Forest harvester localization system

Abstract

A forest harvester includes a cutting arm supporting a cutting head, a cabin serving as a reference point for the position of the cutting head, a positioning system comprising a perception sensor and a gnss receiver, and a computer configured to process data from the perception sensor and the gnss receiver to obtain a global cutting head position. The computer determines the global position data of the cabin according to a global coordinate system and the local cutting head position relative to the cabin. This enables improved accuracy and efficiency in forest harvesting operations.

Description

[0001] Forest Harvester Localization System

[0002] Field

[0003] The technology relates to the field of forestry and forest harvesting equipment, specifically focusing on the development and improvement of forest harvester machines used for tree felling, delimbing, and bucking operations. This field aims to enhance the efficiency, accu- racy, and safety of forest harvesting processes while minimizing the environmental impact and ensuring sustainable forest management practices.

[0004] Background

[0005] Forest harvesting operations are essential for the management and utilization of forest resources. These operations involve the felling, delimbing, and bucking of trees in des- ignated areas within a forest. The planning and execution of forest harvesting operations are based on maps that indicate the specific areas to be harvested, the forest management prescriptions to be applied, and the areas to be left untouched. The accurate and efficient execution of these operations is crucial for the sustainable management of forest resources and the economic viability of the forestry industry. In the prior art, forest areas designated for harvesting are marked using physical markers, such as ribbons or paint, which are manually placed by forestry workers. This method of marking forest areas is labor-intensive and time-consuming, leading to inefficiencies in the forest marking process. Furthermore, the reliance on physical markers can result in inaccuracies in the positioning of the harvesting boundaries, leading to potential damage to the surrounding environment and non-targeted trees.

[0006] Existing forest harvester positioning systems are not sufficiently precise and robust for accurate and efficient forest harvesting operations. The majority of these systems only provide the position of the cabin of the machine, not the harvester head, which can extend more than 10 meters away from the cabin. This lack of precision makes it difficult to replace physical markers with digital positioning systems, limiting the potential for streamlining and improving the forest marking process.

[0007] Moreover, current positioning systems rely on direct Global Navigation Satellite System (GNSS) positioning, which is prone to errors in forest environments due to interference from standing trees. This affects the accuracy of harvester positioning and makes it unsuitable for replacing physical markers in the forest. Additionally, existing precise positioning systems cannot be easily retrofitted to older machines due to the dependency on joint encoders in the harvester arm. This limits the adoption of more accurate positioning systems in existing forest harvesting equipment.

[0008] In summary, the prior art suffers from several shortcomings, including the labor-intensive and time-consuming process of marking forest areas using physical markers, the lack of precision in existing harvester positioning systems, the vulnerability of GNSS positioning to interference from standing trees, and the difficulty of retrofitting older machines with more accurate positioning systems. These shortcomings result in inefficiencies and inaccuracies in forest harvesting operations, negatively impacting the sustainability and economic viability of the forestry industry.

[0009] Summary According to a first aspect of the disclosure, there is provided a forest harvester comprising: a cutting arm supporting a cutting head, a reference point for the positioning ,a positioning system including a perception sensor and a gnss receiver, and a computer configured to process data from the perception sensor and the gnss receiver to obtain a continuous global position and orientation of the reference point , wherein the computer determines the global position data of the reference point according to a global coordinate system. This aspect provides the advantage of accurately determining the position of the cutting head, which can improve the efficiency and precision of forest harvesting operations.

[0010] Optionally the cutting arm is coupled to a machine frame and the reference point is a reference point of the machine frame.

[0011] Optionally the computer is configured to obtain a global cutting head position and wherein the computer determines the local cutting head position relative to the reference point in dependence of the data from the perception sensor.

[0012] Optionally in some examples, the perception sensor generates perception sensor data comprising a digital representation of at least parts of the cutting arm and surrounding environment. This feature provides the advantage of creating a detailed and accurate digital representation of the forest harvester and its surroundings, which can enhance the precision of the cutting head positioning.

[0013] Optionally in some examples, the digital representation includes a 3D point cloud, camera images, or radar images. This feature provides the advantage of offering a variety of data types for representing the forest harvester and its surroundings, which can improve the robustness and reliability of the cutting head positioning.

[0014] Optionally in some examples, the gnss receiver is configured to receive signals from satellites in the GNSS constellation and determine the global position data of the cabin. This feature provides the advantage of enabling the forest harvester to determine its global position, which can improve the accuracy of the cutting head positioning. Optionally in some examples, the computer is configured to determine a local 3d position and orientation of the cabin based on the perception sensor data. This feature provides the advantage of enabling the forest harvester to determine its local position and orientation, which can enhance the precision of the cutting head positioning.

[0015] Optionally in some examples, the computer is configured to determine a continuous global position and orientation of the cabin in dependence on a sequence of the local 3d position and orientation of the cabin and a sequence of the global position data. This feature provides the advantage of enabling the forest harvester to determine its global position and orientation with high precision, which can further improve the accuracy of the cutting head positioning. Optionally the determination of the continuous global position and orientation of the reference point comprising matching the sequence of global position data and the sequence of the local 3d position and orientation.

[0016] Optionally in some examples, the computer is configured to determine the local cutting head position in dependence on the perception sensor data of the cutting head or the cutting arm. This feature provides the advantage of enabling the forest harvester to determine the local position of the cutting head, which can enhance the precision of the cutting head positioning.

[0017] Optionally in some examples, the computer is configured to determine the global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the cabin. This feature provides the advantage of enabling the forest harvester to determine the global position of the cutting head, which can improve the accuracy of the cutting head positioning and enhance the efficiency and precision of forest harvesting operations. Optionally in some examples, the forest harvester further comprises a display configured to display real-time data on the global cutting head position in relation to a harvesting map. This feature provides the advantage of providing real-time feedback to the operator, which can improve the operator's situational awareness and enhance the efficiency and precision of forest harvesting operations.

[0018] Optionally in some examples, the harvesting map includes harvesting boundaries defining areas within the forest where harvesting operations are to take place and virtual boundary indicators indicating whether a tree is inside or outside the harvesting boundaries. This feature provides the advantage of guiding the operator in conducting harvesting opera- tions within specified boundaries, which can enhance the efficiency and precision of forest harvesting operations and ensure compliance with forest management prescriptions.

[0019] Optionally in some examples, the perception sensor is a lidar sensor, a radar sensor, a camera sensor, or a combination of multiple sensor modalities. This feature provides the advantage of offering a variety of sensor modalities for generating perception sensor data, which can improve the robustness and reliability of the cutting head positioning.

[0020] Optionally in some examples, the computer comprises a cutting head positioning algorithm selected from a group consisting of machine learning algorithms, clustering algorithms, object detection algorithms, rule-based algorithms, semantic segmentation algorithms, instance segmentation algorithms, and point cloud segmentation algorithms. This feature provides the advantage of offering a variety of algorithms for processing perception sensor data, which can enhance the precision of the cutting head positioning.

[0021] Optionally in some examples, the cutting head is moveable relative to the cabin and is configured for felling, delimbing, and bucking operations. This feature provides the advantage of enabling the cutting head to perform a variety of operations, which can enhance the versatility and efficiency of the forest harvester.

[0022] According to a second aspect of the disclosure, a method for global localization of a cutting head in a forest harvester is provided. The method comprises processing perception sensor data generated by a perception sensor to determine a local 3d position and orientation of a cabin of the forest harvester, determining a continuous global position and orientation of the cabin in dependence on the local 3d position and orientation of the cabin and global position data generated by a gnss receiver, determining a local cutting head position in dependence on the perception sensor data, and determining a global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the cabin. This aspect provides the advantage of accurately de- termining the position of the cutting head, which can improve the efficiency and precision of forest harvesting operations.

[0023] According to a third aspect of the disclosure, a positioning system for a forest harvester is provided. The positioning system comprises a perception sensor configured to gener- ate perception sensor data of the forest harvester and surrounding environment, a gnss receiver configured to generate global position data of a cabin of the forest harvester according to a global coordinate system, and a computer configured to process the perception sensor data and the global position data to determine a global cutting head position of a cutting head supported by a cutting arm of the forest harvester. The computer de- termines the global position data of the cabin and the local cutting head position relative to the cabin. This aspect provides the advantage of accurately determining the position of the cutting head, which can improve the efficiency and precision of forest harvesting operations.

[0024] Brief Description of the Drawings Examples are described in more detail below with reference to the appended drawings.

[0025] Figure 1 is a side view of a forest harvester with a cutting arm supporting a cutting head, a cabin, and the positioning system including a perception sensor and a gnss receiver.

[0026] Figure 2 is a block diagram illustrating the components of the computer configured to process data from the perception sensor and the gnss receiver to obtain a global cutting head position.

[0027] Figure 3 is a flowchart illustrating the method for global localization of the cutting head in the forest harvester.

[0028] Figure 4 is a perspective view of the forest harvester with a display configured to display real-time data on the global cutting head position in relation to a harvesting map. Detailed Description

[0029] The detailed description set forth below provides information and examples of the disclosed technology with sufficient detail to enable those skilled in the art to practice the disclosure. Figure 1 shows a side view of a forest harvester 100 with a cutting arm 80 supporting a cutting head 10, a cabin 20, and the positioning system 30 including a perception sensor 35 and a gnss receiver 40 and a computer 50. The cutting arm 80 is coupled to a machine frame and is moveable with respect to the machine frame. The cabin 20 is fixed with respect to the machine frame. The cutting arm 80 is connected to the cabin 20 and is able to articulate and extend to reach distant trees. The cutting head 10 is moveable relative to the cabin 20 and is configured for felling, delimbing, and bucking operations. A reference point 25 is using for positioning of the cutting head 10 and the machine frame. The cabin 20 houses the operator interface and controls for operating the forest harvester 100. The positioning system 30 is attached to the cabin 20 and is responsible for determining the global cutting head position.

[0030] Figure 2 is a block diagram illustrating the components of the positioning system 30, including a computer 50 configured to process data from the perception sensor 35 and the gnss receiver 40 to obtain a global cutting head position. The perception sensor 35 gen- erates perception sensor data, which includes a digital representation of at least parts of the cutting arm 80 and surrounding environment. The perception sensor data comprises data covering the surrounding environment and covering cutting head(s) 10 and / or cutting arm(s) 80. The gnss receiver 40 receives signals from satellites in the GNSS constellation and determines the global position data of the reference point, which may be intermittent global position data without orientation data. The computer 50 processes the perception sensor data and the global position data to determine the local cutting head position relative to the reference point 25 e.g. the cabin 20 and the global cutting head position according to a global coordinate system. In particular, the computer 50 is configured to run an algorithm to determine a sequence of positions and orientations of a reference point 25 (e.g. cabin 20) relative to the local environment. The computer 50 is also configured to run an algorithm to determine the harvester head position relative to reference point 25 (e.g. cabin 20). The computer 50 is configured to run an algorithm to associate the determined local reference point 25 position sequence and orientation to a global position data sequence. The computer 50 is also configured to run an algorithm to determine a global cutting head position based on the output from one or more of the other algorithms.

[0031] The display 60 can show the determined global cutting head position and the determined global position and orientation of the reference point 25.

[0032] Figure 3 is a flowchart illustrating the method for global localization of the cutting head 10 in the forest harvester 100. The method includes processing perception sensor data generated by the perception sensor 35 to determine a sequence of determinations of a local 3d position and orientation of a reference point 25. The method comprises determin- ing the harvester head position relative to cabin 20 based on the perception sensor data. The method comprises a step of associating a sequence of determined local 3d position and orientation to a sequence of received intermittent global positions. The method further comprises determining a continuous global position and orientation of the reference point 25 in dependence on the matched sequence of determined local 3d position and orientation and sequence of global position data generated by the gnss receiver 40. The method also comprises determining continuous global harvester head position based on the matched sequence of the determined local 3d position and orientation and sequence of global position data generated by the gnss receiver 40. Figure 4 is an example view of display 60 configured to visualize real-time data on the global cutting head position in relation to a harvesting map. The harvesting map includes harvesting boundaries defining areas within the forest where harvesting operations are to take place and virtual boundary indicators indicating whether a tree is inside or outside the harvesting boundaries. The display 60 helps improve the forest harvesting process for operators of different skill levels by providing real-time information on the global cutting head position and its relation to the harvesting boundaries.

[0033] Figure 5 highlights how a digital representation (e.g., 3d point cloud) of the harvester arm 80 and the surrounding environment is used to determine the position of the harvester head 10 relative to the reference point 25 e.g. cabin 20, and determine the position and orientation of the reference point 25 relative to the local environment. Further, the global position and orientation of the reference point 25 is determined by associating the local position and orientation to the global position from the gnss receiver 40. The global position of the harvester head 10 is determined from the global position and orientation of the reference point 25 and the position of the harvester head relative to the reference point 25.

[0034] 1 Forest Harvester Details

[0035] The forest harvester 100 may be a large, heavy-duty machine designed for the purpose of felling, delimbing, and bucking trees in a forest environment. The forest harvester 100 is typically equipped with a variety of components and systems that enable it to perform these tasks efficiently and safely. These components and systems may include, but are not limited to, a cutting crane 70 and a cabin 20. The forest harvester 100 may also include other components and systems not specifically mentioned here, such as an engine, a transmission, a hydraulic system, a fuel system, a cooling system, a braking system, a suspension system, and a control system, among others. 1.1 Cutting Crane

[0036] The cutting crane 70 is mounted on the forest harvester 100 and is designed to support and manipulate the cutting head 10. The cutting crane 70 may include a cutting arm 80 as well as the cutting head 10, and other components not specifically mentioned here.

[0037] 1.1.1 Cutting Arm

[0038] In some configurations, the cutting arm 80 is a component of the cutting crane 70. The cutting arm 80 is typically designed to support the cutting head 10 and connect it to the forest harvester 100. The cutting arm 80 may have a length of between 6 and 14 meters, although other lengths are also possible. The cutting arm 80 is typically designed to be flexible, able to articulate and extend to reach distant trees but strong enough to support the weight of the cutting head 10 and withstand the forces exerted during tree felling operations.

[0039] 1.1.2 Cutting Head

[0040] The cutting head 10 is typically designed to perform the actual cutting operations, such as felling, delimbing, and bucking. The cutting head 10 may include a variety of components, such as a saw, a delimbing device, a bucking device, and other components not specifically mentioned here. The cutting head 10 is moveable relative to the reference point 25.

[0041] Local Cutting Head Position The local cutting head position refers to the position of the cutting head 10 relative to the reference point 25 on the forest harvester 100. The local cutting head position may be determined by the computer 50 based on data from the perception sensor 35 and the gnss receiver 40. The local cutting head position may be represented in a variety of ways, such as a set of coordinates in a local coordinate system, a distance and angle from the reference point 25, or other representations. The local cutting head position may be used by the computer 50 to control the movement of the cutting head 10 and to determine the global cutting head position.

[0042] Global Cutting Head Position

[0043] The global cutting head position refers to the position of the cutting head 10 relative to a global coordinate system. The global cutting head position may be determined by the computer 50 based on the local cutting head position and the continuous global position and orientation of the reference point 25. The global cutting head position may be repre- sented in a variety of ways, such as a set of coordinates in a global coordinate system, a distance and angle from a global reference point 25, or other representations. The global cutting head position may be used by the computer 50 to control the movement of the cutting head 10 and to provide information to the operator of the forest harvester 100. 1.2 Reference Point

[0044] The reference point 25 is a position on the forest harvester 100 that can be used to provide a reference for the position and orientation of the forest harvester 100 as well as a reference for the position of the cutting head 10. The reference point 25 may be located at any suitable location on the forest harvester 100, such as the cabin 20, a specific point on the body of forest harvester 100, or other locations. The reference point 25 may be used by the computer 50 to determine the local and global positions of the cutting head 10.

[0045] 1.2.1 Local 3D Position And Local 3D Position and Orientation

[0046] The local 3d position and orientation refers to the position and orientation of the reference point 25 relative to the surrounding environment. A sequence of local 3d position and orientation may be determined by the computer 50 based on data from the perception sensor 35 In some examples, the sequence of local 3d position and orientation is a plurality of sequential determinations of the local 3d position and orientation. In some other less preferred examples, the computer 50 can carry out one local 3d position and orientation determination. The local 3d position and orientation may be represented in a variety of ways, such as a set of coordinates and angles in a local coordinate system, a distance and angles from a local reference point 25, or other representations.

[0047] The computer 50 is configured to match a sequence of local 3d position and orientation e.g. a plurality of positions and orientations to a sequence of global position data. As discussed below, the global position data is determined in dependence of the gnss receiver 40. As discussed below the sequence of local 3d position and orientation are determined in dependence of the perception sensor 35.

[0048] The sequence of the local 3d position and orientation and the sequence of the global position data may differ in several ways. The local 3d position and orientation are stable and continuous but not global. The global position data are global but may be intermittent and do not contain orientation data. However, over a window of time the computer 50 can match the two sequences (e.g. local 3d position and orientation and global position data) to obtain a stable, continuous (not intermittent), global position and orientation of the reference point. This in turn is the basis for determining the global cutting head position as discussed in more detail below.

[0049] 1.2.2 Continuous Global Position And Orientation

[0050] The continuous global position and orientation refers to the position and orientation of the reference point 25 relative to a global coordinate system. The continuous global position and orientation may be determined by the computer 50 based on the local 3d position and orientation sequence of the reference point 25 and the intermittent global position data from the gnss receiver 40. The continuous global position and orientation may be represented in a variety of ways, such as a set of coordinates and angles in a global coordinate system, a distance and angles from a global reference point 25, or other representations. The continuous global position and orientation may be used by the computer 50 to determine the global position of the cutting head 10 or passed on to the operator interface. The continuous global position and orientation is resilient to intermittent GNSS signal, which is especially relevant where the forest harvester 100 is travelling through forest and the GNSS signal is weak. This enables continuous and accurate positioning of the harvester head even in areas with weak or intermittent GNSS signals, ensuring reliable operation in all forest environments.

[0051] Continuous global precision orientation data represents a high-frequency stream of three- dimensional position information, referenced to a global coordinate system (e.g., WGS84), and characterized by a consistent periodicity in both time and / or distance domains. This data stream is generated through the fusion of multiple sensor modalities, providing realtime positional updates at intervals between 1ms and 50ms, and in some other examples up to 1 second, ensuring precise tracking of the harvester head's movement even during rapid operation. In other examples the interval can be any suitable interval e.g. 20ms, 30ms, 40ms, 50ms, 60ms, 70ms, 80ms, 100ms, 200ms, 500ms, Is, 2s, 3s, 4s etc. Furthermore, the system maintains continuous positional updates with a spatial resolution between 1cm and 10cm, guaranteeing precise location information regardless of the harvester's speed or the complexity of the terrain. This combination of high temporal and spatial resolution, coupled with the global coordinate referencing, enables seamless integration with digital harvesting plans and provides robust performance even in envi- ronments with intermittent or absent GNSS signals. This continuous data stream aids precise control, real-time monitoring, and accurate mapping of the harvester head's position throughout the harvesting operation. 1.2.3 Cabin

[0052] In one configuration, a cabin 20 serves as the reference point 25. The cabin 20 is typically the central operative area for the forest harvester 100, housing the operator interface and controls for operating the forest harvester 100. The cabin 20 may be designed to provide a comfortable and safe environment for the operator of the forest harvester 100, with features such as a seat, a control panel, a display 60, windows, doors, air conditioning, heating, lighting, and other features.

[0053] The cabin 20 may also include mounting points for attaching the positioning system 30.

[0054] The mounting points may be configured to allow for easy installation and removal of the positioning system 30 for maintenance and upgrades. The mounting points may be located at any suitable location on the cabin 20, such as on the roof, the sides, the front, the rear, or other locations. The mounting points may be designed to securely hold the positioning system 30 in place, while also allowing for easy access for installation, removal, and maintenance. The mounting points may include features such as brackets, bolts, clamps, hooks, or other features.

[0055] 1.3 Positioning System

[0056] The positioning system 30 is configured to determine the global position and orientation of the reference point and the global position of the cutting head 10 . The positioning system 30 may comprise a perception sensor 35, a gnss receiver 40, a computer 50, and other components not specifically mentioned here.

[0057] 1.3.1 Perception Sensor

[0058] The perception sensor 35 is configured to generate perception sensor data of the forest harvester 100 and the surrounding environment. The perception sensor 35 may include a variety of sensor types, such as a lidar sensor, a radar sensor, a camera sensor, or a combination of multiple sensor modalities.

[0059] Perception Sensor Data

[0060] The perception sensor data typically comprises a digital representation or model of at least parts of the cutting crane 70 and the surrounding environment. The perception sensor data may include a variety of data types, such as 3D point clouds, 2D camera images, radar images, or other data types. The perception sensor data may be used by the computer 50 to determine the sequence of local 3d position and orientation of the reference point 25, the local cutting head position, and the continuous global position and orientation of the reference point 25.

[0061] The surrounding environment typically includes trees, rocks, or other terrain features that are in the vicinity of the forest harvester 100. The surrounding environment may also be used by the computer 50 to issue a control signal e.g. such as issuing a notification to the user such as warning the user that the felling of a tree in a certain area is not allowed etc.

[0062] Optionally, the control signal may prevent the control of the movement of the cutting head 10 and the forest harvester lOOwithin a certain area.

[0063] 1.3.2 Gnss Receiver

[0064] In some implementations, the gnss receiver 40 is a component of the positioning system 30. The gnss receiver 40 is configured to receive signals from satellites in the GNSS constellation and determine the global position data of the reference point 25 according to a global coordinate system. The gnss receiver 40 may be designed to be robust and durable, capable of withstanding the harsh conditions of a forest environment. The gnss receiver 40 may also be designed to be accurate and reliable, capable of providing precise global position data for the computer 50 to process. However, due to the forest machine operating under the tree canopy in varying terrain, there will always be cases when gnss receiver 40 loses connection with the satellites, resulting in intermittent global position data.

[0065] Global Position Data In some examples, the global position data is a feature of the gnss receiver 40. The global position data comprises the coordinates of the reference point 25 according to a global coordinate system. The global position data may be determined by the gnss receiver 40 based on signals received from satellites in the GNSS constellation. The global position data may be used by the computer 50 to determine the continuous global position and orientation of the reference point 25 and the global cutting head position. The global position data may be vulnerable to imprecision or temporary outages in the GNSS signal due to tree canopies interfering with satellite communications.

[0066] 1.3.3 Computer

[0067] In one configuration, the computer 50 processes data from the perception sensor 35 and the gnss receiver 40 to obtain the global cutting head position. The computer 50 may include a variety of components, such as a processor, a memory, a storage device, a network interface, a user interface, and other components not specifically mentioned here. Cutting Head Positioning Algorithm

[0068] In some examples, the computer 50 comprises a cutting head positioning algorithm. The cutting head positioning algorithm is configured to process data from the perception sensor 35 to determine the local position of the cutting head 10 relative to the reference point 25. The cutting head positioning algorithm may also process data from the perception sensor 35 to determine the local position of the cutting arm 80 relative to the reference point 25 and determine the position of the cutting head 10 from the position of the cutting arm 80. The cutting head positioning algorithm may be of various types, such as a machine learning algorithm, a clustering algorithm, an object detection algorithm, a rule-based algorithm, a semantic segmentation algorithm, an instance segmentation algorithm, a point cloud segmentation algorithm, or other types of algorithms. The cutting head positioning algorithm may be designed to be accurate and reliable, capable of providing precise position data for the cutting head 10.

[0069] In some examples, the cutting head positioning algorithm can optionally use information including orientation of the cutting head 10 to further improve user functionality. This information can include details such as recording where the cut logs end up based on the orientation of the cutting head 10 during bucking.

[0070] 1.4 Display

[0071] The forest harvester 100 may comprise a display 60. The display 60 is configured to display 60 real-time data on the global cutting head position in relation to a harvesting map.

[0072] The display 60 may include a variety of components, such as a screen, a user interface, a control panel, and other components not specifically mentioned here. The display 60 may also include a harvesting map, which may include harvesting boundaries defining areas within the forest where harvesting operations are to take place and virtual boundary indicators indicating whether a tree is inside or outside the harvesting boundaries.

[0073] 1.4.1 Harvesting Map

[0074] In some configurations, display 60 may show a harvesting map. The harvesting map may comprise harvesting boundaries defining areas within the forest where harvesting operations are to take place. The harvesting map may also include virtual boundary indicators indicating whether a tree is inside or outside the harvesting boundaries. The harvesting map may be displayed on the display 60 and used by the operator of the forest harvester 100 to plan and control the movement of the cutting head 10 and the forest harvester 100.

[0075] In some examples, the harvesting boundaries are a feature of the harvesting map. The harvesting boundaries typically define the areas within the forest where harvesting operations are to take place. The harvesting boundaries may be represented in a variety of ways, such as lines, polygons, or other shapes. The harvesting boundaries may be determined based on a variety of factors, such as the type of trees to be harvested, the size of the trees, the density of the trees, the terrain, the accessibility, the environmental considerations, and other factors. The harvesting boundaries may be used by the operator of the forest harvester 100 to plan and control the movement of the cutting head 10 and the forest harvester 100.

[0076] In one configuration, virtual boundary indicators are shown on the harvesting map. The virtual boundary indicators may indicate whether a tree is inside or outside the harvesting boundaries. The virtual boundary indicators may be represented in a variety of ways, such as colors, symbols, or other indicators. The virtual boundary indicators may be determined based on the position of the tree relative to the harvesting boundaries. The virtual boundary indicators may be used by the operator of the forest harvester 100 to plan and control the movement of the cutting head 10 and the forest harvester 100.

[0077] 2 Global Localization Of The Cutting Head Method Details

[0078] In one implementation, the method for global localization of the cutting head 10 in the forest harvester 100 includes several steps.

[0079] 2.1 Perception Sensor Data Generation The perception sensor data generation comprises the perception sensor 35 generating perception sensor data of the forest harvester 100 and the surrounding environment. The perception sensor data may include a digital representation of at least parts of the cutting crane 70 and the surrounding environment. The digital representation may comprise a digital model of at least parts of the cutting crane 70 and the surrounding environment. The digital representation may be in various forms, such as a 3D point cloud, 2D camera images, radar images, or other forms. The digital representation may include the cutting head 10 and / or the cutting arm 80, as well as other components of the forest harvester 100 and the surrounding environment.

[0080] 2.2 Global Position Data Generation Gnss receiver 40 generates global position data of the reference point 25 of the forest harvester 100 according to a global coordinate system. 2.3 Local 3D Position and Orientation Determination

[0081] This step comprises computer 50 processing the perception sensor data to determine a local 3d position and orientation of the reference point 25 of the forest harvester 100. The local 3d position and orientation may be represented in a variety of ways, such as a set of coordinates and angles in a local coordinate system, a distance and angle from a local reference point 25, or other representations. The perception sensor data may be used to identify the position and orientation of the reference point 25 relative to the surrounding environment, which may include trees, rocks, or other terrain features. Examples of algorithms that can be employed in this step include simultaneous localization and map- ping (SLAM) algorithms, lidar odometry algorithms, visual odometry algorithms or radar odometry algorithms. As mentioned below the determination of the local 3d position and orientation may be a sequence of separate determinations of the local 3d position and orientation. In this case, each determination of the local 3d position and orientation in the sequence is performed according to the same process e.g. according to the examples discussed herein with respect to the determination of the local 3d position and orientation.

[0082] 2.4 Continuous Global Position and Orientation Determination

[0083] This step comprises computer 50 determining a continuous global position and orientation of the reference point 25 in dependence on the local 3d position and orientation sequence of the reference point 25 and the intermittent global position data sequence generated by the gnss receiver 40. The continuous global position and orientation may be represented in a variety of ways, such as a set of coordinates and angles in a global coordinate system , a distance and angles from a global reference point 25, or other representations.

[0084] In one example, the computer 50 performs outlier robust matching of the time stamped positions of the local 3d position and orientation sequence of the reference point 25, as determined from the perception sensor data, with the time stamped positions of the intermittent global position data sequence, as determined by the gnss receiver 40, to associate global positions to the local 3d position and local 3d position and orientation sequence, and thus determine a continuous global position and orientation sequence, from which the current continuous global position and orientation of the reference point 25 can be ob- tained. This matching process may involve finding the mathematical transformations, such as translations, rotations, scalings, or other transformations, to associate the sequence of the local 3d position and orientation with global positions, and thereby obtaining a global position and orientation. Examples of algorithms that can be employed in this step include point set registration algorithms. An alternative approach would be to include the global position data in a SLAM algorithm and solve for global position and orientation directly. In one example, a series of global position data is matched with a corresponding series of local 3d position and orientation data, to provide a series of global position and orientation data. In one example, the series of global position and orientation data is a time series col- lected over a window of between 10 seconds and 60 seconds. In another example, each data point of the series of global position and orientation data, is collected in response to a local 3D position or orientation change of the forest harvester 100. In this example, the precision and robustness of the global position and orientation data of the forest harvester 100 improves over time or the more the forest harvester 100 travels. 2.5 Local Cutting Head Position Determination

[0085] This step comprises computer 50 determining the local cutting head position in dependence on the perception sensor data. The local cutting head position refers to the position of the cutting head 10 relative to the reference point 25 on the forest harvester 100. The local cutting head position may be represented in a variety of ways, such as a set of co- ordinates in a local coordinate system, a distance and angle from the reference point 25, or other representations. In one implementation, computer 50 processes the perception sensor data, which includes a digital representation of at least parts of the cutting crane 70 and the surrounding environment, to determine the local position of the cutting head 10 relative to the reference point 25. The perception sensor data may be used to identify the position of the cutting head 10 relative to the reference point 25, which may be determined based on the position and orientation of the cutting arm 80 and the cutting head 10. Examples of algorithms that can be employed in this step include object detection algorithms.

[0086] 2.6 Global Cutting Head Position Determination This step comprises computer 50 determining the global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the reference point 25. The global cutting head position refers to the position of the cutting head 10 relative to a global coordinate system. The global cutting head position may be represented in a variety of ways, such as a set of coordinates in a global coordinate system, a distance and angle from a global reference point 25, or other representations.

[0087] In one configuration, the computer 50 integrates the local cutting head position, as determined from the perception sensor data, with the continuous global position and orientation of the reference point 25, as determined from the sequences of local 3d position and orientation and the global position data respectively, to determine the global cutting head po- sition. This integration process may involve various mathematical transformations, such as translations, rotations, scalings, or other transformations, to convert the local cutting head position into a global cutting head position.

[0088] 3 Operational Process In one implementation, the operational process of the forest harvester 100 involves several stages, including an initial setup and calibration stage, a real-time operation stage, and a system adjustments and error handling stage. These stages may involve various tasks and operations performed by the operator of the forest harvester 100, the computer 50, and other components of the forest harvester 100. 3.1 Initial Setup and Calibration

[0089] In some configurations, the initial setup and calibration involves installing and calibrating the positioning system 30 on the forest harvester 100. The initial setup and calibration may involve various tasks, such as mounting the positioning system 30 on the cabin 20, connecting the positioning system 30 to the power supply and the control system of the forest harvester 100, configuring the settings of the positioning system 30, calibrating the perception sensor 35 and the gnss receiver 40, and other tasks. The initial setup and calibration may be performed by the operator of the forest harvester 100, a technician, or other personnel.

[0090] 3.1.1 Positioning System Installation and Calibration In one example, the positioning system 30 installation and calibration involves installing the positioning system 30 on the cabin 20 of the forest harvester 100 and calibrating the perception sensor 35 and the gnss receiver 40. The positioning system 30 installation may involve attaching the positioning system 30 to the mounting points on the cabin 20, connecting the positioning system 30 to the power supply and the control system of the forest harvester 100, and other tasks. The positioning system 30 calibration may involve adjusting the settings of the perception sensor 35 and the gnss receiver 40, testing the performance of the perception sensor 35 and the gnss receiver 40, and other tasks. The positioning system 30 installation and calibration may be performed by the operator of the forest harvester 100, a technician, or other personnel. 3.2 Real-time Operation

[0091] In some implementations, the real-time operation involves the forest harvester 100 performing tree felling operations while the positioning system 30 determines the global cutting head position in real time. The real-time operation may involve various tasks, such as moving the forest harvester 100 to the desired location, operating the cutting head 10 to fell, delimbing, and bucking trees, monitoring the global cutting head position on the display 60, and other tasks. The real-time operation may be performed by the operator of the forest harvester 100, the computer 50, and other components of the forest harvester 100. 3.2.1 Data Collection and Processing

[0092] In one configuration, the data collection and processing involves the perception sensor 35 generating perception sensor data of the forest harvester 100 and the surrounding environment, and the gnss receiver 40 generating global position data of the reference point 25. The computer 50 then processes the perception sensor data and the global position data to determine the sequence of the local 3d position and orientation of the reference point 25, the continuous global position and orientation of the reference point 25, the local cutting head position, and the global cutting head position. The data collection and processing may be performed continuously or at regular intervals during the real-time operation of the forest harvester 100. 3.2.2 Position Determination and Display

[0093] In some examples, the position determination and display 60 involves the computer 50 determining the global cutting head position based on the perception sensor data and the global position data, and displaying the global cutting head position on the display 60 in relation to a harvesting map. The position determination and display 60 may provide real- time information to the operator of the forest harvester 100, helping to improve the forest harvesting process for operators of different skill levels.

[0094] 3.3 System Adjustments and Error Handling

[0095] In some configurations, the system adjustments and error handling involves the computer 50 adjusting the settings of the positioning system 30 and handling errors that may occur during the operation of the forest harvester 100. The system adjustments may involve tasks such as recalibrating the perception sensor 35 and the gnss receiver 40, adjusting the settings of the cutting head positioning algorithm, updating the software of the position- ing system 30, and other tasks. The error handling may involve tasks such as diagnosing and fixing problems with the positioning system 30, dealing with an intermittent GNSS signal, correcting errors in the perception sensor data or the global position data, and other tasks. The system adjustments and error handling may be performed by the operator of the forest harvester 100, a technician, or other personnel.

[0096] 3.3.1 Dealing with GNSS Signal Interruptions

[0097] In one example, dealing with GNSS signal interruptions involves the computer 50 handling interruptions in the GNSS signal that may occur during the operation of the forest harvester 100. The GNSS signal may be interrupted due to various factors, such as tree canopies interfering with satellite communications, atmospheric conditions, or other factors. When an interruption in the GNSS signal occurs, the computer 50 may use the last known global position data and the perception sensor data to estimate the current global position and orientation of the reference point 25. The computer 50 may also use various error correction techniques to improve the accuracy of the estimated global position and orientation. This task helps to ensure continuous and accurate positioning of the cutting head 10 even in areas with weak or intermittent GNSS signals, ensuring reliable operation in all forest environments.

[0098] 3.3.2 System Calibration and Error Correction

[0099] In some configurations, the system calibration and error correction involves the computer 50 calibrating the perception sensor 35 and the gnss receiver 40, and correcting errors in the perception sensor data or the global position data. The system calibration may involve adjusting the settings of the perception sensor 35 and the gnss receiver 40, testing the performance of the perception sensor 35 and the gnss receiver 40, and other tasks. The error correction may involve identifying and correcting errors in the perception sensor data or the global position data, such as noise, outliers, or other errors. The system calibration and error correction may be performed by the operator of the forest harvester 100, a technician, or other personnel.

[0100] 4 Description of Examples of the Disclosure

[0101] The following sections provide examples of the disclosure, illustrating various configurations and implementations of the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100. 4.1 Example of Forest Harvester with Specific Sensor Modalities

[0102] In one example, the forest harvester 100 includes a positioning system 30 with specific sensor modalities. The positioning system 30 includes a perception sensor 35 that may be a lidar sensor, a radar sensor, a camera sensor, or a combination of multiple sensor modalities.

[0103] 4.1.1 Use of Lidar Sensor

[0104] In some configurations, the perception sensor 35 is a lidar sensor. The lidar sensor is typically designed to emit laser beams and measure the time it takes for the beams to return after hitting an object. The lidar sensor can generate a 3D point cloud of the forest harvester 100 and the surrounding environment, providing detailed and accurate perception sensor data for the computer 50 to process. The lidar sensor may be particularly useful in forest environments, where it can penetrate tree canopies and provide detailed information about the trees and the terrain.

[0105] 4.1.2 Use of Camera Sensor In some examples, the perception sensor 35 is a camera sensor. The camera sensor is typically designed to capture 2D images of the forest harvester 100 and the surrounding environment. The camera sensor can provide visual perception sensor data for the computer 50 to process, which may be particularly useful for tasks such as object detection, image segmentation, and other tasks. The camera sensor may be a color camera, a monochrome camera, an infrared camera, a thermal camera, or other types of cameras.

[0106] 4.2 Example of Forest Harvester with Specific Algorithm Types

[0107] In one implementation, the computer 50 of the forest harvester 100 includes a cutting head positioning algorithm of specific types. The cutting head positioning algorithm may be a machine learning algorithm, a clustering algorithm, an object detection algorithm, a rule-based algorithm, a semantic segmentation algorithm, an instance segmentation algorithm, a point cloud segmentation algorithm, or other types of algorithms. The cutting head positioning algorithm is typically designed to process the perception sensor data and the global position data to determine the local and global positions of the cutting head 10. The cutting head positioning algorithm may be designed to be accurate and reliable, capable of providing precise position data for the cutting head 10. 4.2.1 Use of Machine Learning Algorithm

[0108] In some configurations, the cutting head positioning algorithm is a machine learning algorithm. The machine learning algorithm is typically designed to learn from the perception sensor data and the global position data, and improve its performance over time. The machine learning algorithm may use various techniques, such as supervised learning, unsupervised learning, reinforcement learning, deep learning, or other techniques. The machine learning algorithm may be particularly useful for tasks such as object detection, image segmentation, and other tasks that require the ability to learn and adapt to changing conditions. 4.2.2 Use of Clustering Algorithm

[0109] In some examples, the cutting head positioning algorithm is a clustering algorithm. The clustering algorithm is typically designed to group similar data points together based on their features. The clustering algorithm may use various techniques, such as k-means clustering, hierarchical clustering, density-based clustering, or other techniques. The clus- tering algorithm may be particularly useful for tasks such as object detection, where it can group together data points that belong to the same object.

[0110] 5 Potential Applications

[0111] The following sections provide potential applications of the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100. 5.1 Application in Different Forest Environments

[0112] In one implementation, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in different forest environments. These forest environments may include dense forests, open forests, mixed forests, or other types of forests. 5.1.1 Dense Forests

[0113] In some configurations, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in dense forests. Dense forests typically have a high density of trees, which can make it difficult for the operator of the forest harvester 100 to see the trees and the terrain clearly. The positioning system 30 and the method for global localization of the cutting head 10 can provide accurate and reliable position data for the cutting head 10, helping the operator to navigate the dense forest and perform tree felling operations efficiently and safely.

[0114] 5.1.2 Open Forests

[0115] In some examples, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in open forests. Open forests typically have a low density of trees, which can make it easier for the operator of the forest harvester 100 to see the trees and the terrain. However, the trees in open forests may be spread out over a large area, which can make it difficult for the operator to plan and control the movement of the cutting head 10. The positioning system 30 and the method for global localization of the cutting head 10 can provide accurate and reliable position data for the cutting head 10, helping the operator to navigate the open forest and perform tree felling operations efficiently and safely.

[0116] 5.2 Application in Different Harvesting Operations

[0117] In one implementation, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in different harvesting operations. These harvesting operations may include felling operations, delimbing operations, bucking operations, or other operations.

[0118] In some configurations, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in felling operations. Felling operations typically involve cutting down trees at their base. The positioning system 30 and the method for global localization of the cutting head 10 can provide accurate and reliable position data for the cutting head 10, helping the operator to position the cutting head 10 accurately for felling operations.

[0119] In some examples, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in delimbing operations. Delimbing operations typically involve removing the branches from a felled tree. The positioning system 30 and the method for global localization of the cutting head 10 can provide accurate and reliable position data for the cutting head 10, helping the operator to position the cutting head 10 accurately for delimbing operations. In one configuration, the forest harvester 100 and the method for global localization of the cutting head 10 in the forest harvester 100 can be applied in bucking operations. Bucking operations typically involve cutting a felled and delimbed tree into logs. The positioning system 30 and the method for global localization of the cutting head 10 can provide accu- rate and reliable position data for the cutting head 10, helping the operator to position the cutting head 10 accurately for bucking operations.

[0120] Example 1: A forest harvester 100 comprising a cutting arm 80 supporting a cutting head 10, a reference point 25 for the position of the cutting head 10, a positioning system 30 including a perception sensor 35 and a gnss receiver 40, and a computer 50 configured to process data from the perception sensor 35 and the gnss receiver 40 to obtain a global cutting head position, wherein the computer 50 determines the global position data of the reference point 25 according to a global coordinate system and the local cutting head position relative to the reference point 25. Example 2: The forest harvester 100 of example 1, wherein the perception sensor 35 generates perception sensor data comprising a digital representation of at least parts of the cutting arm 80 and surrounding environment.

[0121] Example 3: The forest harvester 100 of example 2, wherein the digital representation includes a 3D point cloud, camera images, or radar images. Example 4: The forest harvester 100 of any of the examples of 1 to 3, wherein the gnss receiver 40 is configured to receive signals from satellites in the GNSS constellation and determine the global position data of the reference point 25.

[0122] Example 5: The forest harvester 100 of any of the examples 1 to 4, wherein the computer 50 is configured to determine a local 3d position and orientation of the reference point 25 based on the perception sensor data.

[0123] Example 6: The forest harvester 100 of any of the examples 1 to 5, wherein the computer 50 is configured to determine a continuous global position and orientation of the reference point 25 in dependence on the local 3d position and orientation of the reference point 25 and the global position data. Example 7: The forest harvester 100 of any of the examples 1 to 6, wherein the computer 50 is configured to determine the local cutting head position in dependence on the perception sensor data of the cutting head 10 or the cutting arm 80.

[0124] Example 8: The forest harvester 100 of any of the examples 1 to 7, wherein the computer 50 is configured to determine the global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the reference point 25. Example 9: The forest harvester 100 of any of the examples 1 to 8, further comprising a display 60 configured to display 60 real-time data on the global cutting head position in relation to a harvesting map.

[0125] Example 10: The forest harvester 100 of example 9, wherein the harvesting map includes harvesting boundaries defining areas within the forest where harvesting operations are to take place and virtual boundary indicators indicating whether a tree is inside or outside the harvesting boundaries.

[0126] Example 11: The forest harvester 100 of any of the examples 1 to 10, wherein the perception sensor 35 is a lidar sensor, a radar sensor, a camera sensor, or a combination of multiple sensor modalities.

[0127] Example 12: The forest harvester 100 of any of the examples 1 to 11, wherein the computer 50 comprises a cutting head positioning algorithm selected from a group consisting of machine learning algorithms, clustering algorithms, object detection algorithms, rulebased algorithms, semantic segmentation algorithms, instance segmentation algorithms, and point cloud segmentation algorithms.

[0128] Example 13: The forest harvester 100 of any of the examples 1 to 12, wherein the cutting head 10 is moveable relative to the reference point 25 and is configured for felling, delimbing, and bucking operations.

[0129] Example 14: The forest harvester 100 of any of the examples 1 to 12, wherein the refer- ence point 25 is a cabin 20 of the forest harvester 100.

[0130] Example 15: A method for global localization of a cutting head 10 in a forest harvester 100 according to any of the examples 1 to 14, the method comprising processing perception sensor data generated by a perception sensor 35 to determine a local 3d position and orientation of a reference point 25 of the forest harvester 100, determining a continuous global position and orientation of the reference point 25 in dependence on the local 3d position and orientation of the reference point 25 and global position data generated by a gnss receiver 40, determining a local cutting head position in dependence on the perception sensor data, and determining a global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the reference point 25.

[0131] Example 16: A positioning system 30 for a forest harvester 100, the positioning system 30 comprising a perception sensor 35 configured to generate perception sensor data of the forest harvester 100 and surrounding environment, a gnss receiver 40 configured to generate global position data of a reference point 25 of the forest harvester 100 according to a global coordinate system, and a computer 50 configured to process the perception sensor data and the global position data to determine a global cutting head position of a cutting head 10 supported by a cutting arm 80 of the forest harvester 100, wherein the computer 50 determines the global position data of the reference point 25 and the local cutting head position relative to the reference point 25.

[0132] The terminology used herein is for the purpose of describing particular aspects only and is not intended to be limiting of the disclosure. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items. It will be further understood that the terms "comprises," "comprising," "includes," and / or "including" when used herein specify the presence of stated features, integers, actions, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, actions, steps, operations, elements, components, and / or groups thereof.

[0133] It will be understood that, although the terms first, second, etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element without departing from the scope of the present disclosure.

[0134] Relative terms such as "below" or "above" or "upper" or "lower" or "horizontal" or "vertical" may be used herein to describe a relationship of one element to another element as illustrated in the Figures. It will be understood that these terms and those discussed above are intended to encompass different orientations of the device in addition to the orientation depicted in the Figures. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or intervening elements may be present. In contrast, when an element is referred to as being "directly connected" or "directly coupled" to another element, there are no intervening elements present. Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. It will be further understood that terms used herein should be interpreted as having a meaning consistent with their meaning in the context of this specification and the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0135] It is to be understood that the present disclosure is not limited to the aspects described above and illustrated in the drawings; rather, the skilled person will recognize that many changes and modifications may be made within the scope of the present disclosure and appended claims. In the drawings and specification, there have been disclosed aspects for purposes of illustration only and not for purposes of limitation, the scope of the disclosure being set forth in the following claims.

Claims

Claims1. A forest harvester comprising: a cutting arm supporting a cutting head, a reference point for the positioning , a positioning system including a perception sensor and a gnss receiver, and a computer configured to process data from the perception sensor and the gnss receiver to obtain a continuous global position and orientation of the reference point wherein the computer determines the global position data of the reference point according to a global coordinate system .

2. The forest harvester according to claim 1, wherein the cutting arm is coupled to a machine frame and the reference point is a reference point of the machine frame.

3. The forest harvester according to claims Ito 2 wherein the computer is configured to obtain a global cutting head position and wherein the computer determines the local cutting head position relative to the reference point in dependence of the data from the perception sensor.

4. The forest harvester according to any of claims 1 to 3, wherein the perception sensor generates perception sensor data comprising a digital representation of at least parts of the cutting arm and surrounding environment.

5. The forest harvester according to claim 4, wherein the digital representation includes a 3D point cloud, camera images, or radar images.

6. The forest harvester according to any of claims 1 to 5, wherein the gnss receiver is configured to receive signals from satellites in the GNSS constellation and determine the global position data of the reference point.

7. The forest harvester according to any of claims 1 to 6, wherein the computer is configured to determine a local 3d position and orientation of the reference point based on the perception sensor data.

8. The forest harvester according to any of claims 1 to 7, wherein the computer is configured to determine a continuous global position and orientation of the reference point in dependence on the sequence of local 3d position and orientation of the reference point and the sequence of global position data.

9. The forest harvester according to claim 8 wherein the determination of the contin-uous global position and orientation of the reference point comprises matching the sequence of global position data and the sequence of the local 3d position and orientation.

10. The forest harvester according to any of claims 1 to 9, wherein the computer is configured to determine the local cutting head position in dependence on the perception sensor data of the cutting head or the cutting arm.

11. The forest harvester according to any of claims 1 to 10, wherein the computer is configured to determine the global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the reference point.

12. The forest harvester according to any of claims 1 to 11, further comprising a display configured to display real-time data on the global cutting head position in relation to a harvesting map.

13. The forest harvester according to claim 12, wherein the harvesting map includes harvesting boundaries defining areas within the forest where harvesting operations are to take place and virtual boundary indicators indicating whether a tree is inside or outside the harvesting boundaries.

14. The forest harvester according to any of claims 1 to 13, wherein the perception sensor is a lidar sensor, a radar sensor, a camera sensor, or a combination of multiple sensor modalities.

15. The forest harvester according to any of claims 1 to 14, wherein the computer comprises a cutting head positioning algorithm selected from a group consisting of machine learning algorithms, clustering algorithms, object detection algorithms, rulebased algorithms, semantic segmentation algorithms, instance segmentation algorithms, and point cloud segmentation algorithms.

16. The forest harvester according to any of claims 1 to 15, wherein the cutting head is moveable relative to the reference point and is configured for felling, delimbing, and bucking operations.

17. The forest harvester according to any of claims 1 to 16, wherein the reference point is a cabin of the forest harvester.

18. A method for global localization of a cutting head in a forest harvester according to any of claims 1 to 17, the method comprising:processing perception sensor data generated by a perception sensor to determine a local 3d position and orientation of a reference point of the forest harvester, determining a continuous global position and orientation of the reference point in dependence on the local 3d position and orientation of the reference point and global position data generated by a gnss receiver, determining a local cutting head position in dependence on the perception sensor data, and determining a global cutting head position in dependence on the local cutting head position and the continuous global position and orientation of the reference point.

19. A positioning system for a forest harvester, the positioning system comprising: a perception sensor configured to generate perception sensor data of the forest harvester and surrounding environment, a gnss receiver configured to generate global position data of a reference point of the forest harvester according to a global coordinate system, and a computer configured to process the perception sensor data and the global position data to determine a continuous global position and orientation of the reference point, wherein the computer determines the global position data of the reference point .

Images