Device and method for loading or unloading sheet metal processing machine or wood working machine

Abstract

The invention relates to a device (26) for loading or unloading a sheet metal processing machine or a wood working machine (12), comprising: a sensor interface (38) for receiving a sensor signal containing information of a workpiece (16) in a loading area (24) of the sheet metal processing machine or the wood working machine; a representation unit (40) for creating a representation of the workpiece and the loading area based on the sensor signals; a user interface (42) for providing the representation to a machine operator (28) and for receiving user input from the machine operator, the user input containing position information of the workpiece in the representation; an evaluation unit (44) for determining a modified representation of the workpiece and of the loading area on the basis of the sensor signals and user input, the user interface being designed for providing the modified representation to a machine operator and for receiving further user input containing information on which a workpiece loading operation is to be performed; a planning unit (46) for determining control instructions based on the sensor signal and the further user input, enabling the loading robot (14) to perform a loading operation to be performed; and a control interface (48) for activating and controlling the loading robot to perform a loading operation based on the control instruction. The invention further relates to a system (10), an associated method and a computer program product.

Description

Technical Field

[0001] This invention relates to an apparatus and method for loading or unloading a sheet metal processing machine or a timber processing machine. The invention also relates to a system and related computer program products for loading or unloading a sheet metal processing machine or a timber processing machine. Background Technology

[0002] During the production or processing of sheet metal products, stress or bending often occurs. To reduce these adverse effects, leveling machines, especially roller levelers, are used to process such sheet metal workpieces. When cutting and stamping sheet metal parts, unwanted burrs are formed at the cut edges and the edges of notches and grooves. These burrs protrude from the top or bottom surface of the steel sheet. Especially for steel sheet parts several centimeters thick, considerable technical work may be required to remove these burrs, namely, rounding the edges and polishing the surface.

[0003] It is worth noting that these sheet metal processing machines, as well as machines used in the woodworking field, typically operate in a pass-through manner. Workpieces to be processed are delivered via pallets or other types of load-bearing racks, placed on a conveyor belt, and fed into the machine for processing. Once the processing is complete, the workpiece is placed back onto the load-bearing rack.

[0004] To improve the efficiency of such operations, automated methods have been developed, in which workpieces are grasped by robots and transferred, for example, from one processing station to the next within a material handling machine. Similarly, methods designed to automate loading and unloading also exist.

[0005] Against this backdrop, DE102018218828A1 discloses a method and system for modifying a manufacturing process in a machine shop and / or a virtual 3D model of the machine shop. The method is designed for use in a real machine shop having at least one machine, particularly programmable and specifically a robot, and / or is designed for use in a virtual 3D model of the machine shop. The method includes the steps of: acquiring machine shop signals while a manufacturing process is being performed in the machine shop; storing a given point in time along with associated machine data based on the acquired machine shop signals; determining the position, motion, and / or state data of at least one object in the virtual model at at least one specific point in the past based on the stored point in time and its associated machine data; and representing and / or virtually measuring the virtual model of the machine shop based on the determined position, motion, and / or state data.

[0006] However, the operations of placing workpieces onto the conveyor belt and picking them up after processing are still often performed manually, especially when the number of units processed per batch is relatively small. For many applications, the cost of setting up and / or teaching or calibrating such automation solutions outweighs the benefits. Summary of the Invention

[0007] Therefore, the object of this invention is to provide a method for efficiently loading and unloading sheet metal processing machines or woodworking machines. The aim is to achieve high throughput and low error rate while maintaining low teach-in requirements and necessary technical implementation requirements. The object of this invention is to provide a method that can be implemented in a cost-effective manner while significantly improving efficiency.

[0008] To achieve this objective, the invention relates in a first aspect to an apparatus for loading or unloading a sheet metal processing machine or a woodworking machine, comprising:

[0009] - Sensor interface for receiving sensor signals containing workpiece information in the loading area of ​​the sheet metal processing machine or woodworking machine;

[0010] - Representation unit, used to create a representation of the workpiece and loading area based on sensor signals;

[0011] - A user interface for providing the representation to a machine operator and for receiving user input from the machine operator, the user input containing the position information of the workpiece in the representation;

[0012] - An evaluation unit for determining a corrected representation of the workpiece and loading area based on sensor signals and user input. The user interface is designed to provide the corrected representation to the machine operator and receive further user input containing information about the workpiece loading operation to be performed.

[0013] - A planning unit, used to determine control commands based on sensor signals and further user input, enabling the loading robot to perform the loading operation to be performed; and

[0014] - Control interface, used to activate and control the loading robot to perform loading operations based on control commands.

[0015] On the other hand, the present invention relates to a system for loading or unloading sheet metal processing machines or woodworking machines, comprising:

[0016] - The apparatus as described above;

[0017] - Sensors for covering the loading area; and

[0018] - Loading robots, especially industrial robots with robotic arms, are used to perform loading operations based on control commands.

[0019] Other aspects of the invention relate to a method designed for use with respect to the aforementioned apparatus, and a computer program product having program code for performing the steps of the method when the program code is executed on a computer. The invention also relates to a storage device on which the computer program is stored, causing the methods described herein to be performed when the computer program is executed on a computer.

[0020] Preferred configurations of the invention will be described in the dependent claims. It should be understood that the above features, and those further described below, can be used not only in their respective specified combinations, but also in other combinations or individually, without departing from the scope of the invention. Specifically, apparatus, systems, methods, and computer program products can be designed to correspond to the configurations described in the dependent claims for apparatus and systems.

[0021] According to the present invention, an automation of the loading and / or unloading process is provided, taking into account operator intervention at two points. The current status of the loading and / or unloading areas (loading zones) is acquired via sensors. More precisely, workpieces (metal sheets or wooden pieces) in the loading zone are detected. These workpieces can, for example, be placed on pallets or other load-bearing racks. Multiple workpieces, particularly those of the same type, can be stacked together (separately). Advantageously, the workpieces are oriented in the same direction within the stack. The workpieces themselves are preferably planar or flat parts. For example, parts weighing between several hundred grams and 20 kilograms or more, and with an area of ​​several square centimeters to several square decimeters or more, can be processed. Notably, irregularly shaped metal sheets can also be processed as workpieces.

[0022] Based on sensor signals, the first step is to create a representation of the workpiece and the loading area. This representation is provided to the machine operator, who then selects the workpiece according to it. This selection is received into the device via user input. By selecting, the user indicates the workpiece's position in the representation. The user input then serves as the basis for automatically correcting the representation.

[0023] In the second step, the corrected representation is provided to the machine operator. Through further user input, the machine operator provides information about the workpiece loading operation to be performed. He or she can specify the gripping method of the workpiece and the position to which the workpiece needs to be moved to perform the loading operation. Therefore, the loading position can be specifically indicated when loading the machine; the unloading position of the workpiece can be specifically indicated when unloading. During loading, the position relative to the through feed width of the machine can be specifically indicated.

[0024] Based on this further user input, the loading operation to be performed can be planned, that is, the control commands for the loading robot can be determined, and then the loading robot can be activated and controlled to perform the loading operation. Specifically, an industrial robot including an appropriate manipulator and an appropriate gripping device can thus be activated and controlled.

[0025] Compared to existing methods of manually loading and / or unloading workpieces, the method according to the present invention improves efficiency. This allows for a high degree of automation in the loading process, thereby saving labor costs. This results in an efficiency advantage. The machine operator only needs to intervene when two user inputs are specified. The machine operator's intervention occurs upstream of the processing. This may result in a slightly longer preparation time. On the other hand, the loading operation can be performed automatically by a loading robot without any further intervention from the machine operator.

[0026] Compared to fully automated processes involving, for example, workpiece image acquisition, gripping position, and placement position, the method of this invention improves safety and efficiency. In particular, it reduces expenditures on monitoring the loading area (required sensors, etc.) and on automatically processing the acquired sensor data (using automated data processing solutions, etc.). Furthermore, no additional machine protection devices are required to prevent errors. Moreover, complex and demanding teaching and calibration operations are unnecessary. The robot acquires clear specifications regarding workpiece picking and placement without requiring extensive and detailed training on teaching and / or calibration operations. By considering two user inputs, loading operations can be automated, resulting in efficient processing, especially when batch numbers are relatively low and workpiece shape freedom is high.

[0027] In a preferred configuration, the sensor interface is designed to receive sensor signals associated with color (optional: monochrome) and depth images of the workpiece in the loading area. Specifically, the sensor signals can be received from one, two, or more cameras. These cameras are oriented towards the loading area. By using color and depth cameras, different information sources can be provided, and these signals can be advantageously combined in subsequent stages. Furthermore, this combined camera and depth camera system is technologically advanced and readily available at low cost. Additionally, universally applicable and highly sophisticated image processing algorithms can be used, thereby reducing the error rate. Ultimately, high efficiency is achieved at low cost.

[0028] In the preferred configuration, the evaluation unit is designed to create an image representation. Notably, when a signal is received from the camera via the sensor interface, this signal can be used to create an image representation and provide it to the machine operator. For this purpose, processing steps such as contrast and / or brightness adjustment can be provided. However, more complex processing operations, such as coordinate transformations, can also be performed. This representation is provided to the machine operator, enabling them to efficiently provide user input. A particular advantage of using an image representation is that even untrained machine operators, or those without prior training in this area, can intuitively understand the representation and provide appropriate user input.

[0029] In a preferred configuration, the user interface includes a display, particularly a touchscreen display. A touchscreen display ensures easy and efficient display of information and reception of user input. Therefore, operation is intuitive. Extensive training and / or lengthy familiarization processes for machine operators are avoided. Furthermore, the possibility of errors is reduced.

[0030] In a preferred configuration, the user interface is designed to receive further user input, including the position of the gripping point of the loading robot on the workpiece. Alternatively, the user interface is designed to receive further user input, including the placement position of the workpiece. Alternatively, the user interface is designed to receive the orientation of the workpiece during placement. The gripping point here should be understood as relating to the point, area, or multiple points that the workpiece contacts when gripped. Therefore, depending on the gripping system of the loading robot (one or more points of contact with the workpiece, magnetic grippers, etc.), the position of the gripping point should be specifically understood as the orientation of the gripping system relative to the workpiece. The placement position of the workpiece refers to the point where the workpiece is placed for further processing during a loading operation, or the point where the workpiece is placed for storage / further processing / further transport during an unloading operation. The orientation of the workpiece refers to its three-dimensional or two-dimensional position when placed. This further user input can be received, in particular, via touch on a touchscreen. Of course, other user input methods are also possible. The further user input can also include multiple corresponding information items. The result is high usability. Required process parameters can be specified in a simple manner.

[0031] In a preferred configuration, the evaluation unit for determining the surface plane of the workpiece is designed to be segmentation-based and user-input-based. For this purpose, the RANSAC algorithm (Random Sample Consensus Algorithm) can be specifically employed. Alternatively, the evaluation unit is designed to determine an orthogonal view. Image data from both the color camera and the depth camera can be segmented. Notably, segmentation based on color image data can be provided. In other words, effective identification of the workpiece can be achieved based on simple user specifications and / or user selections, such as the machine operator specifying a single point belonging to the workpiece. Based on this identification, an orthogonal view can be established. Specifically, an orthogonal view should be understood as a view in which all parts of the workpiece are displayed at the same scale, thus enabling a simple evaluation of the workpiece's relative size and orientation. More specifically, converting the perspective view pointing to the loading area provided by the sensors into an orthogonal view allows the machine operator to perform a more comprehensive and efficient evaluation of the workpiece and, based on this, specify the loading operation to be performed. Ultimately, efficient processing is achieved.

[0032] In the preferred configuration, the planning unit is designed to determine control commands, including the workpiece's gripping and placement coordinates. Specifically, 6D coordinates (position and orientation) can be determined. Then, path planning can be performed according to a predetermined standard method. Additionally, appropriate transformations can be made. Ultimately, this achieves high-efficiency machining performance.

[0033] In a preferred configuration, the planning unit is designed to determine control commands that include predefined path default data for the movement of the loading robot relative to the loading area. This path default data is preferably determined during a calibration process. Alternatively or additionally, the planning unit is designed to determine control commands that include predefined sensor position data for the position of sensors relative to the loading area. This sensor position data is preferably determined during a calibration process, preferably performed during startup. The inclusion of path default data can particularly correspond to the inclusion of data specifically related to the loading area. Specifically, a one-time operation can be performed during loading area setup to determine how the loading robot moves within the loading area, thereby defining which types of movement are feasible and / or advantageous. This default specification can be a one-time operation. According to the invention, this data can be considered during the setup / initialization process of a single loading operation. Therefore, it is not necessary to perform corresponding calibrations for each loading operation of multiple parts. For example, a one-time calibration performed during startup is sufficient to enable subsequent loading of various types of workpieces. The same applies to sensor position data. This sensor position data specifically describes the position and orientation of the sensors relative to the loading area. It's worth noting that the camera's orientation and position can be defined as sensor position data in a single operation and then taken into account in the planning job executed within the planning unit. This further improves efficiency, as the path default data and sensor position data only need to be determined once and can then be used in various different loading operations.

[0034] In the preferred configuration, the user interface is designed to receive user input including the workpiece thickness. Especially when workpieces are stacked or about to be stacked, inputting their thickness is a valuable additional piece of information. Given this thickness, appropriate activation and control of the loading robot can be enabled. For example, there is no need to obtain this thickness via additional sensors. Therefore, efficiency can be further improved, thereby reducing costs. Typically, parts with the same profile also have the same thickness. If this is not the case, their thickness can also be obtained through appropriate sensor placement.

[0035] In a preferred configuration, the user interface is designed to receive user input containing information about the position of the workpiece located on the top layer of a multi-layered workpiece in the representation. The planning unit is designed to identify the offset of a workpiece located on the second layer after a loading operation has been performed on a workpiece already placed on the first layer. The planning unit is designed to determine control instructions based on the offset. Therefore, stacked workpieces can also be processed, in particular. For example, workpieces can be supplied in a stacked form placed on a pallet. The machine operator selects the position of the workpiece placed on the top layer. Based on this instruction, once the offset is determined, the entire stack of workpieces can be processed automatically. Furthermore, it is particularly advantageous if the workpiece thickness can be obtained as additional information.

[0036] In a preferred configuration of the system according to the invention, the sensor comprises a color camera and / or a depth camera, facing the loading area of ​​the sheet metal processing machine or woodworking machine. Using such cameras enables efficient implementation at a relatively low cost. Furthermore, standard image processing methods can be employed, thereby achieving high efficiency.

[0037] In a preferred configuration of the system, a loading robot is designed to perform loading operations. Sensors are designed to cover the loading area. The system includes another loading robot for performing unloading operations. The system also includes another sensor for covering the unloading area and providing another sensor signal containing information about the workpiece in the unloading area. A planning unit is designed to determine control commands that enable the other loading robot to perform the unloading operation based on sensor signals, another sensor signal, and another user input. A control interface is designed to activate and control the other loading robot to perform the unloading operation based on control commands. In addition to the loading robot, another loading robot is used. Thus, the loading robot is used to load the sheet metal processing machine or woodworking machine, while the other loading robot is used for unloading. The other sensor covers the unloading area. According to the invention, the specific workpiece information acquired in relation to the loading process is also used for the unloading operation. For example, a mesh model of the workpiece can be created based on user input and other user input entered during the loading operation. This mesh model or any other information can then be provided to the other loading robot during machine unloading, or it can be used for activation and control, thereby further improving efficiency.

[0038] The sheet metal processing machine or woodworking machine should be understood herein as a machine for processing individual parts whose size allows them to be handled by an industrial robot. Specifically, it is for processing workpieces transported by pallets or similar load-bearing racks. It is noteworthy that workpieces with different contours and / or different shapes are being processed. Multiple identical workpieces are processed in one process step. However, in subsequent processes, workpieces with different shapes will be processed. The term "representation" should be understood as particularly, but not limited to, image representations and / or graphic representations. The term "industrial robot" or "loading robot" specifically refers to an industrial manipulator or a robot with a manipulator arm designed for handling workpieces. The loading area should be understood herein specifically as the area adjacent to and / or surrounding the sheet metal processing machine or woodworking machine. The loading area includes the floor space for accommodating the load-bearing rack and / or the area for placing workpieces. The loading area also includes the machine's feeding area and / or unloading area (machine feeding area and machine unloading area). Both the loading area and the unloading area can be collectively referred to as the loading area. However, in a narrower sense, the unloading area can be used as a synonym for the loading area.

[0039] The method according to the invention can also implement user input related to the workpiece orientation during grinding or leveling operations. Grinding operations typically result in a predefined grinding pattern along one direction. Therefore, the machine operator can specify a specific orientation of the grinding pattern when placing the workpiece. After leveling, the operator can select the orientation based on the direction of the bow or bulge in order to flatten it in a controlled manner.

[0040] The contours and gripping points input by the machine operator can be stored. This allows for the creation of a database containing workpieces, gripping points and orientations, as well as related machine settings (such as the abrasive used, abrasive feed, abrasive speed, alignment gap adjustment, conveyor speed, etc.). For repetitive tasks, this data can be retrieved, reducing setup time. For example, identified workpieces can be recommended to the machine operator in the form of images, and confirmed by the operator where applicable. Furthermore, settings associated with the load cell can be established. For repetitive load cells, this configuration can be retrieved and re-executed.

[0041] Multiple gripping systems can be used. These systems can be automatically changed via a magazine. Therefore, workpieces of different geometries can be safely gripped. The gripping system can be displayed in the user interface. If the gripping system contains multiple gripping points, the operator can be instructed on the position of each gripping point on the workpiece as needed. The operator can then position the gripping points on the workpiece, ensuring that they are (mostly) not located in open areas (i.e., holes) to guarantee safe gripping. If the workpiece cannot be safely gripped in this manner, the operator can select another gripping system more suitable for the workpiece.

[0042] If non-planar workpieces are involved, the surface of each workpiece can be recalculated as needed.

[0043] As the machine operator approaches the loading position, a visual image of the nearest workpiece in the stack can be displayed based on the known conveyor belt speed. Starting from the workpiece surface and stack height, the wear distribution along the machine width can be determined. This wear distribution can be indicated to the machine operator. Based on this distribution, the machine operator can specify a uniform distribution to achieve uniform wear of the abrasive in the grinder and / or uniform wear of the leveling rollers in the leveler. Optionally, an ideal loading position can be suggested to the machine operator.

[0044] In the preferred configuration, workpieces with the same profile are stacked together on a load-bearing frame. All possible profiles are identifiable in the captured images. There are no uncovered portions and / or no profiles that were initially covered. Once preparation is complete, the machine operator can leave the factory, and the workpieces can be processed automatically. Attached Figure Description

[0045] The invention will now be described and illustrated in more detail with reference to several selected embodiments and the accompanying drawings. In the drawings:

[0046] Figure 1 This is a schematic diagram of a system according to the present invention for loading or unloading a sheet metal processing machine or a woodworking machine;

[0047] Figure 2 This is a schematic diagram of the device according to the present invention;

[0048] Figure 3 This is a schematic diagram illustrating the data flow and interaction between a machine operator and the inventive system according to one embodiment;

[0049] Figure 4 This is a schematic diagram of the steps of an inventive method according to one embodiment;

[0050] Figure 5 A schematic diagram illustrating optional modifications to the steps of the method according to the present invention;

[0051] Figure 6 A schematic diagram illustrating an optional modification of the method according to the present invention;

[0052] Figure 7 This is a schematic diagram of the method according to the present invention. Detailed Implementation

[0053] Figure 1 A system 10 according to the invention for loading or unloading a sheet metal processing machine or timber processing machine 12 is schematically shown. The schematic diagram specifically shows a sheet metal processing machine, particularly a deburring machine. The sheet metal processing machine or timber processing machine 12 operates in pass-through mode. The machine is specifically equipped with a conveyor belt for transporting workpieces 16 from a loading area 13 to an unloading area 15. In the sheet metal processing machine or timber processing machine 12, processing is performed at one (or more) processing stations 17. System 10 includes a loading robot 14 designed to transport workpieces 16 from a load rack 18 to the feed area 20 of the sheet metal processing machine during loading operations. The workpieces 16 may be placed in a stacked state on the load rack 18 when transported to the machine. System 10 also includes a sensor 22, in the illustrated embodiment a camera sensor, designed to cover the loading area 24 (in the illustrated embodiment, particularly the loading area). Furthermore, system 10 includes a device 26 for interacting with the sensor 22 and the loading robot 14. In the illustrated embodiment, device 26 is an industrial computer used by machine operator 28.

[0054] The method according to the invention relates particularly to the partial automation of loading and / or unloading operations of sheet metal processing machines or timber processing machines. Its aim is, in particular, to achieve efficient and easily implemented partial automation by cleverly incorporating some user input.

[0055] The illustrated embodiment demonstrates an (optional) embodiment of system 10, in which another loading robot 30 is provided. This additional loading robot 30 removes the processed workpiece 16 from the conveyor belt of the sheet metal processing machine or woodworking machine 12. The workpiece is then placed in another load rack 32. The unloading area 15 is covered by another sensor 36. The generated sensor signals contain information about the workpiece 16 in the unloading area 15. With this optional supplement, data acquired during loading can be reused for the unloading operation.

[0056] Figure 2 A schematic diagram illustrates a device 26 according to the invention for loading or unloading a sheet metal processing machine or a woodworking machine. Device 26 includes a sensor interface 38, a display unit 40, a user interface 42, an evaluation unit 44, a planning unit 46, and a control interface 48. Each unit can be implemented partially or entirely in software and / or hardware. Specifically, these units can be designed as processors, processor modules, or software for processors. For example, device 26 can be implemented as an industrial computer. The device and / or the methods performed using the device can also be designed as software for a sheet metal processing machine system. For example, the functionality of device 26 can be implemented in software that executes on the sheet metal processing machine system. However, device 26 can also be designed as an add-on device, connected wired or wirelessly to the sheet metal processing machine system, a loading robot, and corresponding sensors.

[0057] Sensor interface 38 is connected to a sensor, which may specifically include a color camera and a depth camera. Therefore, color and depth images of the workpiece within the loading area and the entire loading area can be received.

[0058] The representation unit 40 is used to create a representation of the workpiece and loading area based on sensor signals. Based on the sensor signals, a representation, particularly an image, preferably a real-time image, is created for evaluation by the machine operator. The representation unit 40 can either simply transmit the received image or perform image processing operations.

[0059] User interface 42 is designed to allow interaction with a machine operator. User interface 42 may, in particular, include a touchscreen display. Other types of displays may also be used. User interface 42 enables interaction with the machine operator. The machine operator may, in particular, input data and read and / or receive data output. The representation may, in particular, be displayed. The displayed representation may, in particular, be an image representation. Based on the image representation, the machine operator may then input and / or select the position of a workpiece in the representation. The machine operator may, in particular, select a given point on the workpiece. This operation can be accomplished via touch input (clicking on an object). Specifically, the machine operator selects one workpiece from each stack of workpieces (the topmost workpiece), or more precisely, selects a given point for each topmost workpiece. This avoids technically demanding image data processing operations and automatic workpiece identification. This greatly simplifies the process, as otherwise cumbersome teaching operations and / or evaluation processes would be required, especially for workpieces with complex shapes.

[0060] Evaluation unit 44 determines a corrected representation of the workpiece and loading area based on user input provided by the machine operator. For this purpose, both user input and sensor signals are used / evaluated. Notably, this allows for the determination of the workpiece's surface planes. This can be achieved, in particular, by using segmentation. The RANSAC algorithm can be employed for this. That is, based on input defining a point on the workpiece, the entire contour of the workpiece can be determined. Additionally, an orthogonal view of the workpiece can be created. This view should be understood, in particular, as a scaled view that has eliminated all perspective distortion. This method is especially advantageous, for example, if images are acquired using a tilted-mounted camera, and the machine operator selects the workpiece on that image for subsequent further processing.

[0061] Once the corrected representation is determined, it is fed back to the machine operator via user interface 42. With updated user input, the operator provides information about the workpiece loading operation to be performed. Specifically, further user input includes defining the gripping point position of the loading robot on the workpiece. Depending on the type of workpiece, gripping points or engagement points can be defined to ensure the workpiece can be effectively gripped and secured to the loading robot's manipulator arm. Combined with the machine operator's data input, their experience can be relied upon. Specifically, the machine operator manually defines the workpiece gripping method, thus avoiding more cumbersome data processing steps.

[0062] Additionally, the further user input can include the placement position of the workpiece. It is worth noting that the placement position can be specified within a placement area, particularly on the conveyor belt of a woodworking machine or sheet metal processing machine. Therefore, for example, by processing a stored image of the loading area, the machine operator can specify a suitable placement position through graphical visualization. In this way, the machine operator can preset the placement position. Clearly, further automation can be achieved, for example, by sequentially presetting different / variable placement positions to fully utilize the entire width of the conveyor belt. Furthermore, it is even more advantageous if the machine operator specifies the orientation of the workpiece when placing it. The key point here is particularly the orientation in which the workpiece is fed into the machine. In all cases, the data input by the machine operator when performing the further user input can replace relatively more complex data processing steps, thereby improving efficiency.

[0063] Planning unit 46 is used to determine the control commands for the loading robot. These commands are based on further user input and sensor signals. This data forms the basis for determining how the loading robot places the workpiece into the sheet metal processing machine or woodworking machine 12. Notably, appropriate gripping and placement coordinates can be determined. Appropriate transformations are required for this. Planning can optionally be based on the use of path default data. Such path default data is used to specify default values ​​relative to the loading area. These default values ​​can be, for example, waypoints on the loading robot's path. Additionally, the direction of travel and travel speed can be preset. This path default data can, for example, be stored in memory when the system according to the invention is put into operation and can reflect the characteristics of the current operating location and / or robot unit. Furthermore, sensor position data can be considered accordingly. This data reflects the position of the sensors, i.e., the position and orientation of the sensors relative to the loading area. Moreover, the sensor position data is specific to the application case and / or application field and only needs to be stored once, for example, when the device and / or system of the invention are put into operation at a given operating site. In this case, the path default data and sensor position data can be determined in a single calibration operation.

[0064] For example, the machine operator selects the topmost workpiece in the stack. Once gripped, the selected part will no longer appear in the image. The captured / determined geometry (profile) and gripping position of that part can then be used specifically for two purposes:

[0065] - After processing, identify and grasp the part for unloading; and

[0066] - Identify and grab subsequent parts in the stack.

[0067] If the load cell comprises multiple identical parts, a selected part or its outline can be searched in the image, and it can be identified in other stacks in this way. The indicated gripping location can then be transferred to other stacks.

[0068] Additionally, the user interface 42 can also be designed to receive the thickness of the workpiece. This is particularly advantageous when workpieces to be processed are supplied in a stacked manner. Having the machine operator specify the workpiece thickness further simplifies the operation. This is especially beneficial because the loading robot can travel to the appropriate gripping position at a relatively high speed without the risk of collision. It is worth noting that the user interface 42 can receive position information for multiple layers of the same stacked workpieces.

[0069] Control interface 48 is used to activate and control the loading robot. In particular, it can thus activate and control industrial robots with manipulators.

[0070] The method according to the invention aims to automate, or more precisely, partially automate, loading and unloading operations in sheet metal processing machines and woodworking machines. Specifically, the system according to the invention is used to transfer sheet metal parts from the load rack in the machine's feed area onto the conveyor belt of the processing machine. Additionally, the sheet metal parts can be removed from the conveyor belt again and placed back onto the load rack in the machine's discharge area.

[0071] Essentially, the process of loading workpieces onto a machine, or more precisely, onto a handling system connected to the machine, requires information about the location of the workpiece to be loaded and the points that the loading system and / or loading robot need to navigate to and / or grip. These gripping coordinates can be retrieved (predefined layers on the pallet: "configuration"), determined (AI camera systems: generating gripping point determination on unknown parts, or identifying parts and retrieving stored gripping points on parts), or passed on during the learning process (teaching: defining the points to navigate to by activating and controlling the robot to perform appropriate actions). Placement points can be established accordingly through coordinate generation, predefined coordinates, or teaching steps. These gripping points can then be used to plan the loading robot's path.

[0072] The method according to the invention simplifies processes, reduces error sensitivity, and enhances robustness, especially under varying auxiliary conditions (e.g., lighting conditions). Teaching operations can also lead to considerable costs. The method according to the invention involves a machine operator at both stages of the process and enhances robustness while significantly reducing costs by taking their input into account. Its core proposal is to combine or connect AI models with the intelligence of the machine operator to reduce costs. Partial automation can be achieved through a certain degree of user input from the machine operator.

[0073] Figure 3This is a schematic diagram of the method according to the invention. Specifically, it shows a machine operator 28 interacting with the device 26 according to the invention via a user interface 42. The device receives data from sensor 22 via sensor interface 38. Data items D1…Dn received via user interface 42 and sensor interface 38 may relate to / include, for example, gripping points, the number of workpieces in a stack, part orientation, machine start / stop, the necessity of part flipping, the necessity of providing temporary storage space, bimetallic strip identification, placement point, metal plate thickness, and other information. Sensor 22 may in particular include a combination of color images and a depth camera. Additionally, the sensor may also include other data sources S1…Sn, such as bimetallic strip sensors, metal plate thickness sensors, force sensors, gripping sequence sensors, and other sensors. Evaluation unit 44 is used to process these different types of data. In planning unit 46, the loading operation is being planned. In the illustrated embodiment, evaluation unit 44 and planning unit 46 are implemented together. Planning unit 46 is connected to loading robot 14 via control interface 48. Other systems / actuators A1…An may also be connected. For example, activation and control of the conveying system, flipping unit, thin liner remover, gripper hopper, bimetallic strip separator, or other units can also be provided. The sheet metal processing machine or woodworking machine 12 can also be connected to the evaluation unit and / or planning unit via (optional) machine interface 50.

[0074] According to a preferred configuration of the method of the present invention, a color and depth combined camera is used to capture color and depth images of a tray and the workpiece placed on it. Subsequently, in the first step, the machine operator defines the position in the image and / or the location of the workpiece via user input. This input can then be used to segment the color image and generate a mask. An image mask containing only the workpiece is created. This image mask can then be applied to the depth and color images. The generated depth image fragments can be used to extract the surface planes of the component. Since the sensor signal of the depth camera typically contains relatively much noise, applying the RANSAC algorithm is highly advantageous. The color image fragments are combined with the planar data to eliminate perspective distortion, thereby creating an orthographic view of the component. The gripper point can be defined before or after the creation of the orthographic view. Specifically, this can be accomplished through further user input from the machine operator. Visualizing the gripper in the orthographic view of the workpiece facilitates operation by the machine operator, enabling even inexperienced operators to effectively use the method according to the present invention. The orthographic view can define the orientation of the component to be placed and the placement point. Due to the absence of perspective distortion, the orthographic view can effectively predict potential collisions with the workplace environment or other workpieces.

[0075] Figure 4 A flowchart illustrating the steps of the method of the present invention in one configuration is provided.

[0076] Specifically, assuming the workpieces are supplied to the load rack with appropriate stacking accuracy, this method is sufficient to achieve full automation of the loading process.

[0077] On the other hand, assuming the workpiece is supplied with insufficient stacking accuracy, the workpiece offset can be identified by searching for predefined color image segments in the orthogonal view of the currently acquired image, or based on updated sensor signals. Then, new gripping points can be determined using the relative positions of the gripping points and their orientation on the component. Similarly, a mesh file for the component can be created using the orthogonal view. The corresponding process is as follows: Figure 5 As shown.

[0078] The mesh model used to search for workpieces in newly acquired images (updated sensor signals) can also be used during unloading when another loading robot is deployed in the unloading area. In some cases, a 2D model of the workpiece may suffice when dealing with relatively thin sheet metal parts, or when the camera is positioned vertically above the pallet and / or conveyor belt.

[0079] Figure 6 This indicates another optional supplement. By attaching fixed position markers relative to the loading robot, the camera can be placed in any location, such as... Figure 6 As shown.

[0080] Figure 7 A method for loading or unloading a sheet metal processing machine or a woodworking machine according to the present invention is illustrated schematically. The method includes the following steps: receiving a sensor signal (S10); creating a representation of the workpiece (S12); providing the representation (S14); receiving user input (S16); determining a corrected representation (S18); providing the corrected representation (S20); receiving further user input (S22); determining control commands (S24); and activating and controlling the loading robot (S26). This process can be implemented, for example, as software executed on the machine system of the sheet metal processing machine.

[0081] The invention has been fully described and illustrated with reference to the accompanying drawings and specification. These descriptions and illustrations should be understood as exemplary rather than restrictive. The invention is not limited to the disclosed embodiments. Other embodiments or variations will become apparent to those skilled in the art upon using the invention and thoroughly analyzing the drawings, this disclosure, and the appended claims.

[0082] In the claims, words such as "comprising" and "having / including" do not exclude the presence of other elements or steps. The indefinite articles "a" or "an" used with these words do not exclude the presence of multiple said items. A single element or unit can perform the function of multiple units mentioned in the patent claims. Elements, units, devices, and systems can be implemented partially or wholly in hardware and / or software. The mere mention of certain measures in multiple different dependent patent claims should not be construed as meaning that a combination of these measures cannot be used advantageously. Computer programs can be stored / distributed on non-volatile data carriers, such as optical storage or solid-state drives (SSDs). Computer programs can be distributed, for example, via the Internet, wired or wireless communication systems, along with the hardware, or distributed as part of the hardware. Reference numerals in the patent claims should not be construed as restrictive.

Claims

1. A device (26) for loading or unloading a sheet metal processing machine or a woodworking machine (12), comprising: A sensor interface (38) is used to receive sensor signals containing information about the workpiece (16) present in the loading area (24) of the sheet metal processing machine or the woodworking machine; The representation unit (40) is used to create a representation of the workpiece and the loading area based on the sensor signals; User interface (42) is used to provide the representation to machine operator (28) and to receive user input from the machine operator, the user input containing the position information of the workpiece in the representation; Evaluation unit (44) is used to determine a corrected representation of the workpiece and the loading area based on the sensor signal and the user input. The user interface is designed to provide the corrected representation to the machine operator and receive further user input, which includes information about the workpiece loading operation to be performed. Planning unit (46) is used to determine control instructions based on the sensor signals and the further user input, the control instructions enabling the loading robot (14) to perform the loading operation to be performed; as well as The control interface (48) is used to activate and control the loading robot to perform the loading operation based on the control command.

2. The apparatus (26) according to claim 1, wherein the sensor interface (38) is designed to receive sensor signals associated with color and depth images of the workpiece (16) present in the loading area (24).

3. The apparatus (26) according to any of the preceding claims, wherein the representation unit (40) is designed for creating an image representation.

4. The apparatus (26) according to any of the preceding claims, wherein the user interface (42) includes a display, particularly a touch screen display.

5. The apparatus (26) according to any of the preceding claims, wherein the user interface is designed to receive the further user input, the further user input including: The position of the gripping point of the loading robot (14) on the workpiece (16); The placement position of the workpiece; And / or The orientation of the workpiece when it is placed.

6. The apparatus (26) according to any of the preceding claims, wherein the evaluation unit (44) is designed to: The surface planes of the workpiece (16) are determined based on segmentation and user input, particularly by employing the RANSAC algorithm; and / or Determine the orthographic view.

7. The apparatus (26) according to any of the preceding claims, wherein the planning unit (46) is designed to determine control instructions, the control instructions including the gripping coordinates and placement coordinates of the workpiece (16).

8. The apparatus (26) according to any of the preceding claims, wherein the planning unit (46) is designed to determine the control command in such a way as: Based on predefined path default data of the motion of the loading robot (14) relative to the loading area (24), the path default data is preferably determined during the calibration process; and / or Based on predefined sensor position data of the sensor's position relative to the loading area, the sensor position data is preferably determined during the calibration process.

9. The apparatus (26) according to any of the preceding claims, wherein the user interface (42) is designed to receive user input including the thickness of the workpiece (16).

10. The apparatus (26) according to any of the preceding claims, wherein The user interface (42) is designed to receive user input, which includes position information of the uppermost workpiece (16) in the multi-layered workpiece in the representation; The planning unit (46) is designed to identify the offset of the workpiece located in the second layer after the loading operation of the workpiece placed in the first layer; as well as The planning unit is designed to determine the control command based on the offset.

11. A system (10) for loading or unloading a sheet metal processing machine or a woodworking machine (12), comprising: The apparatus (26) according to any of the preceding claims; Sensors are used to cover the loading area (24). as well as A loading robot (14), particularly an industrial robot with a manipulator arm, is used to perform the loading operation based on the control commands.

12. The system (10) of claim 11, wherein the sensor comprises a color camera and / or a depth camera facing the loading area (24) of the sheet metal processing machine or the woodworking machine (12).

13. The system (10) of claim 11 or 12, wherein the loading robot is designed to perform loading operations, and the sensors are designed to cover the loading area, including: Another loading robot (30) is used to perform the unloading operation; as well as Another sensor (36) is used to cover the unloading area and to provide another sensor signal including information about the workpiece (16) present in the unloading area. The planning unit (46) is designed to determine, based on the sensor signal, the other sensor signal, and the further user input, control instructions to cause the other loading robot to perform an unloading operation; and The control interface (48) is designed to activate and control the other loading robot to perform the unloading operation based on the control command.

14. A method for loading or unloading a sheet metal processing machine or a woodworking machine (12), comprising the following steps: Receive (S10) a sensor signal containing information about the workpiece (16) present in the loading area (24) of the sheet metal processing machine or wood processing machine; (S12) A representation of the workpiece and the loading area is created based on the sensor signals; The representation is provided (S14) to the machine operator (28); Receive (S16) user input from the machine operator, the user input containing the position information of the workpiece in the representation; Based on the sensor signals and the user input, a corrected representation of the workpiece and the loading area is created (S18); The corrected representation is provided (S20) to the machine operator; Receive (S22) further user input containing information about the loading operation to be performed on the workpiece; Based on the sensor signals and the further user input, a control command is determined (S24), which enables the loading robot (14) to perform the loading operation to be performed; and Based on the control command, the loading robot is activated and controlled (S26) to perform the loading operation.

15. A computer program product comprising program code, which, when executed on a computer, performs the steps of the method according to claim 14.

Images