Method for maintaining a minimum distance between a robotic manipulator and a contact-critical object

Abstract

The invention relates to a method for controlling a robotic system for maintaining a minimum distance between a robotic manipulator and a contact-critical object, wherein the contact-critical object and the manipulator are located on the basis of sensor information of the robotic system, and a virtual model is created which has at least one virtual contact-critical object and a virtual manipulator. The minimum distance is maintained by enlarging the virtual contact-critical object and / or the virtual manipulator; detecting when the enlarged or non-enlarged virtual contact-critical object and the enlarged or non-enlarged virtual manipulator contact each other; and stopping or slowing down the movement of the manipulator when the enlarged or non-enlarged virtual contact-critical object is contacted by the enlarged or non-enlarged virtual manipulator. The aim of the invention is to maintain a minimum distance between the object (4) and the manipulator (2a). This is achieved by enlarging the virtual object (6) in order to form an enlarged virtual object (9) and / or enlarging the virtual manipulator (7a) in order to form an enlarged virtual manipulator (8a) and determining a contact point between one of the aforementioned virtual components.

Description

[0001] Method for maintaining a minimum distance between a robotic manipulator and a contact-critical object

[0002] The invention relates to a method for controlling a robotic system for maintaining a minimum distance between a robotic manipulator and a touch-critical object.

[0003] It is known from the state of the art that robotic systems used in, for example, orbital robotics, telesurgery, nursing, or telenavigation, which utilize traditional bilateral telemanipulation approaches, often compromise the accuracy and performance of telemanipulation tasks due to technical and physical limitations. To counteract this problem, the model-based haptic telemanipulation (MATM) method is known from the state of the art. It relies on two types of virtual models: a remote virtual model that enables the shared autonomous function of the teleoperated robot, and a local virtual model that aims to provide assistive haptic feedback to the human operator.

[0004] The object of the invention is to provide a safe method which enables the control of a robotic system with a robotic manipulator and prevents unwanted collisions or contact between manipulators and touch-critical objects.

[0005] Touch-critical objects with which a manipulator should not collide are objects that the user does not actively want to manipulate, or that require particularly careful manipulation. In the case of telesurgery, these objects could be, for example, vital organs, blood vessels, or tissues that should not be operated on.

[0006] The object is achieved according to the invention by claims 1, 12 and 13.

[0007] The inventive method for controlling a robotic system to maintain a minimum distance between a robotic manipulator and a contact-critical object relates to a robotic system having a manipulator. This can be, for example, an industrial robot, such as those used in assembly line production, or a surgical robot, such as those used in telemedicine, or a humanoid robot, or a robot such as those used in space or the ISS. In principle, any robotic system designed to accommodate a manipulator is suitable for the inventive method.

[0008] A manipulator according to the invention can be a variety of attachments for the robotic system. A manipulator can be:

[0009] A gripper, a mechanical hand, a cutting tool (scalpel, saw, knife, etc.), a welding device, a rotary tool (drill, milling machine, angle grinder), a geometric shape (rod, hook, cone, cuboid, etc.), a universal gripper holder.

[0010] The robotic system according to the invention further comprises one or more sensors capable of determining the position of the manipulator and / or a touch-critical object in space. Suitable sensors for this purpose include, for example, cameras, lidar sensors, radar sensors, MRI or X-ray sensors, time-of-flight sensors, as well as ultrasonic or infrared sensors. Angle sensors or position sensors on or in the joints of the robotic system are also suitable for determining the position of the robotic manipulator.

[0011] Based on the sensor information, the robotic system creates a virtual model comprising at least one virtual touch-critical object and one virtual manipulator. It is further preferred that the virtual touch-critical object and the virtual manipulator correspond to the touch-critical object and the manipulator between which the minimum distance is to be maintained.

[0012] The robotic system preferably further comprises a processor. Using the sensor data, a processor is able to determine the position of one or more objects and one or more manipulators and, based on this, create a virtual model with at least one virtual manipulator and at least one virtual touch-critical object. Such virtual models can be point clouds, 3D models, voxel maps, or sphere trees. These models can preferably be formed both classically and with probabilities, confidence values, or probability distributions. The method according to the invention maintains a minimum distance between the touch-critical object and the manipulator by enlarging the virtual touch-critical object and / or the virtual manipulator.A preferred magnification can, for example, be an isotropic, static magnification, preferably by a scalar factor, so that the virtual manipulator or the virtual touch-critical object is magnified uniformly in all directions by the scalar factor. This factor could preferably be set statically by the user, meaning it preferably does not change over time.

[0013] The method according to the invention further detects when a magnified or unmagnified virtual touch-critical object and a magnified or unmagnified virtual manipulator touch each other. A preferred contact can be a point, a curve, a surface, or a volume in which the virtual manipulator and the virtual touch-critical object touch and / or intersect. It is the region where two objects, or a virtual touch-critical object and a virtual manipulator, are in contact. A magnification can also be a reduction. It is preferred that a decreasing magnification is accompanied by an actual magnification in another spatial direction. Unmagnified means that the virtual touch-critical object or the virtual manipulator remains unchanged in size.

[0014] The method according to the invention stops or slows down the movement of the manipulator when touching the enlarged or unenlarged virtual touch-critical object and the enlarged or unenlarged virtual manipulator.

[0015] It is preferred that only the virtual touch-critical object, only the virtual manipulator, or both the virtual manipulator and the virtual touch-critical object are enlarged.

[0016] Preferably, the virtual touch-critical object and / or the virtual manipulator are magnified. Magnification can be achieved by scaling the individual spatial axes, by projection, or by any matrix multiplication.

[0017] Preferably, a minimum distance between the touch-critical object and the manipulator is determined based on a contact point between the virtual touch-critical object and / or virtual manipulator or enlarged virtual touch-critical object and / or virtual manipulator or virtual touch-critical object and / or enlarged virtual manipulator or enlarged virtual touch-critical object and / or enlarged virtual manipulator.

[0018] Preferably, the minimum distance according to the invention is thus determined based on the contact point, by defining that the virtual touch-critical object and / or virtual manipulator or enlarged virtual touch-critical object and / or virtual manipulator or virtual touch-critical object and / or enlarged virtual manipulator or enlarged virtual touch-critical object and / or enlarged virtual manipulator must not touch, or must not overlap, or must not overlap by more than a manually set value. If one of the aforementioned criteria is violated, a critical touch may occur.

[0019] Preferably, the magnification of a virtual touch-critical object and / or the virtual manipulator is different for different spatial directions. This has the advantage that the minimum distance from an object on one side, which should not be touched, can be selected to be greater than on another side, where contact should preferably occur. This preferably increases the maneuverability of the manipulator.

[0020] It is further preferred that the virtual model comprises, in addition to the virtual touch-critical object, a virtual object to be manipulated. It is further preferred that the virtual object to be manipulated corresponds to an object to be manipulated with which the manipulator is intended to interact.

[0021] It is preferred that the magnification be realized by scaling the virtual touch-critical object and / or the virtual manipulator by multiplying it by a magnification factor, magnification vector, or a magnification matrix. A magnification factor is, for example, a scalar that enables uniform scaling in all dimensions of the virtual touch-critical object and / or the virtual manipulator.

[0022] A magnification vector is a three-dimensional vector in which the three components represent the scaling factors in the corresponding spatial directions. A scaling factor greater than 1 in one dimension results in an enlargement along that direction, while a factor less than 1 results in a reduction. Advantageously, the magnification vector can also be applied in a rotated reference coordinate system, allowing a touch-critical object, for example, to be magnified more in any spatial direction.

[0023] A magnification matrix is ​​a mathematical matrix used to describe enlargements or reductions of three-dimensional virtual objects in a coordinate-based system. Typically, it is a 3x3 or 4x4 matrix, where the diagonal elements represent the scaling factors for the respective spatial directions. A scaling factor greater than 1 results in enlargement along the corresponding axis, while a factor less than 1 results in a reduction. The magnification matrix allows the size and shape of virtual objects to be modified. By adding translation components, it can also be used to center the magnification around a specific point and thus perform simultaneous translations. The magnification matrix also allows for shearing by adjusting the off-diagonal elements of the matrix for the corresponding axes.

[0024] It is preferred that the magnification be realized by scaling the virtual touch-critical object and / or the virtual manipulator using safety layers. A safety layer, or distance layer in virtual environments, enables the introduction of a safety distance between different virtual objects. The distance layer corresponds to an enlargement of the virtual objects, whereby a shell of constant thickness is placed around the object. The enlargement achieved by this layer does not correspond to classic scaling, but rather to the addition of a protective shell around the virtual object. This enlargement can be implemented using a distance field, for example, stored in a voxel map.

[0025] It is preferred that the enlargement be realized by scaling the virtual touch-critical object and / or the virtual manipulator through an extrusion. A preferred extrusion describes the enlargement of a volume by translating or rotating a 2D object. It is further preferred that an extrusion describes a translating or rotating a 3D object.

[0026] It is preferred that the magnification factor for magnifying the virtual touch-critical object be different for different spatial axes. The spatial axes do not necessarily have to coincide with the axes of a world coordinate system, but can also be the axes of a rotated coordinate system. The different magnification in different spatial axes can result, for example, from a different nature of the object along different axes. For example, one side of an organ, which has many blood vessels, could benefit from a greater minimum distance than its back, which is protected, for example, by a layer of fat. Dynamically adjusting the magnification factor results in maximum mobility for the manipulator, with minimal risk of a critical touch.

[0027] It is further preferred that the magnification factor for magnifying the virtual touch-critical object depends on the direction of movement and / or movement speed of the manipulator. The different magnification depending on the manipulator's direction of movement and / or movement speed can be advantageous, for example, when the manipulator is moved toward a touch-critical object. For example, the magnification of the virtual touch-critical object and / or virtual manipulator can be increased along the direction of movement, while the magnification along other axes remains constant. The magnification can also be adjusted depending on the speed and / or rotation speed of the manipulator in or around one or more spatial axes. For example, high manipulator speeds could lead to increased magnification, for example to take the manipulator's inertia into account.

[0028] It is also preferred that the magnification factor for magnifying the virtual touch-critical object depends on the direction of movement and / or speed of the touch-critical object and / or on the relative speed between the touch-critical object and the manipulator. The different magnification depending on the direction of movement and / or speed of the touch-critical object can be advantageous, for example, when the touch-critical object is moving, such as a heart. For example, the magnification of the virtual touch-critical object and / or virtual manipulator can be increased along the contraction direction of the heart, while the magnification along other axes remains constant. The magnification can also be adjusted depending on the speed and / or rotational speed of the touch-critical object in or around one or more spatial axes.For example, high touch-critical object speeds and / or contraction speeds could lead to increased magnification, for example to take into account the inertia of the touch-critical object.

[0029] It is also preferred that the magnification factor for magnifying the virtual touch-critical object depends on the accuracy of the sensors used to create the virtual touch-critical object and the virtual manipulator in the robotic system. The magnification factor for magnifying the virtual touch-critical object depends on the accuracy of the sensors. This accuracy can be affected, for example, by: the design and quality of the sensor, lighting, fog, smoke, vibrations, obscuring the touch-critical object and / or manipulator by, for example, parts of the robotic system, or other external components or people. A change in the accuracy of the sensor data can thus be used to change the magnification factor of the virtual touch-critical object.

[0030] It is also preferred that the magnification factor for magnifying the virtual touch-critical object is combined with a translational or rotational displacement. A magnification factor of 1 can preferably be used here. A magnification factor of 1 does not result in any effective magnification, so that purely rotational or translational displacements are also possible to ensure the minimum distance. This can be advantageous if the position of the virtual touch-critical object in question is not plausible. This can occur, for example, when penetrating rigid objects. By shifting with simultaneous magnification with a magnification factor greater than 1, the safety distance on one side of a virtual touch-critical object / touch-critical object can be set larger than on its opposite side.This can be advantageous, for example, if one side of a touch-sensitive object is sensitive or dangerous and should be avoided with particular care. It can also be advantageous if movement of the touch-sensitive object is expected or possible.

[0031] An error of the virtual model or touch-critical object calculated by the robotic system or a processor of the robotic system can also influence the magnification factor of the virtual touch-critical object. If the calculated error, for example, the statistical error, changes, the magnification factor of the virtual touch-critical object changes. For example, a larger error leads to a larger magnification factor.

[0032] It is also preferred that the magnification factor of the virtual touch-critical object be the same in all directions. This can be advantageous for using the inventive method on robotic systems with limited computing power.

[0033] It is preferred that the magnification factor for magnifying the virtual manipulator be different for different spatial axes. The different magnification in different spatial axes can result, for example, from a different nature of the manipulator along different axes. For example, one side of a manipulator could have a blade, which justifies a larger magnification factor in the direction in which the blade can cut. The back of the manipulator, on the other hand, could be blunt, which would justify a smaller magnification factor orthogonal to the back of the manipulator. Dynamically adjusting the magnification factor results in maximum mobility for the manipulator, with minimal risk of critical contact.

[0034] It is also preferred that the magnification factor for magnifying the virtual manipulator depends on the direction of movement and / or speed of the manipulator and / or on the relative speed between the touch-critical object and the manipulator. The different magnification depending on the manipulator's direction of movement and / or speed can be advantageous, for example, when the manipulator is moved toward a touch-critical object. For example, the magnification of the virtual touch-critical object and / or virtual manipulator can be increased along the direction of movement, while the magnification along other axes remains constant. The magnification can also be adjusted depending on the speed and / or rotational speed of the manipulator in or around one or more spatial axes.For example, high manipulator speeds could lead to increased magnification, for example to take into account the inertia of the touch-critical object.

[0035] It is also preferred that the magnification factor for magnifying the virtual manipulator depends on the direction of movement and / or speed of the touch-critical object. The different magnification depending on the direction of movement and / or speed of the touch-critical object can be advantageous, for example, when the touch-critical object is moving, such as a heart. For example, the magnification of the virtual touch-critical object and / or virtual manipulator can be increased along the contraction direction of the heart, while the magnification along other axes remains constant. The magnification can also be adjusted depending on the speed and / or rotation speed of the touch-critical object in or around one or more spatial axes.For example, high touch-critical object speeds and / or contraction speeds could lead to increased magnification, for example to take into account the inertia of the touch-critical object.

[0036] It is also preferred that the magnification factor for magnifying the virtual manipulator depends on the accuracy of the sensors used to create the virtual touch-critical object and the virtual manipulator in the robotic system. The magnification factor for magnifying the virtual manipulator depends on the accuracy of the sensors. This accuracy can be affected, for example, by: the design and quality of the sensor, lighting, fog, smoke, vibrations, obscuring the touch-critical object and / or manipulator by, for example, parts of the robotic system, an object, other external components, or people. A change in the accuracy of the sensor data can thus be used to change the magnification factor of the virtual manipulator.

[0037] It is also preferred that the magnification factor for magnifying the virtual manipulator be combined with a translational or rotational displacement. A magnification factor of 1 can preferably be used here. A magnification factor of 1 does not result in any effective magnification, so purely rotational or translational displacements are also possible to ensure the minimum distance. This can be advantageous if the position of the virtual manipulator in question is implausible. This can occur, for example, when penetrating rigid objects.

[0038] The error of the virtual model or virtual manipulator calculated by a processor can also influence the magnification factor of the virtual manipulator. If the calculated error, for example, a statistical error, becomes larger, the magnification factor of the virtual manipulator changes. For example, a larger error leads to a larger magnification factor. It is also preferred that the magnification factor of the virtual manipulator be the same in all directions. This can be advantageous for using the inventive method on robotic systems with limited computing power.

[0039] It is preferred that the robotic system has at least one user terminal with at least one input device capable of receiving inputs for controlling the robotic system from at least one user. One embodiment of the user terminal can be a mobile device, such as a smartphone or tablet.

[0040] Preferably, a user terminal is designed as a computer with a screen, keyboard, mouse, gamepad or joystick.

[0041] Furthermore, a user terminal can preferably be designed as an exoskeleton or robotic arm, which replicates the user's movements and, with the aid of the exoskeleton or a robotic arm and actuators, is capable of providing the user with haptic or physical feedback. Preferably, the exoskeleton limbs or the limbs of the robotic arm cannot be moved closer to a touch-critical object than the specified minimum distance in the method according to the invention.

[0042] Preferably, the restoring forces of the actuators of the exoskeleton or the robot arm would therefore not allow the user to move a manipulator closer to a touch-critical object than the minimum distance determined according to the invention, since the robot arm or the exoskeleton provides the user with force and / or torque feedback.

[0043] It is preferred that the at least one input device is designed to provide the user with feedback as to whether the minimum distance has been reached or exceeded. This feedback can be provided either visually, by means of displays on a screen, or by the lighting up of a display or lighting device. It is further preferred that this feedback is audible, for example by an announcement from a digital assistant, or a warning tone, or both. It is further preferred that the feedback is haptic. For example, an input device could vibrate as soon as the minimum distance is reached or exceeded. It is further preferred that, in the event that the input device is an exoskeleton or a robotic arm, the feedback is provided physically and the user is not even able to exceed the minimum distance with the manipulator.This results in a restoring force or a locking of the exoskeleton or robot arm as soon as the minimum distance is reached or exceeded. Any combination of feedback is conceivable. It is preferred that, in addition to enlarging the virtual model and / or the virtual manipulator, a vector field is calculated that includes restoring vectors proportional to the restoring force of the virtual manipulator. The vector field can be formed based on the position and a probability or confidence distribution for the position of the virtual manipulator and / or virtual touch-critical object.Based on these positions and the calculated position errors, which can be based on statistical errors or sensor errors associated with the sensor data and / or model data, a probability or confidence distribution is created for the position of the virtual touch-critical object and / or the virtual manipulator. This is advantageous because the calculated position probability or confidence distribution continuously decreases toward the edge of the magnification. Thus, the force or stiffness required for collision avoidance can be low at locations with a low probability or confidence distribution, while it is high at locations with a high probability or confidence distribution.

[0044] It is preferred that the confidence values ​​be stored in a spatial data structure, for example, in a voxel map. The values ​​can assume the confidence or accuracy information of the sensors. Depending on the confidence values, the magnification of the virtual manipulator and / or virtual touch-critical object can be adjusted. This is advantageous for reliably avoiding unintentional collisions even with low confidence. If, for example, the manipulator is located in an area of ​​low sensor accuracy, this can be compensated for by increasing the magnification accordingly.

[0045] It is preferred that the user be able to manually change the determined minimum distance with the aid of one of the input devices of the user terminal, preferably in particular by changing the magnification factor of the virtual touch-critical object and / or the virtual manipulator. A change in the minimum distance results from a further magnification of the virtual touch-critical object and / or the virtual manipulator by eliminating a previous overlap of the virtual touch-critical object and / or the virtual manipulator due to their larger size. As already described above, it is preferred that a minimum distance is maintained in that the enlarged virtual components are not allowed to touch or overlap.The input to change the minimum distance can preferably be made either by pressing or turning a button, one or more voice commands, or one or more inputs via a mouse, keyboard, gamepad, or joystick. In the case of an input device designed as an exoskeleton or robotic arm, gestures for changing the minimum distance are also conceivable.

[0046] It is preferred that the changed minimum distance or the changed magnification of the virtual touch-critical object or of the virtual manipulator is positive in order to manually create an additional safety margin between the manipulator and the touch-critical object. It is further preferred that the minimum distance or the changed magnification of the virtual touch-critical object or of the virtual manipulator is negative in order to allow interaction between the manipulator and the touch-critical object. In the preferred scenario, a negative minimum distance or the changed magnification of the virtual touch-critical object or of the virtual manipulator enables contact between the manipulator and the touch-critical object, for example, to allow the two to interact with each other. An interaction could be gripping, pushing, pulling, cutting, welding, pushing, milling, etc.Such a negative minimum distance can be realized by a magnification factor less than 1, i.e. a reduction and / or a translational or rotational displacement of the virtual touch-critical object or virtual manipulator.

[0047] It is preferred that the change in the real distance between the touch-critical object and the manipulator be continuously adjusted based on the minimum distance or a change in the magnification of the virtual touch-critical object or the virtual manipulator to avoid jerky movements. A change in the sensor data quality, for example, can lead to a sudden change in the minimum distance or a change in the magnification of the virtual touch-critical object or the virtual manipulator, which can lead to jerky movements of the manipulator. In addition, jerky movements of the user in the exoskeleton or robot arm can occur, which must be avoided.Therefore, the preferred goal is to prevent such rapid movements and to perform sudden changes in the minimum distance or a change in the magnification of the virtual touch-critical object or the virtual manipulator at a limited and continuous rate, i.e. with a limitation in the time derivative of the magnification.

[0048] An appropriate rate could be determined based on the typical movement speed of the haptic device or robot. This varies for each haptic device and robot, as well as for each operator. For example, it could be set to 1-20% of this typical speed.

[0049] It is further preferred that the actual distance between the touch-critical object and the manipulator be taken into account when enlarging the virtual touch-critical object or the virtual manipulator, ensuring that the minimum distance or the magnification of the virtual touch-critical object or the virtual manipulator cannot be set greater than the actual distance between the manipulator and the touch-critical object in order to avoid jerky movements or active behavior of the algorithm. This scenario is particularly important if a user does not fall below the minimum distance and then the user or the system changes the minimum distance or the magnification of the virtual touch-critical object or the virtual manipulator so that it is greater than the current minimum distance between the touch-critical object and the manipulator.Here, large restoring forces can spontaneously occur both between the contact-critical object and the manipulator, as well as for the user in the exoskeleton or robot arm, which must be avoided.

[0050] It is preferred that the manipulator has additional sensors for more precisely determining the actual distance between the touch-critical object and the manipulator. Such sensors can be used to improve the quality of the virtual model and more accurately determine the exact distances between objects and the manipulator. Such sensors can preferably be configured as: cameras, infrared sensors, ultrasonic sensors, radar sensors, probes, capacitive sensors, resistive sensors, etc.

[0051] It is preferred that the specified minimum distance, or the magnification of the virtual touch-critical object, or the virtual manipulator, depends on the materials of the touch-critical object and / or the manipulator. For example, soft materials could allow for a smaller minimum distance than rigid materials, which carry a higher risk of causing an undesirable effect in the event of a critical touch. For example, a soft manipulator, such as a soft robot manipulator, could allow for a smaller minimum distance.

[0052] It is further preferred that the magnification factor of the virtual touch-critical object and / or the virtual manipulator depends on the materials of the touch-critical object and / or the manipulator. For example, soft materials could enable a smaller magnification factor than solid materials, which carry an increased risk of producing an undesirable effect in the event of a critical touch. For example, a soft manipulator, e.g., a soft robot, could enable a smaller magnification factor. In this scenario, the magnification factor of the manipulator could preferably be reduced if it is made of a soft material, or the magnification factor of the touch-critical object could be reduced if it is made of a soft material.

[0053] It is preferred that a plurality of objects are recorded in the virtual model. It is particularly preferred that the virtual model has at least one virtual object to be manipulated. It is further preferred that different or uniform minimum distances or magnifications are determined for different object-manipulator pairs and / or that the robotic system has a plurality of manipulators, the plurality of manipulators are recorded in the virtual model and different or uniform minimum distances or magnifications are determined for different object-manipulator pairs. A scenario for a plurality of objects could occur, for example, during a medical operation in which the robotic system detects a plurality of organs, a virtual object is generated for each of these organs, and different minimum distances or magnifications are determined depending on its importance for the patient's survival.This gives preference to organs that need to be manipulated and are sensitive to touch.

[0054] Preferably, the same applies to a medical surgical scenario in which multiple manipulators are used during the operation, for example, one for pushing away tissue. Here, the minimum distance between a first manipulator and tissue would be zero or negative, or the magnification of the manipulator or object would be zero or negative. Preferably, a second manipulator could comprise a scalpel, which should have a positive minimum distance between objects and the second manipulator, or have a positive magnification.

[0055] In the following, preferred embodiments of the invention are explained with reference to figures.

[0056] They show:

[0057] Fig. 1 : a preferred robotic system with preferred virtual model

[0058] Fig. 2: a preferred embodiment of the method for enlarging the virtual touch-critical object and / or a virtual manipulator Fig. 3: a preferred embodiment of the user terminal as robot arms

[0059] Fig. 4a - 4f: Different types of magnification of a virtual touch-critical object or virtual manipulator

[0060] The robotic system 1 according to Fig. 1 has two manipulators 2a and 2b configured as grippers. It further has sensors 3 configured to detect the manipulators 2a and 2b and object(s) 4 in the field of view of the sensors 3. Based on this sensor information, a virtual model 5 is created with a virtual object 6 and the virtual manipulators 7a and 7b.

[0061] Fig. 2a shows a close-up of an object 4 with manipulator 2a. Fig. 2b shows the virtual manipulator 7a and the virtual object 6. Fig. 2b also shows the enlarged virtual object 9 and the enlarged virtual manipulator 8a.

[0062] Fig. 3 shows a preferred embodiment of the input device as robot arms 10. The seating 11 allows the user to comfortably grasp and move the two robot arms 12a and 12b. Sensors in the robot arms are capable of transmitting movements of the user terminal to a robotic system and, thanks to the method according to the invention, maintaining safety distances from contact-critical objects. These minimum distances are preferably perceptible to the user, with a robot arm developing a restoring force counter to the user's force should a critical contact occur.

[0063] Fig. 4a shows a virtual touch-critical object 6 in the center with two different magnifications. These magnifications are uniform translational scalings around the object center. These are realized, for example, by a scaling factor or a vector with three equal values ​​for the three spatial directions. This type of magnification can also be applied to the virtual manipulator.

[0064] Fig. 4b shows a virtual touch-critical object 6 in the center with two different magnifications. The magnifications are uniform translational scalings around a point other than the object center. These are realized, for example, using a 4x4 scaling matrix that contains a corresponding shift to the top right as translation elements and three equal values ​​as scaling factors. This type of magnification can also be applied to the virtual manipulator. Fig. 4c shows a virtual touch-critical object 6 in the center with two different magnifications. The magnifications are safety layers (safety margin) realized, for example, using a geometric algorithm such as the VPS (Voxmap PointShell algorithm). This type of magnification can also be applied to the virtual manipulator.

[0065] Fig. 4d shows a virtual touch-critical object 6 in the center with two different magnifications. These magnifications represent non-uniform safety layers, realized, for example, by a geometric algorithm with a larger scaling of the distance values ​​along the vertical axis. This type of magnification can also be applied to the virtual manipulator.

[0066] Fig. 4e shows a virtual touch-critical object 6 in the center with a magnification. The magnification is a volumetric extrusion by rotation around a central axis 14. The central axis 14 is represented as a line in the center of the object. The rotation occurs clockwise and counterclockwise around the central axis by a few degrees. This type of magnification can also be applied to the virtual manipulator.

[0067] Fig. 4f shows a virtual touch-critical object 6 in the center with a magnification. The magnification is a volumetric extrusion by rotation around an axis 15 at the left edge of the object. The axis 15 at the left edge of the object is represented as a line on the left edge of the object. The rotation occurs clockwise and counterclockwise around this axis by a few degrees. This type of magnification can also be applied to the virtual manipulator.

Claims

Patent claims 1. A method for controlling a robotic system for maintaining a minimum distance between a robotic manipulator and a touch-critical object, wherein the robotic system comprises the manipulator and sensors, wherein based on the sensor information the touch-critical object and the manipulator are localized and a virtual model is created which has at least one virtual touch-critical object and one virtual manipulator, wherein the minimum distance is maintained by: Magnification of the virtual touch-critical object and / or the virtual manipulator, Detect as soon as an enlarged or un-enlarged virtual touch-critical object and an enlarged or un-enlarged virtual manipulator touch each other, Stop or slow down the movement of the manipulator when touching the enlarged or un-enlarged virtual touch-critical object and the enlarged or un-enlarged virtual manipulator.

2. Method according to one of the preceding claims, wherein the enlargement is realized by scaling the virtual touch-critical object and / or the virtual manipulator by multiplying by a magnification factor, magnification vector or a magnification matrix and / or wherein the enlargement of the virtual touch-critical object and / or the virtual manipulator is realized by security layers and / or wherein the enlargement of the virtual touch-critical object and / or the virtual manipulator is realized by extrusion.

3. Method according to one of the preceding claims, wherein the magnification factor for enlarging the virtual touch-critical object is different for different spatial axes, and / or the magnification factor for enlarging the virtual touch-critical object is dependent on the direction of movement and / or the speed of movement of the manipulator and / or the magnification factor for enlarging the virtual touch-critical object is dependent on the direction of movement and / or the speed of movement of the touch-critical object and / or the magnification factor for magnifying the virtual touch-critical object depends on the accuracy of the sensors used to create the virtual touch-critical object and the virtual manipulator in the robotic system, and / or the magnification factor of the virtual touch-critical object is the same in all directions.

4. Method according to one of the preceding claims, wherein the magnification factor for magnifying the virtual manipulator is different for different spatial axes, and / or the magnification factor for magnifying the virtual manipulator is dependent on the direction of movement and / or speed of movement of the manipulator, and / or the magnification factor for magnifying the virtual manipulator is dependent on the direction of movement and / or speed of movement of the touch-critical object, and / or the magnification factor for magnifying the virtual manipulator is dependent on the accuracy of the sensors used to create the virtual touch-critical object and the virtual manipulator in the robotic system, and / or the magnification factor of the virtual manipulator is the same in all directions.

5. Method according to one of the preceding claims, wherein the virtual model comprises, in addition to the virtual touch-critical object, a virtual object to be manipulated.

6. Method according to one of the preceding claims, wherein the robotic system has at least one user terminal with at least one input device which is capable of receiving inputs for controlling the robotic system from at least one user.

7. The method according to claim 6, wherein the at least one input device is designed to provide the user with feedback, in particular a force or torque feedback, as to whether the minimum distance has been reached or exceeded.

8. Method according to claim 6 or 7, wherein the user, with the aid of one of the input devices of the user terminal, is able to manually set the minimum distance to be maintained to change, in particular by changing the magnification factor of the virtual touch-critical object and / or the virtual manipulator.

9. The method according to claim 8, wherein the manually changed minimum distance is positive, or the minimum distance is negative to enable interaction between the manipulator and the touch-critical object.

10. Method according to one of the preceding claims, wherein the change in the real distance between the touch-critical object and the manipulator is adjusted at an appropriate rate based on the minimum distance in order to avoid jerky movements, or the real distance between the touch-critical object and the manipulator is taken into account, ensuring that the minimum distance cannot be set greater than the real distance between the manipulator and the touch-critical object in order to avoid jerky movements.

11. Method according to one of the preceding claims, wherein the robotic manipulator has additional sensors for more accurate determination of the actual distance between the contact-critical object and the robotic manipulator.

12. Method according to one of the preceding claims, wherein the minimum distance to be maintained depends on the materials of the contact-critical object and / or the manipulator.

13. Method according to one of the preceding claims, wherein a plurality of objects are recorded in the virtual model and different or uniform minimum distances are to be maintained for different object-manipulator pairs and / or wherein the robotic system has a plurality of manipulators, the plurality of manipulators are recorded in the virtual model and different or uniform minimum distances are to be maintained for different object-manipulator pairs.

14. Robotic system for carrying out the method according to one of the preceding claims.

15. A computer-readable medium comprising instructions which, when executed by a computer and / or a robotic system, cause the computer and / or robotic system to perform the method / steps of the method according to any one of the preceding claims.

Images