A cutting apparatus and a method for carrying out cutting tasks on food items

Abstract

A cutting apparatus (1) for executing cutting tasks on food items (4), comprising - a conveying device (2) having a support surface (3) configured to convey at least one row of food items (4) in a downstream direction between an inlet (5) and an outlet (6), - at least one multi-axis cutting device (8) having a water jet cutting nozzle (9) being movable in at least two dimensions over the support surface (3). To improve quality and processing capacity, and reduce processing costs, and / or reduce footprint, the cutting apparatus further comprises at least one single-axis cutting device (10) having a water jet cutting nozzle (11) being movable in at most one dimension over the support surface (3).

Description

[0001] A CUTTING APPARATUS AND A METHOD FOR CARRYING OUT CUTTING TASKS ON FOOD

[0002] ITEMS

[0003] FIELD OF THE INVENTION

[0004] The present invention relates to a cutting apparatus for executing cutting tasks on food items.

[0005] BACKGROUND OF THE INVENTION

[0006] In food processing, including beef, poultry, fish, pork, and other meat products, the food items may be cut in various ways, e.g., depending on shapes, dimensions, and weights before the items are processed or after the items are processed, or depending on a desired cut such as to obtain a filet of poultry breast or a filet of fish, or a steak cut from a roast etc., or depending on a desired size or weight after the food items are processed.

[0007] High-speed cutting and portioning machines exist in which incoming food items are analyzed to determine optimal portioning and / or trimming. Industrial machines can include several cutting devices, e.g., cutting devices based on water jet cutting, or traditional knife-based cutting.

[0008] Depending on the desired cut, the existing cutting devices may operate sub optimally and sometimes utilize a mechanical setup which is more advanced than needed for trivial cutting tasks, or alternatively, lack the sufficient flexibility required for obtaining a more advanced cutting. In the first case, the costs for the processing of the food item may be higher than necessary, in the latter case, the device may malfunction and provide non-correct cutting. In both cases, the user may experience processing which is not necessarily optimal.

[0009] Further, the existing cutting devices may not be able to cut with the desired speed, i.e., the desired amount of food items per hour may not always be satisfactory.

[0010] SUMMARY OF THE INVENTION

[0011] It is an object of embodiments of the disclosure to reduce costs, to increase throughput, and potentially to improve the quality of cutting in food items. Additionally, it may be an object to reduce the footprint in relation to the number of cutting devices in the machine and thereby potentially increase the capacity per unit of area in the factory.

[0012] According to these and other objects, the present disclosure, in a first aspect, provides a cutting apparatus with a conveying device having a support surface configured to convey at least one row i.e., e.g., two rows of food items in a downstream direction between an inlet and an outlet.

[0013] The apparatus comprises at least one multi-axis cutting device having a water jet cutting nozzle being movable in at least two dimensions over the support surface, and at least one single-axis cutting device having a water jet cutting nozzle being movable in at most one dimension over the support surface.

[0014] Since the cutting apparatus utilizes both a multi-axis cutting device and a single-axis cutting device, the cutting tasks can be shared between the cutting devices depending on the complexity of the cutting tasks and thus the need for movability of the water jet cutting nozzle.

[0015] Moreover, when a single-axis cutting device is sufficient for a specific cutting task, the control can be simplified, not least because inverse kinematic considerations are much less complicated when it comes to single-axis manipulation as compared to multi-axis manipulation. Accordingly, the computation involved in the control of the cutting device becomes simpler, and the cutting speed may potentially be increased, i.e., the total amount of cutting meters performed by the apparatus may be increased.

[0016] Moreover, when a single-axis cutting device is sufficient for a specific cutting task, the machine costs can be reduced, not least because a single-axis manipulator is typically cheaper to produce and easier to maintain. Additionally, it may be more robust than a multiaxis manipulator.

[0017] Accordingly, the disclosure provides the opportunity to choose between advanced cutting devices and less advanced cutting devices for optimizing the processing by smart allocation of cutting tasks.

[0018] Finally, the single-axis cutting device may potentially take up less space along the support surface of the conveying device and therefore allow more cutting devices to be used in the same cutting apparatus, thereby further improving the capacity of the cutting apparatus. The reason for the reduced footprint is that the single-axis cutting device may typically have a much smaller size in all other directions than the direction of the single-axis of freedom, particularly when constituted by a linear actuator with a cutting tool.

[0019] Herein, a cutting pattern is a pattern in which the food item is to be cut. The cutting pattern could comprise one or more cutting tasks. A cutting task is a cut performed in one continuous movement of the water jet cutting nozzle. A simple cutting task would be a single straight cut through the food item. This could belong to a cutting pattern by which the food item is simply divided in two. If the cutting pattern divides the food item in four, it will typically consist of at least two cutting tasks, e.g. straight or curved cuts each dividing the food items in two.

[0020] The cutting apparatus may be configured for cutting any kind of food items. However, poultry, pork, beef, fish, and similar animal products may be particularly suitable for the cutting apparatus.

[0021] The conveying device may have a support surface to allow water from the water jet nozzles to penetrate through the support surface. The surface may e.g., be constituted by an endless belt of a metal mesh or similar open mesh allowing penetration of water.

[0022] The inlet and the outlet are simply defined as those locations where the food items enter and leave the support surface. It could also be constituted by a dedicated inlet and outlet, such as a hopper etc.

[0023] Each multi-axis cutting device may have one or more water jet cutting nozzles. Such nozzles are known in the art, and they are typically handled by multi-axis robots such as Delta robots. Typically, such robots offer movement of a tool, in this case the water jet cutting nozzle, in three directions. Additionally, the tool may be rotatable around one, two, or three axes of rotation to thereby provide tilting / angulation of the cutting nozzle.

[0024] Each single-axis cutting device may have one or more water jet cutting nozzles, e.g., nozzles which are identical to those of the multi-axis cutting devices. These cutting nozzles are movable in only one dimension over the support surface, and they may additionally be rotatable about one, two, or three axes of rotation for tilting / angulation of the cutting nozzle, or they may be fixed statically to the single degree of freedom thereby only facilitating cutting along a line in one specific angle.

[0025] The cutting apparatus may comprise a computer system configured to allocate selected cutting tasks to multi-axis cutting devices and other cutting tasks to single-axis cutting devices. The computer system may use a dynamic allocation based on different variables selected ad hoc during processing of the food item, or the computer system may use a more static allocation based on a fixed selection of variables. Moreover, the allocation may be based on an algorithm or based on fixed, empirical, datasets which define the best cutting device for a specific cutting task. In any event, the computer system will consider the complexity of a cutting task and select a single-axis cutting device or a multi-axis cutting device based on the complexity.

[0026] In one example, the allocation which the computer system is programmed to perform for defining a cutting operation comprising the following steps: a) a cutting pattern for a specific food item is received together with a position of the specific food item and thus the position of the cutting pattern on the support surface. b) the cutting pattern is analyzed and converted into a number n of cutting tasks each having a location on the support surface. c) the computer system uses an inverse kinematic model of the single-axis cutting device to confirm which m of the n cutting tasks could kinematically be performed by the single axis cutting device. This step could be carried out in an associated motion controller be external to the computer system, e.g. a motion controller being integrated in the single-axis cutting device, or it could be carried out in a motion control sequence within the computer system. d) the computer system uses a scheduler to allocate as many of the n cutting tasks, herein referred to as m out of the n cutting tasks, which are confirmed in step c) to the single axis cutting device considering the duration of each cut and the available time. e) the computer system uses an inverse kinematic model of the multi-axis cutting device to confirm which q of the n-m cutting tasks could kinematically be performed by the multi-axis cutting device. f) the computer system uses a scheduler to allocate as many of the q cutting tasks which are confirmed in step e) to the multi axis cutting device considering the duration of each cut and the available time. g) the computer system optionally, allocates remaining cutting tasks to subsequent singleaxis or multi-axis cutting devices along the support surface of the conveyor.

[0027] In another example, the allocation which the computer system is programmed to perform comprises the following steps: a) a cutting pattern for a specific food item is received together with a position of the specific food item and thus the position of the cutting pattern on the support surface. b) the cutting pattern is analyzed and converted into a number n of cutting tasks each having a location on the support surface. c) the computer system uses a simple analysis comparing each cutting task with one or more constraints to confirm which, m, of the n cutting tasks could kinematically be performed by the single axis cutting device. The constraints could define e.g., one or more of the following:

[0028] - an angle of a cutting task relative to the conveying direction;

[0029] - a duration of a cutting task and available processing capacity of the single-axis cutting device or multi-axis cutting device; and / or

[0030] - a curvature of the cutting task. d) the computer system uses a scheduler to allocate as many of the n cutting tasks, herein referred to as m out of the n cutting tasks, which are confirmed in step c) to the single axis cutting device considering the duration of each cut and the available time. e) the computer system uses an inverse kinematic model of the multi-axis cutting device to confirm which q of the n-m cutting tasks could kinematically be performed by the multi-axis cutting device. This step could be carried out in an associated motion controller be external to the computer system, e.g. a motion controller being integrated in the multi-axis cutting device, or it could be carried out in a motion control sequence within the computer system. f) the computer system uses a scheduler to allocate as many of the q cutting tasks which are confirmed in step e) to the multi axis cutting device considering the duration of each cut and the available time. g) the computer optionally, allocates remaining cutting tasks to subsequent single-axis or multi-axis cutting devices along the support surface of the conveyor.

[0031] An example of the dynamic allocation based on ad hoc defined variables could be if a number of variables may determine that a single-axis cutting device is assigned to a cutting task. By means of an example, the three variables could be three different constraints, e.g. : 1) the single-axis cutting device is free,

[0032] 2) the cutting task specifies a straight cut at an angle less than 45 degrees to the downstream direction, and

[0033] 3) the cutting task is independent on other cutting tasks, i.e. other cutting tasks do not form an extension of the performed cut.

[0034] During the course of food item processing, the computer system may dynamically adapt and use one or more of these criteria for selection of a single-axis cutting device as opposed to a multi-axis cutting device.

[0035] An example of a static allocation could be if one single variable is always used for the allocation. An example could be that all cuts which are straight cuts, e.g., straight cuts throughout the food item to split the food item in two, or just internal straight cuts within the boundary of the food item, and particularly cuts at an angle less than 45 degrees to the downstream direction defined by the conveying device will always be allocated to a singleaxis cutting device, and all cuts where there is more than 45 degrees angle to the downstream direction will always be allocated to a multi-axis cutting device.

[0036] The computer system may additionally be configured to control each multi-axis cutting device and each single-axis cutting device to execute the allocated cutting tasks on food items in the row of food items. The computer system may, e.g., carry out the functions of providing inverse kinematic solutions for the single-axis cutting device and for the multi-axis cutting device, and to execute, e.g., through a servo controller, the control of the individual drives and the nozzle.

[0037] The computer system may be constituted by one or more CPU(s) with memory and computer system executable code for enabling various functions. Software program instructions and data may be stored on a non-transitory or a transitory computer-readable storage medium, and when the instructions are executed by the CPU the functions associated with those instructions are carried out. The computer system may be implemented in a controller for the cutting apparatus such that the allocation and the control of the individual cutting devices and / or the water jet cutting nozzle are carried out by the computer system.

[0038] In one embodiment, the computer system comprises several independently operating computer units, e.g., formed by a traditional computer unit, e.g., a PC etc. and configured for carrying out the allocation of selected cutting tasks to the multi-axis or single-axis cutting devices. Another, independently operated computer unit could be constituted by a servodrive system for manipulators of the cutting devices, e.g., implemented in a PLC-computer unit. Together the computer units could define the computer system within the meaning of this document.

[0039] When the computer system which controls each multi-axis cutting device is the same computer system which allocates the selected cutting tasks between the cutting devices, a particular robust and simple control structure may be defined. Particularly, knowledge about the capabilities of each cutting device may be used by one and the same computer system both when allocating cutting tasks to a cutting device and when controlling the operation of that cutting device. Should one cutting device be changed, e.g. such that the kinematics of the cutting device changes, then the changes may be entered only in one single system which is then again capable of allocating cutting tasks and controlling the cutting with the changed cutting device.

[0040] The water jet cutting nozzle of each single-axis cutting device may be movable transverse to the downstream direction, particularly perpendicular to the downstream direction.

[0041] The water jet cutting nozzle of each single-axis cutting device may exclusively be movable along a single linear axis. That will simplify the mechanical layout and the inverse kinematic solution and thereby provide a simple task for the computer system allowing potentially faster operation of the cutting apparatus. Moreover, the single linear axis may reduce the complexity of the mechanical layout and thus reduce the costs of the apparatus.

[0042] The water jet cutting nozzle of each single-axis cutting device could be movable while the row(s) of food items are conveyed in the downstream direction, i.e., simultaneously. That means that motion of the single axis cutting nozzle and the support surface in combination forms the shape of the cutting, and it will allow curved cut with a single-axis cutting device which is only movable along a straight axis. For this purpose, the computer system may be configured also to control the operation of the conveying device, i.e., to control servo motors driving the movement of the support surface. Accordingly, the computer system may be configured to carry out an inverse kinematic calculation including the single axis of the single-axis cutting device and the movement of the support surface to provide a solution to each cutting task allocated to the single-axis cutting device.

[0043] Accordingly, the computer system may be configured to control each of the single-axis cutting devices to move the water jet cutting nozzle transverse to the downstream direction while the row(s) of food items are conveyed in the downstream direction to thereby obtain cuts of different shape, including curved or straight cuts. At least one of the at least one single-axis cutting devices may have a water jet cutting nozzle being configured to eject water in a direction which is variable relative to the support surface. This may imply that the water jet cutting nozzle is fixed to the manipulator which makes the single axis move via a bearing, e.g. a regular, single axis bearing, a gimbal bearing, or spherical bearing, i.e., simply a bearing allowing rotation about a single rotation axis, or a bearing allowing rotation about multiple axes of rotation for tilting / angulation purpose. With this feature, cutting angles may be variable relative to the support surface, and the food items may therefore be cut in angled cuts.

[0044] In one embodiment, the single-axis cutting device and / or the multi-axis cutting device is / are mounted with a rotatable two-headed nozzle to refine the cutting process and inter alia allow the keel strip of a poultry butterfly fillet to be removed in one single cutting operation. Such rotatable two-headed nozzles are disclosed in the patents US10611046, US11318580, and US11717939.

[0045] The computer system may be configured for receiving a food item characterizing parameter. This parameter may characterize inter alia a thickness, a change of thickness, a fat and / or membrane content, an existence of a defect, a color of the food item, or a location of the food item. The food item characterizing parameter may be generated in a vision system and it may be communicated in digital form to the computer system.

[0046] The computer system may be configured, based on the received food item characterizing parameter, to make the allocation of the cutting task.

[0047] The food item characterizing parameter may be generated by a vision system.

[0048] The vision system provides visual information related to the food item and the support surface. It then processes the images using a combination of hardware and software and prepares the information for use by the computer system for defining cutting tasks, for allocation of the cutting tasks and / or for other control purposes.

[0049] The vision system may be based e.g. on area scan cameras and / or line scan cameras. Area scan cameras take pictures in a single frame using e.g., a rectangular sensor. The number of pixels in the sensor corresponds to the width and height of the image. Area scan cameras are used for scanning objects that are the same size in terms of width and height. Line scan cameras build an image pixel by pixel. They're suited for taking images of items in motion or of irregular sizes. The sensor passes in a linear motion over an object when taking the pictu re. The vision system provides the ability of the computer system to register various features of the food item being decisive for generation of the cutting pattern and the allocation of the cutting task between the single-axis and multi-axis cutting devices and decisive for the control of the cutting devices, inter alia: the position of the food item on the support surface, the size and shape of the food item and thus e.g. a desired length of the cutting if it is supposed to divide the food item in two, an undesired element in the food item, e.g. fat or bones to be cut away, or the height of the food item, if it is decisive for the way the cutting is planned.

[0050] In one example, the vision system may provide the computer system with sufficient data to perform the cutting to obtain specific weight, size, and / or shape criteria for the food items being cut.

[0051] The vision system may employ one or more cameras, e.g. a video camera, and further image processing equipment, e.g. including analog-to-digital conversion and digital signal processing and / or x-ray. The resulting data is received by the computer system which controls the allocation and / or the cutting devices.

[0052] The vision system may include an optical scanner for generating at least one of a visible light (e.g., greyscale) image, a laser light scattering image, a height map, a hyperspectral image, a multispectral image, etc., of the food item to show one or more of the overall shape / size of the food item, a composition of the food item (e.g., fat. v. lean meat), a height or thickness over the area of the food item, etc. Scanning with the optical scanner can be carried out using a variety of techniques, such as the techniques shown and described in U.S. Patent No. 10654185 as well as U.S. Patent No. 10721947, incorporated by reference herein.

[0053] The vision system may include a video camera to view a food item illuminated by one or more light sources. Light from the light source is extended across the moving conveyor belt to define a sharp shadow or light stripe line, with the area forwardly of the transverse beam being dark. When no food item is being carried by the conveyor belt, the shadow line / light stripe forms a straight line across the belt. However, when a food item passes across the shadow line / light stripe, the upper, irregular surface of the food item produces an irregular shadow line / light stripe as viewed by a video camera (not shown) directed diagonally downwardly on the food item and the shadow line / light stripe. The video camera detects the displacement of the shadow line / light stripe from the position it would occupy if no food item were present on the conveyor belt. This displacement represents the thickness of the food item along the shadow line / light stripe. In some examples, the optical scanner is a single SICK® camera with a single laser light source that is suitable for capturing optical data and generating two or more images / views based on the optical data. For instance, the single camera may be in communication with a separate processor (having or more feature recognition modules or the like) and / or directly with the computer system for generating one or more views from the captured optical data, such as a fat recognition (FRS) object view, a laser scatter object view, and a height mode object view.

[0054] In some examples at least two optical cameras each equipped with a different imaging processor are used. For example, a simple optical camera, for example a greyscale camera, and / or RGB camera and / or IR and / or UV camera and / or a charge coupled device (CCD) and / or a Time-of-Flight (ToF) stereoscopic camera, a stereo camera, a lidar sensor, a structured light sensor, or the like, or combinations thereof, can be used to acquire and / or generate one or more complete images of the food item for detecting certain characteristics, such as, e.g., the outer contour of the food item. Moreover, a second, special camera, for example a multispectral or hyperspectral camera, can be used to acquire images / data of specific regions or characteristics of the food item, such as blood spots, streaks of fat or the like. It should be appreciated that a single camera / scanner may instead be used to capture all the data needed to generate the various images, such as with various imaging processes.

[0055] In some examples at least one X-ray camera is used for identifying e.g., internal parts inter alia unwanted bones which are to be cut away.

[0056] The results of the scanning of the vision system are transmitted to the computer system. The computer system may include circuitry for executing one or more feature recognition modules in a sensor data pre-processing engine for generating views / images from the scan data and / or processing data from the different views. For instance, the sensor data preprocessing engine of the computer system may be configured to generate at least one of a fat recognition (FRS) object view, a laser scatter object view, and a height mode object view of a food item, such as from data captured with the optical scanner.

[0057] The vision system may include artificial intelligence (Al), machine learning and / or deep learning to improve and / or accelerate image processing and to define the cutting pattern and cutting tasks.

[0058] For instance, one or more classification machine learning models may be used for determining a type of food item, one or more image segmentation machine learning models may be used to identify features of a food item, one or more fat / lean image segmentation machine learning models may be used to identify cut paths between fat and lean of a food item, one or more multi-class image segmentation machine learning models may be used to identify three or more regions or features in a food item image, and one or more region of interest (ROI) machine learning models may be used to define a proposed portion or outline of an area / object of a food item.

[0059] To support processing of the machine learning models, a computing device associated with the vision system may be configured as a local, high power computing device.

[0060] In that regard, labeled image data may be used as training data for machine learning models that are used for defining cutting patterns for cutting the food items of the labeled food item type (e.g., food item source, food item orientation, food item maturity level, etc.) and / or control parameters of the cutting process (e.g., belt speed at which the image data was captured, belt type, environment, etc.). For instance, image data that has a label indicating a food item type and / or a food item process will be used to train a machine learning model used for outputting information relevant for cutting the food item type and / or using the food item cutting process. In that manner, when a food item type and / or food item machine process is identified during data processing, the machine learning model specifically trained using training data labeled as that food item type and / or food item cutting process may be executed by the computer system to define data for the allocation of cutting tasks and for controlling the single-axis and the multi-axis cutting device.

[0061] For efficient management of the Al system, the computer system may utilize or include a NVIDIA Jetson Orin package, such as an Advantech MIC-711-OX.

[0062] The computer system may execute one or more machine learning models that output information based on an image(s) from the vision system. For instance, the computer system may output information regarding a location of a food item or features therein, e.g., bones, sciatic nerve, cut lines, outline, fat or lean area, or an outline of the food item and any features therein, e.g., bones, fat / lean, foreign objects, etc., or a region of interest of the food item, e.g., an area comprising a maximum nominal height of the food item, or a classification of the food item, e.g., sirloin pork chop, center loin pork chop, etc., or a location of a conveyor belt component relative to a coordinate system, etc.

[0063] In one example, a classification machine learning model may be configured to classify a type of food item, such as a type of sub-primal cut.

[0064] In other examples, an image segmentation machine learning model of the machine learning model may be configured to identify features of a food item, identify separate food items on the conveying device by segmenting or "cutting out" an object(s), feature(s), etc., in an image as output. The image segmentation machine learning model may incorporate the Segment Anything Model (SAM) available from Meta Al, FastSAM from Ultralytics, or another suitable image segmentation model using image segmentation techniques.

[0065] An image segmentation machine learning model may use images from the vision system to identify features of a food item. For instance, a food item feature image segmentation machine learning model may provide an outline of bones as output based on an X-ray image. The output may be a binary image or a map showing the location of the bones, with every pixel indicating the presence or absence of bone. In such an instance, only a single channel image, such as an x-ray image, may be needed as input, saving processing time and capacity. For instance, an image segmentation machine learning model may be trained to identify / outline bones of a food item (and optionally an overall outline of a food item).

[0066] A binary output image of the model may include the input image of the food item with an outline of the bones (e.g., a beef rib) and an overall outline of the food item (e.g., a rack of beef ribs).

[0067] The output can define cut lines for the food item around one or more undesired objects.

[0068] A fat / lean boundary image segmentation machine learning model may also be used to identify fat / lean boundaries in various food items. For instance, the fat / lean boundary image segmentation machine learning model may provide an image having an outline of fat and / or lean areas in a food item as output based on an optical image(s) sent from the vision system. The model output may be, for instance, a marked-up version of the input image with computer-generated annotations showing outlines of the fat and / or lean areas in a food item.

[0069] In one example, a fat / lean boundary image segmentation machine learning model may be configured to identify fat caps at a top and bottom of a poultry thigh, such as a chicken thigh. For instance, a fat / lean boundary image segmentation machine learning model may be trained to identify / outline a mid-section or lean meat section of a chicken thigh between top and bottom fat caps of the thigh.

[0070] A binary output image of the model may include the input image of e.g. a keel strip of a poultry butterfly fillet to be removed, e.g., with an outline of substantially only the meat section of the chicken breast. The output image can be used by the computer system to define cut lines for splitting the chicken breast of a butterfly, to trim or cut off the extraneous fat pieces, or to exclude the extraneous fat pieces from an estimated size of the food item (e.g., a weight or volume). In one example, the allocation of cutting tasks could be carried out by Artificial intelligence. The cutting apparatus may use a cutting device selection machine learning model. This machine learning model may use the defined cut lines and provide cutting tasks each comprising one or more cut lines. A binary output from the cutting device selection machine learning model may define allocation of cutting tasks to individual single-axis and multi-axis cutting devices. Training data for cutting device selection machine learning model may include a plurality of cutting patterns tagged with identification of single-axis cutting device or multi-axis cutting device for individual cutting tasks of the cutting pattern.

[0071] The cutting apparatus may comprise a weight or volume estimation system for determining the weight or volume of each food item in the row(s) of food items. In this case, the computer system is configured to receive the weight or volume and to make the allocation of the cutting tasks based on the weight or volume, e.g. such that each piece of food has a desired weight or volume after the cutting process.

[0072] The cutting apparatus may include a data storage with predefined cutting tasks and the computer system may be configured to make the allocation of at least one of the predefined cutting tasks to one of the at least one multi-axis cutting devices and to allocate another one of the predefined cutting tasks to one of the at least one single-axis cutting devices, such that the single-axis cutting devices and the multi-axis cutting devices carry out different predefined cutting tasks.

[0073] Examples of predefined cutting tasks include linear cutting tasks at different angles to the downstream direction, or curved cuts with different curvatures etc.

[0074] Particularly when the single-axis cutting device moves perpendicular to the conveying direction, it becomes impossible to cut the food item perpendicular to the conveying direction unless the support surface is stopped. Since stopping of the continuous process is generally undesired, the single-axis cutting devices may become unable to carry out cutting tasks which are "too-close" to a perpendicular direction relative to the conveying direction. For this purpose, the predefined cutting tasks may be defined with an angle relative to the downstream direction and the allocation may be based on the angle.

[0075] In one example, a threshold angle may be statically provided in the data storage, and each cutting task may be compared with the threshold angle. If the angle of a cutting task exceeds the threshold angle, the cutting task is allocated to a multi-axis cutting device, and if not, it is allocated to a single-axis cutting device. In one example, a threshold angle may be dynamically varied, e.g. based on a speed of the conveying device, and / or a length of the cutting task. The dynamic parameter is stored in the data storage, and each cutting task may be compared with the dynamically defined threshold angle. If the angle of a cutting task exceeds the threshold angle, the cutting task is allocated to a multi-axis cutting device, and if not, it is allocated to a single-axis cutting device.

[0076] Since the ability of the single-axis cutting device to cut close to the perpendicular direction depends on the conveying speed of the food items on the support surface, the allocation may also be based on the speed in addition to the angle. In one example, the threshold may be dynamically updated by the speed of the conveyor device.

[0077] The data storage may comprise data related to predefined types of food items, e.g. including one or more predefined cutting patterns for each predefined food item type.

[0078] The computer system may be configured either to receive information defining that the food item to be processes is of the predefined type, or to recognize if a food item in the row(s) of food items is one of the predefined types of food items. The computer system may be configured e.g., to recognize food items like poultry breast fillets, poultry butterflies, salmon fillets, specific kinds of fruits or vegetables, and / or generally various cuts of meat, e.g., by use of aforementioned image segmentation machine learning model.

[0079] The computer system may be configured to make the allocation of cutting task to one of the at least one multi-axis cutting devices or to one of the at least one single-axis cutting devices based on the recognition, such that the single-axis cutting devices and the multi-axis cutting devices carry out cutting tasks on different types of food items.

[0080] The computer system may be configured to determine the duration of each cutting task when executed on the single-axis cutting device or the multi-axis cutting device, and the allocation could be based on the duration. This may be implemented by incorporating a single-axis cutting duration parameter for the single-axis cutting device and a multi-axis cutting duration parameter for the multi-axis cutting device in the computer system and using the parameter with a defined length of a cutting task to find the estimated duration for the single -axis cutting device and the multi-axis cutting device respectively.

[0081] The computer system may, e.g., determine that it takes longer time to execute a cutting task by a multi-axis cutting device than by a single-axis cutting device. In that case, the cutting task could be allocated to the fastest of the cutting devices. For this allocation, a secondary consideration may also be included, e.g., relative to capacity. I.e. even though one cutting device is the fastest cutting device, the cutting task may be allocated to another cutting device which is less loaded and has free capacity. Accordingly, the allocation may be based on several factors, inter alia capacity, price, speed, or quality.

[0082] The water jet cutting nozzle of each multi-axis cutting device may be able to move in the downstream direction, and the water jet cutting nozzle of each single-axis cutting device may be unable to move in the downstream direction. This may allow only a subset of the cutting devices to follow the food items when they move in the downstream direction. An advantage of following the food item could be that more time-consuming cuts could be carried out while the food item is conveyed, and a downside could be that the cutting task will then take more space along the length of the support surface. When the cutting apparatus has at least one cutting device with this capability and at least one cutting device without this capability, it enables both advanced cutting and less space requiring cutting to be performed.

[0083] At least one of the at least one single-axis cutting devices and at least one of the at least one multi-axis cutting devices may be configured to cut simultaneously in the same food item or in different food items, and optionally to cut independently, such that each cutting device of the apparatus cuts independent of the other cutting devices. That may reduce processing time by allowing more cutting devices to process the same or different food items simultaneously.

[0084] Each single-axis cutting device and each multi-axis cutting device may be configured for cutting food items placed in different rows on the support surface, and the single-axis cutting devices and the multi-axis cutting devices may be located consecutively, i.e., one after the other in the downstream direction.

[0085] At least one of the at least one single-axis cutting devices could be arranged between two multi-axis cutting devices, or wherein at least one of the at least one single-axis cutting devices is arranged before or after a multi-axis cutting device, or between two multi-axis cutting devices.

[0086] At least one of the at least one multi-axis cutting devices could be configured for movement in a direction towards or away from the support surface, e.g. in a direction perpendicular to the support surface.

[0087] At least one of the at least one single-axis cutting devices comprises a linear bearing with at least one bearing component slidable along a linear rod.

[0088] If each of the single-axis cutting devices and each of the multi-axis cutting devices extends a certain distance away from the support surface, then at least one multi-axis cutting device may extend a larger distance away from the support surface. I.e. the height of the multi-axis cutting devices may exceed the height of the single-axis cutting devices. This may allow the single-axis cutting devices to be arranged at least partly under the multi-axis cutting devices which may save space in the downstream direction along the conveying device.

[0089] In a second aspect, the disclosure provides a method of executing cutting tasks for cutting a food item in a cutting apparatus of the kind pertaining to the first aspect of the disclosure.

[0090] The method comprises:

[0091] - conveying the food item on the support surface,

[0092] - determining a cutting pattern of the food item,

[0093] - defining based on the cutting pattern, a number of cutting tasks for the food item

[0094] - allocating, based on the cutting pattern, at least one cutting task to at least one of the at least one multi-axis cutting devises and / or to at least one of the at least one single-axis cutting devices, and

[0095] - executing the at least one allocated cutting task and thereby obtaining a cut food item.

[0096] The method may further comprise determining a 3D shape, a thickness, a change of thickness, a fat content, an existence of a defect, a color of the food item, or a location(s) on the conveyor of the food item to be cut. This may be determined by a vision system communicating a digital representation of the food item.

[0097] The step of defining the cutting tasks may comprise a step of selecting pre-defined cutting tasks for the food item based on the cutting pattern, wherein the pre-defined cutting tasks are selected from a list of predefined cutting tasks. The step may include defining a cutting pattern, and both steps may be carried out by use of Al, e.g., using the previously mentioned machine learning segmentation models trained with tagged images including the cutting pattern and / or cutting tasks for food items of a specific kind, e.g., images of a poultry butterfly fillet with a keel strip to be removed in one single cutting operation where the breast fillets are separated.

[0098] The cutting pattern defines a pattern by which the food item is cut, whereas each cutting task solves the cutting of a subset of the cutting pattern. Accordingly, a cutting pattern may require one or more cutting tasks to be defined. How many cutting tasks that are required for a cutting pattern may inter alia depend on the complexity of the cutting pattern and the desired use of either single-axis or multi-axis cutting devices. The method may particularly comprise establishing an identifier selected from the group consisting of a shape, a size, a weight, a food type, and a user demand related to the food item or to the cutting task, and making the allocation based on the identifier.

[0099] This could be exemplified as follows:

[0100] A shape of a food item, e.g. a shape with a curvature below a certain radius may trigger the allocation of cutting to a multi-axis cutting device, and a food item with a very straight shape in a specific direction may trigger allocation of cutting to a single-axis cutting device.

[0101] A size of a food item above a certain threshold may trigger the allocation of cutting to a multi-axis cutting device, and size below the threshold may trigger allocation of cutting to a single-axis cutting device.

[0102] A weight of a food item above a certain threshold may trigger the allocation of cutting to a multi-axis cutting device, and weight below the threshold may trigger allocation of cutting to a single-axis cutting device.

[0103] A specific food type, e.g. a chicken butterfly may trigger the allocation of specific cuts to a single-axis cutting device, e.g., for splitting the butterfly into two breast fillets.

[0104] The method may further comprise establishing a user demand, e.g., a demand which is keyed in or selected on a human interface and being related to the food item or to the cutting task. The method may include making the allocation based on the user demand. As an example, the user demand may specify a specific way of portioning or cutting a specific food item, e.g., that a poultry butterfly is to be split through the keel strip, or that the poultry fillets should be split by removal of the keel strip. Another demand could be to make a specific kind of cutting, e.g. cube cutting of a chicken fillet may trigger the allocation of specific cuts to a single-axis cutting device or a multi-axis cutting device.

[0105] The method may include any step which is implicit in view of the cutting apparatus according to the first aspect of the disclosure.

[0106] LEGENDS TO THE FIGURE

[0107] Fig. 1 illustrates a side view of a cutting apparatus with four multi-axis and one single-axis cutting devices; Fig. 2 illustrates the device in a perspective view;

[0108] Fig. 3 illustrates a top view of an apparatus for cutting food items;

[0109] Figs. 4-5 illustrate salmon fillets with indicated cutting tasks;

[0110] Fig. 6 illustrates a section end-view of a multi-axis and a single-axis cutting device;

[0111] Figs. 7 and 8 illustrate details related to angles of cutting tasks;

[0112] Figs. 9 and 10 illustrate details related to cutting patterns and cutting path;

[0113] Fig. 11 illustrates different modules for a cutting apparatus; and

[0114] Figs. 12-13 are flowcharts illustrating functions of the computer system.

[0115] DETAILED DISCLOSURE OF THE INVENTION

[0116] Fig. 1 illustrates a cutting apparatus 1 for executing cutting tasks on food items seen in a side view, and Fig. 2 illustrates the device in a perspective view. The device comprises a conveying device 2 having a support surface 3 configured to convey at least one row of food items 4 in the downstream direction between an inlet 5 and an outlet 6. The downstream direction is indicated by the arrow 7.

[0117] The support surface 3 is constituted by a metal mesh facilitating water jet cutting and allowing the water to pass through the support surface.

[0118] Four multi-axis cutting devices 8 are equipped with water jet cutting nozzles 9. The illustrated multi-axis cutting devices are parallel arm robots, also called delta robots, but it could be any kind of multi-axis manipulator facilitating movement in more than one dimension, e.g. a Selective compliance articulated robot arm (SCARA), a regular industrial robot (anthropomorphic robot) e.g. with five or six rotational joints, or any kind of robot with more than one degree of translatory freedom, or different kinds of cartesian robots in which two or more linear axes are combined, e.g., two or more belt driven linear drives providing two or more linear degrees of freedom. A delta robot of the kind illustrated in Figs. 1 and 2 is a type of parallel robot having three arms connected to universal joints at the base at the top. Parallelograms in the arms maintain the orientation of the water jet nozzle 9. Delta robots can be fast and precise.

[0119] The cutting apparatus 1 further comprises one single-axis cutting device 10 having a water jet cutting nozzle 11. This nozzle is only movable in one dimension over the support surface. In this specific example, the nozzle moves on a linear bearing in a direction perpendicular to the downstream direction.

[0120] In addition to the movability of the multi-axis cutting devices and single-axis cutting device, one or more of the five cutting devices may be able to change the angle of the water jet cutting nozzle relative to the support surface.

[0121] The cutting apparatus comprises a vision system with an imaging device, in this case a camera 12, pointing towards the support surface for capturing images or video of the food items when they are conveyed on the support surface. As an alternative to the illustrated camara, the imaging device could be a line scanner, e.g. a one-dimension line camera.

[0122] Computer system 13 is programmed to control the cutting operation. The computer system is configured to allocate selected cutting tasks to multi-axis cutting devices and other cutting tasks to single-axis cutting devices. Computer system 13 is in data communication with each of the multi-axis and single-axis cutting devices and with the camera 12, and preferably also with motors 14 (illustrated only in Fig. 2), driving the conveyor via the communication line. The communication line is illustrated by the dotted line 15. It could be wireless or wired communication.

[0123] The Delta robots each define a work area illustrated by the circle 16. This work area is larger than the work area of the single-axis cutting device 10. The single-axis cutting device 10 is located outside the work area of the adjacent multi-axis cutting device.

[0124] The vision system is configured for determining a geometrical description, such as a height, a width, a length, or a characteristic shape of each food item 4 in the row(s) of food items.

[0125] The vision system could be implemented in the illustrated computer system 13, and / or in another computer system. The vision system sends digital data to the computer system related to the physical parameter(s) on interest. The digital data may include one or more outputs of a machine learning model(s), such as described above. The computer system receives digital data from the vision system pertaining to the geometrical description of interest, e.g. in the form of a picture, a classification, or a mask segmentation etc.

[0126] Based on the geometrical description, computer system 13 allocates the cutting tasks between the cutting devices. The vision system is placed upstream the cutting devices to establish the geometrical description prior to the allocation of cutting tasks to specific cutting devices. Alternatively, or additionally, the apparatus comprises an imaging device after one or all of the cutting devices to capture images for quality control or documentation or exact orientation of the individual food items. In this case, an imaging device arranged after the first cutting device but before the last cutting device may validate the cutting tasks executed by the cutting devices before the imaging device. Should cutting tasks be missing, or be carried out with a bad quality, computer system 13 may instruct the cutting devices after the imaging device to correct the error.

[0127] The imaging device, in this case the camera 12, is in a fixed position relative to the support surface and therefore captures the geometrical description of the food item relative to the position of the food item on the support surface.

[0128] Fig. 3 illustrates a top view of an apparatus 1 comprising a multi-axis cutting device 8 in the form of a SCARA robot, and a single-axis cutting device 10 in the form of a nozzle 11 suspended on a linear bearing structure comprising a rod 30, and a cart 31 movable linearly back and forth on the rod 30. The rod may constitute a stator, and the cart may constitute a rotor of a stator-and-rotor electro-motor structure whereby the linear movement can be electrically actuated and controlled. Alternatively, the single-axis cutting device 10 could have a piston and cylinder structure where the piston is moved by pressure of a medium, e.g., water or compressed air etc.

[0129] Fig. 3 illustrates that the single-axis cutting device 10 as allocated to the longitudinal cuts extending essentially in the downstream direction. These cuts require only very little movement of cart 31 on the rod 30 and can therefore be carried out with a high level of precision and at high speed. The multi-axis cutting device 8, i.e. in this case the SCARA robot, is allocated to the transverse cuts, and particularly to cuts having a curved shape. These curved shapes fit very well to the rotational joints of the SCARA robot, and it is therefore suitable for these cutting tasks.

[0130] In addition to, or as an alternative to the SCARA robot, the apparatus in Fig. 3 could have several other multi-axis robots, e.g. a Delta robot of the kind illustrated in Figs. 1 and 2. In addition to the vision system 12, the apparatus in Fig. 3 comprises a weight estimation system, e.g., comprising the illustrated weight 32, for determining the weight of each food item in the row(s) of food items. A signal indicative of the weight is sent to the computer system 13 which makes the allocation of the cutting tasks based on the weight and / or based on the vision system. The apparatus may in general have a vision system and / or a weighing system.

[0131] In addition to or as an alternative to the weight 32, the weight estimation system may comprise a 3D image-based system configured to estimate the weight based on an image, particularly a 3D image. This may also be based on a known density of the food item.

[0132] The apparatus in Fig. 3 comprises data storage 33 in which there are several predefined cutting tasks. By means of examples, such predefined cuts could be keel strip cuts for separating poultry breast fillets of a butterfly, fat trimming cuts for a specific food item, e.g. for trimming fat from poultry butterflies / fillets or from salmon fillets, blood spot removal cuts predefined for cutting away blood spots, or predefined cuts related to predefined ways of portioning a specific food item.

[0133] Fig. 4 illustrates several predefined cuts related to salmon portioning. For each salmon fillet, the white lines form a cutting pattern, and each of the white lines may be seen as individual cutting tasks. Both the cutting pattern and the individual cutting tasks may be predefined in the data storage, and upon receiving a geometrical representation of the food item or upon selecting a food item in the memory of the computer system 13, the computer system may apply predefined cutting patterns and cutting tasks, and the computer system may allocate cutting tasks to cutting devices based on an allocation criterion.

[0134] Additionally, the data storage may comprise data related to predefined types of food items, e.g., poultry butterfly, salmon fillet, steak etc., and the computer system is configured to recognize e.g. from the shape obtained by vision, or from the weight, or a specific dimensions etc., or from a user input in a machine interface, if a food item in the row(s) of food item is one of the predefined types of food items.

[0135] Once the food item is identified, the computer system can make the allocation of cutting task to one of the at least one multi-axis cutting devices or to one of the at least one single-axis cutting devices based on the recognition.

[0136] The computer system may be configured to determine the duration of each cutting task when executed on the single-axis cutting device or the multi-axis cutting device. Fig. 5 illustrates standard cutting tasks 50, 51, 52, and 53 for a salmon filet. The standard cutting task 50 is a transverse cut not extending through the width of the salmon, the standard cutting task 51 is a transverse cut extending through the width of the salmon, the standard cutting task 52 is a longitudinal cut not extending through the length of the salmon, and the cutting task 53 is a longitudinal cut extending lengthwise in the salmon filet and separating the lower piece 54.

[0137] For each pre-defined cut, the computer system may know:

[0138] - which of the cutting devices can carry out the cut,

[0139] - what speed each of the cutting devices identified in a) can operate with when carrying out the cut, and

[0140] - an expected quality which can be expected to be achieved by each cutting device.

[0141] With this knowledge, the computer system may make the allocation of a cutting device to a specific cutting task based on one or more of the below allocation criteria:

[0142] - to obtain the fastest cutting

[0143] - to load balance all the available cutting devices, and / or

[0144] - to obtain the best quality.

[0145] Each single-axis cutting device and each multi-axis cutting device are configured for cutting food items placed in different rows on the support surface. This is illustrated in Fig. 3 where the food item 4' is displaced relative to the other food items and thereby defines its own row on the support surface. The computer system may be configured to make the allocation based on the location of the food item on the support surface, e.g. such that all food items conveyed in a left-side row of the support surface is allocated to a specific one of the cutting devices and the other food items are allocated to another cutting device.

[0146] Fig. 6 illustrates a cross-sectional end view of a single-axis and a multi-axis cutting device. The single-axis cutting devices are located in a space illustrated with the parenthesis 60, i.e., closer to the support surface than the multi-axis cutting devices illustrated behind the singleaxis cutting devices. The multi-axis cutting device defines a free space marked with 61, i.e., under its control unit 62 at the top of the multi-axis cutting device. In this way, the height of the cutting devices illustrated with umbrella-parenthesis 63 is utilized very well.

[0147] The single-axis cutting device is partly located below this control unit and the apparatus may therefore be made with the same number of water jet nozzles, or an increased number of water jet nozzles but at similar or less space in the downstream direction thereby potentially reducing the footprint when comparing to cutting apparatus with only multi-axis cutting devices.

[0148] Fig. 7 illustrates standard cutting tasks 70, 71, 72, 73, 74, and 75 for a poultry butterfly. The cutting task 75 comprises cutting segments extending transverse to the conveying direction. Accordingly, the cutting task 75 includes segments with angles to the downstream direction which exceeds a threshold defining the limit for the single-axis cutting device.

[0149] Due to the ability of the single-axis cutting device to only move in one direction, it may be impossible to use the single-axis cutting device when the cutting task extends "too" perpendicular to the conveying direction - particularly when the single-axis cutting device moves perpendicular to the conveying direction. For such a single-axis cutting device to move perpendicular to the food item in the conveying direction, it would require the conveyor to stop moving which is undesired.

[0150] The computer system comprises data storage with predefined cutting tasks defined with an angle relative to the downstream direction. Accordingly, the computer system can recognize the cutting task 75 as a cutting task for which the multi-axis cutting device is necessary since the single-axis cutting device is not capable of moving perpendicular to the food object when the food object is moved on the conveyor. The remaining cutting tasks 70-74 do not exceed the threshold angle and could all be performed by single-axis cutting devices, however not by the same single-axis cutting device, but by different single-axis cutting devices along the conveyor.

[0151] Fig. 8 illustrates the aspect of a threshold angle 80 of the cutting direction 81 relative to the conveying direction 82. When the angle 80 exceeds a threshold which may depend on the conveying speed, the single-axis cutting device may not be capable of performing the cutting task which is then allocated to a multi-axis cutting device.

[0152] Figs. 9 and 10 illustrate details related to cutting patterns. In Fig. 9, the cutting pattern 90 encircles a keel strip of a poultry butterfly. This pattern may e.g. indicate an undesired part of the poultry butterfly, e.g., the keel strip, and it may therefore be used for defining cutting tasks. In one example, the corresponding cutting tasks may comprise two independent cutting tasks indicated by numeral 91 and 92 which may be cut individually by two separate single-axis cutting devices, or by one multi-axis cutting device. In another example, the cutting task may be one single task essentially following the entire cutting pattern 90. Such a more complex cutting task could be carried out by a multi-axis cutting device. Fig. 10 illustrates a poultry butterfly which is split into two breast fillets. The two breast fillets are cut according to a specific cutting pattern making a fillet 100 and five cubes 101 on each side of the butterfly split. This pattern could be split into a number of cutting tasks. In Fig. 10, the cutting task 110 for splitting the butterfly into two breast fillets could be carried out by a single-axis cutting device whereas the other cutting tasks could be carried out by multiaxis cutting devices. The task of defining the cutting pattern and the cutting tasks could be carried out by use of Al, or it could use other methods. In one example, it is simply a predefined pattern and / or predefined cutting tasks defined for a specific type of food item being processed in the cutting apparatus.

[0153] Fig. 11 illustrates different modules 120, 121, 122 for a cutting apparatus. The module 120 comprises a multi-axis cutting device and two single-axis cutting devices. The module 121 comprises a multi-axis cutting device. The module 122 comprises a single-axis cutting device. Each module can be combined with other modules to make an apparatus with a desired number of single-axis and multi-axis cutting devices. The set of modules may comprise a single device, a double device or a triple device, or a quattro device module with different combinations of single-axis and double-axis cutting devices. The modular system enables a custom specific capacity to be obtained and further allows retrofitting or removal of modules depending on a changed capacity demand. This is illustrated in Fig. 11.

[0154] Fig. 12 illustrates by a flowchart programming of a CPU constituting at least a part of the computer system.

[0155] In process header A, a cutting pattern for a specific food item is received together with a position of the specific food item. The cutting pattern could be received from a cutting pattern generator illustrated in the flowchart in Fig. 13. Receival of the cutting pattern triggers process A having the title process food item no. n.

[0156] In process B, the cutting pattern is analyzed and converted into a number n of cutting tasks nl-nxeach having a well-defined location on the support surface.

[0157] The result C of the process B is an array of splines, e.g. defined piecewise by polynomials, where each spline defines a cutting task.

[0158] In process D, each spline is divided into a plurality of points, and an inverse kinematic model of the single-axis cutting device including the freedom of movement defined by the conveyor is applied to the points on each spline. Process D could be implemented using inter alia the Jacobian inverse technique or other techniques developed for providing inverse kinematic solutions for robots.

[0159] The result E of process D is an array of positions of the water jet cutting nozzle on the singleaxis cutting device where each position corresponds to a point on the splines, along with the inverse kinematic solution to reach those points.

[0160] The result F is a list of those splines where at least one position of the water jet cutting nozzle on the single-axis cutting device is outside a limit of the single-axis cutting device, i.e. where the inverse kinematic solution is an empty space, i.e. thereby marking that cutting task not executable on the single-axis cutting device.

[0161] In process G, an inverse kinematic model of the multi-axis cutting device including the freedom of movement defined by the conveyor is applied to the points on each spline of the result list F.

[0162] The result H of the process F is an array of positions of the water jet cutting nozzle on the multi-axis cutting device where each position corresponds to a point on the splines in list F.

[0163] In Fig. 13, the flowchart comprises two additional processes, I and J. Process I is a scheduler which schedules the work of the single-axis cutting device, and J is a scheduler which schedules the work of the multi-axis cutting device.

[0164] The scheduler I maps the duration of each cutting task in the result list E against a schedule for the single-axis cutting device, and the scheduler J maps the duration of each cutting task in the result list H against a schedule for the multi-axis cutting device. If there is insufficient time for the single-axis cutting device to cut all tasks in the result list E, the scheduler may move cutting tasks to the result list H, vice versa, and if there is insufficient time for both the single-axis cutting device and the multi-axis cutting device, the scheduler may submit a control command reducing the speed of the conveyor or allocate cutting tasks to other singleaxis or multi-axis cutting devices downstream the conveyor.

Claims

CLAIMS1. A cutting apparatus (1) for executing cutting tasks on food items (4), comprising:- a conveying device (2) having a support surface (3) configured to convey at least one row of food items (4) in a downstream direction between an inlet (5) and an outlet (6),- at least one multi-axis cutting device (8) having a water jet cutting nozzle (9) being movable in at least two dimensions over the support surface (3),- at least one single-axis cutting device (10) having a water jet cutting nozzle (11) being movable in at most one dimension over the support surface (3), the cutting apparatus comprising a computer system (13) configured to allocate selected cutting tasks to multi-axis cutting devices and other cutting tasks to single-axis cutting devices.

2. The cutting apparatus according to claim 1, wherein the computer system is configured to control each multi-axis cutting device and each single-axis cutting device to execute the allocated cutting tasks on food items in the row(s) of food items.

3. The cutting apparatus according to any of the preceding claims, wherein the water jet cutting nozzle of each single-axis cutting device is movable transverse to the downstream direction.

4. The cutting apparatus according to any of the preceding claims, wherein the water jet cutting nozzle of each single-axis cutting device is exclusively movable along a single linear axis.

5. The cutting apparatus according to any of the preceding claims, wherein the water jet cutting nozzle of each single-axis cutting device is movable while the row(s) of food items are conveyed in the downstream direction.

6. The cutting apparatus according to any of the preceding claims, wherein the computer system is configured to control each single-axis cutting devices to move the water jet cutting nozzle transverse to the downstream direction while the row(s) of food items is conveyed in the downstream direction to thereby obtain curved or straight cuts.

7. The cutting apparatus according to any of the preceding claims, wherein at least one of the at least one single-axis cutting devices has a water jet cutting nozzle being configured to eject water in a direction which is variable relative to the support surface.

8. The cutting apparatus according to any of the preceding claims, wherein the computer system is configured to receive a food item characterizing parameter and to make the allocation of the cutting tasks based on the food item characterizing parameter.

9. The cutting apparatus according to claim 8, wherein the food item characterizing parameter defines a shape, a thickness, a change of thickness, a fat content, an existence of a defect, a color of the food item, or a location(s) on the conveyor of the food item to be cut.

10. The cutting apparatus according to claim 9, wherein the shape is relative to the support surface.

11. The cutting apparatus according to any of the preceding claims, comprising a volume and / or weight estimation system (32) for determining a volume and / or a weight of each food item in the row(s) of food items, and wherein the computer system is configured to receive the volume and / or the weight and to make the allocation of the cutting tasks based on the volume and / or the weight.

12. The cutting apparatus according to any of the preceding claims, comprising a data storage with predefined cutting tasks and wherein the computer system is configured to make the allocation of at least one of the predefined cutting tasks to one of the at least one multi-axis cutting devices and to allocate another one of the predefined cutting tasks to one of the at least one single-axis cutting devices, such that the single-axis cutting devices and the multi-axis cutting devices carry out different predefined cutting tasks.

13. The cutting apparatus according to claim 12, wherein the predefined cutting tasks are defined with an angle relative to the downstream direction, and wherein the allocation is based on the angle.

14. The cutting apparatus according to claim 13, wherein the allocation is based on the angle combined with a speed at which the row(s) of food items is conveyed by the conveying device on the support surface.

15. The cutting apparatus according to any of the preceding claims, comprising a data storage comprising data related to predefined types of food items and wherein the computer system is configured to recognize if a food item in the row(s) of food items is one of thepredefined types of food items, and to make the allocation of cutting task to one of the at least one multi-axis cutting devices or to one of the at least one single-axis cutting devices based on the recognition, such that the single-axis cutting devices and the multi-axis cutting devices carry out cutting tasks on different types of food items.

16. The cutting apparatus according to any of the preceding claims, comprising a data storage comprising different cutting programs which are selected depending on weight or volume of food items, or food item type, and the multi-axis or single-axis cutting device is selected based on the selected cutting program.

17. The cutting apparatus according to any of the preceding claims, wherein the computer system is configured to determine a duration of each cutting task when executed on the single-axis cutting device or the multi-axis cutting device, and wherein the allocation is based on the duration.

18. The cutting apparatus according to any of the preceding claims, wherein the water jet cutting nozzle of each multi-axis cutting device is able to move in the upstream or downstream direction, and the water jet cutting nozzle of each single-axis cutting device is unable to move in the upstream or downstream direction.

19. The cutting apparatus according to any of the preceding claims, wherein at least one of the at least one single-axis cutting devices and at least one of the at least one multi-axis cutting devices are configured to cut simultaneously.

20. The cutting apparatus according to any of the preceding claims, wherein each singleaxis cutting device and each multi-axis cutting device are configured for cutting food items placed in different rows on the support surface.

21. The cutting apparatus according to any of the preceding claims, wherein the singleaxis cutting devices and the multi-axis cutting devices are located consecutively in the downstream direction.

22. The cutting apparatus according to any of the preceding claims, wherein at least one of the at least one single-axis cutting devices is arranged between two multi-axis cutting devices, or wherein at least one of the at least one single-axis cutting devices is arranged before or after a multi-axis cutting device.

23. The cutting apparatus according to any of the preceding claims, wherein at least one of the at least one multi-axis cutting devices is configured for movement towards or away from the support surface.

24. The cutting apparatus according to any of the preceding claims, wherein at least one of the at least one single-axis cutting devices comprises a linear bearing with at least one bearing component slidable along a linear rod.

25. The cutting apparatus according to any of the preceding claims, wherein the singleaxis cutting devices and the multi-axis cutting devices each extends a distance away from the support surface, and wherein at least one multi-axis cutting device extends a larger distance away from the support surface.

26. The cutting apparatus according to any of the preceding claims, wherein the computer system (13) comprises a multi-axis kinematic model of the multi-axis cutting device, a single-axis kinematic model of the single-axis cutting device and a task planner, where the task planner comprises a data input for receiving a cutting pattern defining a plurality of cutting tasks associated with one single food item and a task allocator which can allocate each cutting task to the multi-axis cutting device based on the multi-axis kinematic model or to the single-axis cutting device based on the single-axis kinematic model.

27. The cutting apparatus according to claim 26, wherein the task planner comprises a priority module which, based on the single-axis kinematic model can identify all cutting task(s) which can be allocated to the single-axis cutting device and thereby reduce the number of cutting tasks allocated to the multi-axis cutting device to a minimum.

28. A method of executing cutting tasks for cutting food items in a cutting apparatus comprising- a conveying device having a support surface configured to convey at least one row of food items in a downstream direction between an inlet and an outlet,- at least one multi-axis cutting device having a water jet cutting nozzle being movable in at least two dimensions over the support surface,- at least one single-axis cutting device having a water jet cutting nozzle being movable in at most one dimension over the support surface,the method comprising :- conveying the food item on the support surface,- determining cutting pattern of food items,- allocating, based on the cutting pattern, at least one cutting task to at least one of the at least one multi-axis cutting devices and / or to at least one of the at least one single-axis cutting devices, and- executing the at least one allocated cutting tasks and thereby obtaining a cut food item.

29. The method according to claim 28, comprising determining a 3D shape, a thickness, a change of thickness, a fat content, an existence of a defect, a color of the food item, or a location(s) on the conveyor of the food item to be cut.

30. The method according to claim 28 or 29, comprising selecting a number of cutting tasks for the food item based on the cutting pattern, wherein the cutting tasks are selected from a list of predefined cutting tasks.

31. The method according to any of claims 28-30, comprising establishing a user demand related to the food item or to the cutting task, and making the allocation based on the user demand.

32. A module for a cutting apparatus according to any of the claims 1-27, the module comprising a multi-axis cutting device and a single-axis cutting device and being combinable with other modules containing multi-axis and / or single-axis cutting devices and comprising a computer interface configured to receive cutting tasks allocated to the module by the computer system of the cutting apparatus.

Images