow up processes. This vehicle may be the same autonomous agricultural vehicle as used during the row creation process (e.g. an autonomous tractor) but may also differ. If it is the same vehicle, than typically the implement coupled thereto is different, or at least a different implement is being used to perform the agricultural process.

[0048] In a second embodiment of the method according to the invention, in which embodiment the multiple rows are plant rows and / or farmland asset rows (e.g. beds, irrigation lines, growing lines etc), the position and spatial extension of each path on the piece of farmland are calculated in order to minimise damage done to the multiple rows during the performance of the one or more follow up processes. For the invention in its broadest sense, the criterion for optimising the overall cultivation process is not fixed. This can be optimising harvest, cultivation time, labour, use of land, use of consumables (such as for example fertiliser, crop protection products, water) etc. However, in this second embodiment, the criterion is minimising damage done to the multiple rows during the performance of the one or more follow up processes. This means that the position and spatial extension of the rows is such that that during the one or more follow up processes the negative impact on the actual rows (i.e. the damage done) is minimal. So it is acceptable that some damage cannot be prevented, e.g. a vehicle crossing one or more rows, as long as the overall planned damage is minimised. Preferably however, the position and spatial extension of each path on the piece of farmland are calculated to prevent damage done to the multiple rows during the performance of the one or more follow up processes. This means that the rows are created such that in theory no damage needs to be done to any of the rows during the one or more follow up processes. However, in practice this cannot be guaranteed, for example due to an accident, implement drift, a vehicle slipping and sliding due to the weather conditions, an emergency stop etc. Thus, “to prevent” in this embodiment means to plan, with the expectation that it does not happen.

[0049] In yet a further embodiment of the method for generating a cultivation plan according tothe invention, one or more follow up processes are performed by one or more autonomous agricultural vehicles. It has appeared that the current invention is ideally suitable for such methods of cultivation, since one does not expect that operator intervention is needed.

[0050] In yet again a further embodiment of the method for generating a cultivation plan according to the invention, in each of the row creating process and the one or more follow up processes the same autonomous agricultural vehicle is used, in each of the said processed combined with an implement that is constituted to perform the corresponding process. Using one and the same autonomous agricultural vehicle, such as for example the Agxeed AgBot 5.115T2 (available from Agxeed, Grubbenvorst, The Netherlands), means that the overall costs for cultivating the land can be minimised.

[0051] In still a further embodiment of the method according to the invention, the future data pertain to one or more follow up processes that are due to be performed on the piece of farmland during one growing season of a crop. In this embodiment, follow up processes are only taken into account of they are due to be performed during the growing season, i.e. until harvest of the crop. What happens after the growing season is not taken into account.

[0052] In an alternative embodiment of the method according to the invention, the future data pertain to one or more follow up processes that are due to be performed on the piece of farmland until the said multiple rows are removed from the piece of farmland, e.g. by harvesting (of a row of crops), levelling (e.g. of a bed) or otherwise taken away (e.g. a drip line) etc. “Until” in this sense means that the removing process itself may also be included for calculation in the row creation process. This embodiment is suitable for crop rows wherein the rows are kept intact for more than one growing season, such as for example when growing asparagus, berries and other fruits.

[0053] In still again another embodiment of the method according to the invention, the input data input for the processing unit are either actively put in by a human operator, and / or automatically retrieved from a digital source (such as a memory, which could be part of the processing unit, or the internet - for example the weather conditions or forecast- , or for example generated by artificial intelligence). It is foreseen that some data are put in actively by an operator, e.g. a remote operator planning, controlling and overseeing the complete cultivation process, or are simply taken form a data base, e.g. a localdatabase, the internet, the cloud (a distributed collection of servers that host software and infrastructure), etc.

[0054] In yet again another embodiment of the method according to the invention, in which embodiment a user interface (Ul) is provided for communication between a human operator and the processing unit, the Ul includes a display that visually shows the multiple distinct paths to the human operator. This increase the sense of control for the operator and minimises the risk for planning any (what is regarded in hindsight as) unwanted cultivation.

[0055] In an embodiment the processing unit is operatively coupled to the autonomous agricultural vehicle (which may be via a wireless connection, such as a 4G or 5G network). This means that the unit and vehicle can communicate for optimal cultivation. Preferably, the processing unit is permanently operatively coupled to the autonomous agricultural vehicle while processing the piece of farmland.

[0056] It has found to be advantageous if the input data themselves includes a routing pattern. This means that the processing unit calculates the actual order for the multiple distinct paths using this pattern. Such a pattern for example can be a pattern wherein every other path is skipped, or two or more paths are skipped, where neighbouring paths are crossed in opposite directions etc. Preferably, there are a number of default patterns stored in a memory of the system and the human operator selects the routing pattern out of a plurality of such stored routing patterns.

[0057] Preferably, the autonomous agricultural vehicle comprises an autonomous tractor and operatively coupled thereto an agricultural implement that performs the agricultural operation. The implement can be chosen from a group of multiple different agricultural implements.

[0058] It is noted that any and all embodiments as described here above or exemplified here after in the examples section for the method to generate a cultivation plan according to the invention, can also be embodied in the method to cultivate a piece of farmland.

[0059] The invention will now be further illustrated using the following specific examples.EXAMPLES OF THE INVENTION

[0060] Figure 1 schematically shows a system implementing the present invention.

[0061] Figure 2 schematically shows a human operator interacting with the system.

[0062] Figure 3 schematically shows Ul options for a method according to the invention.

[0063] Figure 4 schematically shows how a cultivation plan is displayed on a Ul.

[0064] Figure 5 schematically shows paths for various subsequent agricultural processes performed on a piece of farmland.

[0065] Figure 1

[0066] Figure 1 schematically shows a system 1 configured to perform the methods according to the invention. The system 1 has a central processing unit 2, that is operatively coupled (wireless connection) to local processing unit 11 of vehicle 10 (schematically depicted as a dashed box; such a vehicle is commonly known in the art and for example depicted in more detail for example in WO2023 / 191616)). Note that any connection between electronic components as indicated in figure lean be wired or wireless as commonly known in the art. The system 1 further comprises a memory unit 3 that holds data regarding a piece of land to be cultivated (such as the GPS coordinates, the type of soil, objects in the piece of land, etc), the vehicle (lists with tractor types and implement types), a list with (standard) routing options and all kinds of other data than can be used as input data such that the CPU 2 can generate a particular cultivation plan.

[0067] Unit 4 is a unit that is able to retrieve data from the internet, such as the weather force cast or any other conditions that apply during the period of time estimated to be needed for cultivating the land. The CPU 2 is connected to a desk top computer 5 that can be used by a human operator of the system to input various data needed for planning the cultivation of the piece of farmland, for example by having the operator making choices from the lists as stored in memory unit 3, by putting in additional data, and by retrieving data from the internet. This process as such is known from the art. For example, a process as described in US2020033143 in conjunction with figures 6-11 therein. This way, the system 1 is able to generate a cultivation plan for the vehicle 10, taking into account all input data.

[0068] For the actual operation to perform this plan, the local processing unit of vehicle 10 isconnected to engine 12 and steering unit 13. The cultivation plan as generated by system 1 is stored in unit 14 and may be adapted when needed for example when a sensor picks up an object (e.g. a fallen tree) or person standing in the way of the vehicle when crossing the land. The local processing unit 11 is able to let the vehicle perform this plan via control of the engine 12 and steering unit 13.

[0069] The generation of the cultivation plan comprises the determination of multiple distinct paths that extend over the piece of farmland, their position, their spatial extension and an order in which the paths are to be crossed by the autonomous vehicle. Based on the available information, the CPU 2 receives the minimal required data for generating a cultivation plan via unit 3, and generates a travel route that fits to this field and the cultivation needed. This travel route may be generated automatically based on basic, initial parameters entered via units 3 and 4, and / or based on input parameters substantially defining a travel route entered by an operator via computer 5.

[0070] Figure 2

[0071] Figure 2 schematically shows a human operator interacting with the system via the desk top computer 5. The human operator 15 that is standing behind desk top computer 5, for example at a central office of a large farm (remote from the piece of farmland), having a display 50 as user interface. This user interface allows communication between the human operator and the system 1 (as with any display): the human operator can provide all kinds of input data, for example via keyboard 51 or mouse 52, and is provided with a schematic view of the resulting cultivation plan on the display 50. This way and operator can see a schematic representation of the cultivation plan that has been generated by system 1 (see figure 4).

[0072] Figure 3

[0073] Figure 3, having sub figures 3A and 3B, schematically shows what is shown to a human operator on the display 50 when working through a program to input the data needed for the system 1 to generate a cultivation plan for a piece of farmland, which plan comprises multiple distinct paths that spatially extend over the piece of farmland in parallel. In figure 3A, indicating on the Ul 50’, is the part of the program title “Plan generation”, comprising the display of three distinct setting options, namely the “T ractor setting” 55, the “Field setting” 56 and the “Path setting” 57. These settings can be chosen one afteranother to make sure the system gets the desired input data regarding the type of tractor and implement (setting 55), the data regarding the farmland to be cultivated (setting 56) and data regarding the routing (setting 57). After this, by hitting the Calculate button 100, the system will generate a cultivation plan, including plan multiple distinct paths that spatially extend over the piece of farmland in parallel, and for these multiple distinct paths, a) the positioning of each of the paths on the piece of farmland, b) their spatial extension on the piece of farmland and c) the order in which all of the multiple distinct paths are to be crossed by the autonomous vehicle. It is also possible to skip one or more of these settings 55, 56 and 57, that is, if the system has default input values for the data required to determine a cultivation plan.

[0074] In figure 3B, what is shown on the display 50” is the box an operator sees when he hits the Path setting button 57. There is a Routing pattern box 58, which provides the option to select via a dropdown menu 59, various standard (predetermined) routing options such as “consecutive”, “alternating”, “Skip N” etc). Box 60 indicates the option for manual wayline (path) determination. As a default, this option is not activated (so the system automatically generates the order for all of the waylines in this default setting). If the toggle button 61 is put to the right, and thus the option for manual wayline determination is chosen, the operator will be provided the opportunity to manual determine the order for one or more waylines, using the Ul 50. This will not be further described in this set of examples.

[0075] A third box shown ion display 50”is box 62, denoted “Asses follow up?”. As a default, this option is not activated (so the system automatically generates the cultivation plan in this setting, not taking into account any follow up processes). If the toggle button 63 is put to the right, and thus the option for assessing follow up processes is chosen for making the cultivation plan, the opportunity is provided to put in, using the Ul 50, data regarding any such future processes. This is in particularly useful when the current cultivation plan includes a row creation process, e.g. planting rows of bean bushes. After putting the toggle button to the right, the operator is provided for the intended follow up processes (which may be a default set of processes, or a selection hand picked by the operator, or a mixture of bot, e.g. a default set amended by an operator), the choice of again putting in data for the vehicle (“Tractor Setting”), for the farmland (“Field Setting”) and for the paths (“Path Setting”), as indicated in figure 3A. These data will be taken into account for optimising the row creation process of the actual cultivation plan that is being generated, i.e. the planting of the bean bushes. There is no limit forthe number of follow up processes that is taken into account for the current cultivation plan for the creation of the rows. This is further illustrated with reference to figure 5.

[0076] Figure 4

[0077] Figure 4 shows how a plan in general (thus not restricted to a plan for a row creation process), including multiple distinct cultivation paths 75 (numbered 1-13) on a piece of farmland 70, is displayed on the Ul after the calculate button 100 (see figure 3A) is hit by the operator. In this case, the plan is generated while the option for manual wayline determination is not exercised. As can be seen, the piece of farmland is in essence rectangular, with one corner being cut along inclined border 71. Using all of the input data regarding the type of tractor and implement, the field, the weather, the direction in which the first line should be crossed (in this case from headland section 73 towards headland section 72; chosen by the operator) etc, a plan is generated which includes the multiple distinct paths, the positioning of each of the paths on the piece of farmland, the direction in which the paths extend over the piece of farmland and the order in which all of the multiple distinct paths are to be crossed by the autonomous vehicle. In the present case, the plan is such that the vehicle (tractor plus implement) travels over the piece of farmland in the shortest direction, starting with a first path on the left and working to the right by crossing in each case over a neighbouring path in the opposite direction, such that the whole of the land is cultivated. The paths are connected via sharp turns 80 that run on the headland sections 72 and 73. This is the automatically generated plan that the human operator is presented on the display 50.

[0078] Figure 5

[0079] Figure 5 schematically shows paths for various consecutive agricultural processes performed on a piece of farmland. The first one of these processes (not excluding that this first process is preceded by any other processes such as ploughing, levelling and fertilising) is the process of planting bean bushes, leading to parallel rows of bean crop. The vehicle used for this process is the autonomous AgXeed AgBot 5.115T2, with a coupled automatic seedling transplanting implement such as known form EP 3556194 (assigned to Geurts Investment BV), able to create four parallel rows of bean bushes at a distance between the rows of 20 cm to 1.5 metres. Would any of the future processes that will take place on the piece of farmland, such as spraying, weeding, harvesting the beans and removing the bushes not be taken into account, the distance between therows of bean bushes would for example be set at 65 cm (a common minimum for bean bushes), at the choice of the operator, when striving for maximum crop density.

[0080] However, in the present case the human operator (see figure 2) responsible for the process of generating a cultivation pan for the autonomous vehicle (Tractor + Transplanting machine), is aware of a few future follow up processes that will take place on the piece of farmland to cultivate the land to try and maximise the harvest. In this case, these follow up processes include 1) multiple times spraying of water using a particular 12-row sprayer with a fixed wheel base of 240 cm and variable spray head positions, 2) at least two times spraying of a crop protection product, in this case using an 8-row sprayer with a variable tack width of 160-300 cm, but with fixed spraying heads, 3) weeding with a small 1-row weeder having a fixed width of 40 cm, 4) a 4-row bean harvester with a variable track width and adjustable harvest lines (also in the width direction), although only marginally adjustable between 60 and 90 cm, and 5) a 2-row bush harvester with an adjustable track width between 80 and 110 cm and fixed harvesting width of 170 cm.

[0081] By taking into account all of these future follow up processes (via activating box 62, see figure 3B) that need to take place during the growing season up and until the removal of the bean bushes after harvest, the rows can be created that run at a distance (i.e. the distance in a direction perpendicular to the direction in which the rows extend) that matches both the track width of each of the vehicles (AgBot plus Implements, such that as less rows as possible, preferably none, get overrun during any of the follow up processes, and the process widths (i.e. where the liquids can be sprayed, what width of land can be worked for weeding, the harvest width for both the beans and the bushes), such that all bean bushed can be optimally processed (sprayed, weeded under and harvested). In the present case, this mean that the row distance is set at 80 cm row distance. This means that the crop density is less than maximum (which would be when the row distance would be set at the minimum of 65 cm), but ultimately, due to more optimal processing of the crop in the follow up processes, the yield is higher.

[0082] Figure 5 shows in five consecutive schematic figures A through F the actual crop rows (dotted lines) and the position and spatial extension of the planned cultivation paths (continuous lines) for each of the agricultural processes, but only for a small section of the actual piece of farmland. It starts with figure 5A showing the row creation process of planting the bean bush seedlings. In the figure 2 paths 75 are depicted, paths 1 and 2. The track width of the AgBot is set at its maximum of 3.20 m. When following each path,4 rows of bean bushes are created, 2 left of the path (rows 90 and 91 ) and 2 to the right Rows 92 and 93), at a row distance of 80 cm between the rows. Correspondingly, rows 94 and 95 as well as rows 96 and 97 are created when the vehicle (AgBot + Transplanter) crosses path 2. Figure 5B shows the path 75 for the 12-row sprayer, in this case showing only path 1 , which suffices to spray 6 rows left of the path (90 to 95) and 6 rows to the right of the path (of which only rows 96 and 97 are depicted in figure 5B). Correspondingly, figure 5C shows one path for the 8-row sprayer, spraying 4 paths to the left of path 75 and 4 rows to the right when crossing path 75. Figure 5D shows the paths for the 1-row weeder, which in each case has to travel twice in the space between two neighbouring rows. Figure 5E shows the two paths for the 4-row bean harvester and lastly figure 5F shows the paths for the 2-row bush harvester. This way, during non of the follow up processes any crop in the rows needs to be damaged due to the mere crossing of any of the vehicles over the paths.

[0083] In an alternative method, in each case, i.e. for each follow up process the same autonomous tractor is used (with a heightened bottom to be able and move over fully grown bushes), however in each case with a different implement. This way the complete cultivation process during one growing season can be run autonomously.

Claims

CLAIMS1. A method for generating a cultivation plan for an autonomous agricultural vehicle in order to process a piece of farmland, the vehicle comprising an implement for performing a row creation process to create multiple rows on the piece of farmland, which cultivation plan comprises multiple distinct paths that extend over the piece of farmland, the method comprising:- providing a processing unit to generate the cultivation plan,- providing input data to the processing unit, on the basis of which input data the processing unit calculates the position and spatial extension of each of the multiple distinct paths on the piece of farmland, the input data comprising farmland data and vehicle data,characterised in that in the method, in addition to the farmland data and the vehicle data, future data that pertain to one or more follow up processes that are due to be performed on the piece of farmland after the said row creating process has been performed, is used to calculate the position and spatial extension of each of the multiple distinct paths on the piece of farmland.

2. A method for generating a cultivation plan according to claim 1, characterised in that the future data comprise data of a vehicle that is used to perform the one or more follow up processes.

3. A method for generating a cultivation plan according to any of the preceding claims, wherein the multiple rows are plant rows and / or farmland asset rows, characterised in that the position and spatial extension of each path on the piece of farmland are calculated in order to minimise damage done to the multiple rows during the performance of the one or more follow up processes.

4. A method for generating a cultivation plan according to claim 3, characterised in that the position and spatial extension of each path on the piece of farmland are calculated to prevent damage done to the multiple rows during the performance of the one or more follow up processes.

5. A method for generating a cultivation plan according to any of the preceding claims,characterised in that one or more follow up processes are performed by one or more autonomous agricultural vehicles.

6. A method for generating a cultivation plan according to any of the preceding clams, characterised in that in each of the row creating process and the one or more follow up processes the same autonomous agricultural vehicle is used, in each of the said processes combined with an implement that is constituted to perform the corresponding process.

7. A method for generating a cultivation plan according to any of the preceding claims, characterised in that the future data pertain to one or more follow up processes that are due to be performed on the piece of farmland during one growing season of a crop.

8. A method for generating a cultivation plan according to any of the claims 1 to 6, characterised in that the future data pertain to one or more follow up processes that are due to be performed on the piece of farmland until the said multiple rows are removed from the piece of farmland.

9. A method for generating a cultivation plan according to any of the preceding claims, characterised in that the input data input for the processing unit are either actively put in by a human operator, and / or automatically retrieved from a digital source.

10. A method for generating a cultivation plan according to any of the preceding claims, wherein a user interface (Ul) is provided for communication between a human operator and the processing unit, characterised in that the Ul includes a display that visually shows the multiple distinct paths to the human operator.

11. A method for generating a cultivation plan according to any of the preceding claims, characterised in that the processing unit is operatively coupled to the autonomous agricultural vehicle.

12. A method for generating a cultivation plan according to claim 11, characterised in that the processing unit is permanently operatively coupled to the autonomous agricultural vehicle while processing the piece of farmland.

13. A method to cultivate a piece of farmland by performing multiple consecutiveagricultural processes on the piece of farmland, the first one of these multiple consecutive agricultural processes being a row creation process to create multiple rows in the piece of farmland, the method comprising:- providing an autonomous agricultural vehicle which comprises an implement that is able to perform the row creation process,- providing a processing unit to generate a cultivation plan for performing the row creation process on the piece of land, which cultivation plan comprises multiple distinct paths that spatially extend over the piece of land,- providing input data to the processing unit, on the basis of which input data the processing unit calculates the position and spatial extension of each of the multiple distinct paths on the piece of farmland, the input data comprising farmland data and vehicle data,- controlling the autonomous agricultural vehicle such that it autonomously crosses the piece of farmland land by moving over the said multiple distinct paths, while performing the row creation process,characterised in that in the method, in addition to the farmland data and the vehicle data, future data that pertain to one or more follow up processes of the multiple consecutive agricultural processes, which one or more follow up processes are due to be performed on the piece of farmland after the said row creating process has been performed, is used to calculate the position and spatial extension of each of the multiple distinct paths on the piece of farmland.

Images