Method for determining values influencing the movement of a robot

Abstract

A method for determining values influencing movement of a robot is disclosed. The method includes the following steps: a) provision of a task to be performed by the robot and a worker; b) provision of a layout of a workstation; c) provision of tool data; d) determination of respective axial movement patterns of the robot on the basis of steps a) to c); e) provision of a worker workspace; f) determination of critical path points of the robot, where a specified movement speed is exceeded by the robot and/or a specified mass of an element to be moved by the robot is exceeded, on the basis of the axial movement patterns and the workspace; g) simulation of respective collisions at the critical path points by a second robot; and h) determination of permissible operating speeds of the robot for each critical path point on the basis of the simulated collisions.

Claims

1. A method for determining values influencing a movement of a robot, comprising the steps of: a) providing a task to be performed by a first robot and a worker; b) providing a layout of a workstation in which the task is to be performed; c) providing tool data that characterize a tool to be used by the first robot in performing the task; d) determining axial movement patterns, which are required for performing the task, of the first robot on a basis of information provided in steps a) to c); e) providing a workspace of the worker; f) determining critical path points of the first robot where a specified movement speed is exceeded by the first robot and/or a specified mass of an element to be moved by the first robot is exceeded on a basis of the determined axial movement patterns and the workspace; g) simulating, via a second robot coupled with the first robot, respective collisions of the first robot at the critical path points; and h) determining permissible operating speeds of the first robot for any given path point on a basis of the simulated respective collisions.

2. The method according to claim 1, wherein an enveloping space that surrounds the tool is established on a basis of the tool data.

3. The method according to claim 1, wherein the axial movement patterns are determined by a simulation or a measurement.

4. The method according to claim 1, wherein potential crushing or pinning sites where respective minimum distances between the workstation and the first robot are not maintained are determined on a basis of the axial movement patterns.

5. The method according to claim 1, wherein a temporal progression of reflected masses of the first robot is determined from actual robot impacts.

6. The method according to claim 1, wherein the simulated respective collisions are iteratively repeated with different operating speeds by a pendulum and a load cell until corresponding collision forces, collision pressures, and surface pressures no longer reach respective thresholds for the critical path points.

7. The method according to claim 1, wherein for the step of simulating the respective collisions, the second robot simulates corresponding impact directions and resistances for the critical path points.

8. The method according to claim 1, wherein for the step of simulating the respective collisions, the second robot is equipped with a force measurement system.

9. The method according to claim 1, wherein parameters of impact speed, impact mass and/or contact geometry of the first robot are modified in an automatic iterative process until respective specified limit values are no longer reached.

10. The method according to claim 1, wherein the critical path points are within the workspace of the worker.

11. The method according to claim 1 further comprising the step of: i) operating the first robot in accordance with the determined permissible operating speeds in performing the task.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a schematic, perspective view of a layout of a workstation in which a specified task is performed collaboratively by a lightweight robot and a worker, who is not illustrated in this figure;

(2) FIG. 2 shows a perspective view of a tool used by the lightweight robot shown in FIG. 1 to perform the specified task;

(3) FIG. 3 shows another perspective view of the workstation, wherein the worker, who is to perform the specified task in collaboration with the lightweight robot, and a workspace of the worker are illustrated;

(4) FIG. 4 is an illustration of an industrial robot, which is coupled to the lightweight robot shown in the preceding figures for simulating various impact scenarios;

(5) FIG. 5 shows a pendulum with a load cell by means of which collisions between the worker and the robot can be simulated; and

(6) FIG. 6 is a schematic illustration of different force-time progressions arising from the simulation of collisions by means of the industrial robot and the pendulum.

DETAILED DESCRIPTION OF THE DRAWINGS

(7) A layout of a workstation 10 is shown in a perspective view in FIG. 1. The layout of the workstation 10 can be provided as, for example, a CAD model for a planned station and its surroundings. A lightweight robot 12, which performs a specified task on a workpiece 14, is arranged in the workstation 10.

(8) The workstation serves as a so-called HRC plant, this abbreviation standing for human-robot collaboration. In the plant 10 shown here, this means that a worker (who is not illustrated here) and the lightweight robot 12 collaboratively perform a specified task on the workpiece 14 shown here.

(9) FIG. 2 shows a perspective view of a tool 16 that the lightweight robot 12 will use for performing the aforesaid task. An enveloping space 18 that surrounds the tool 16 is established around the tool 18. A gripper not designated in any further detail has a mass of, for example, ca. 1.2 kilograms, wherein the mass of the workpiece 14 is ca. 10 kilograms. A type of gripper envelope that takes corresponding geometries into account for possible collisions is provided by the enveloping space 18.

(10) In FIG. 3, the layout of the workstation 10 is illustrated again in another perspective view, wherein in this case the worker 20, who is to perform an accordingly specified task on the workpiece 14 in collaboration with the lightweight robot 12, is now illustrated. Furthermore, a workspace 22 of the worker 20 is illustrated around the worker 20. This depicts the workspace of a 95 percentile person.

(11) Provision is therefore made of the task to be performed by the lightweight robot 12 and the worker 20, as well as of the layout of the workstation 10 in which the task is to be performed. Further provision is made of corresponding tool data (e.g., mass, geometry, and the like) for the tool 16 that the lightweight robot 12 will use for performing the task. If this data is not available, contours as well as load data (masses and centers of gravity) estimated by abstraction will be used.

(12) Respective axial movement patterns of the lightweight robot 12 that are required for performing the task can be determined on the basis of this information. This can be achieved, for example, either with a pure simulation or with a corresponding measurement. All robot positions necessary for performing the task can thus be determined beforehand. For example, the axial movement patterns acquired in this manner can be exported and used to provide target points and different axis positions of the lightweight robot 12. The target positions and corresponding tool data can be transformed in a robot control interface, after which a program relating thereto can be run, wherein corresponding axis values can be logged. This can take place on a simulation system as well as on an actual robot.

(13) The so-called reflected mass can then be determined, this determination providing a theoretical progression of the reflected mass over time. This serves as an indicator for large masses in particular that act on the lightweight robot 12 or, in the event of an impact with the worker 20, on the worker 20.

(14) Critical path points of the lightweight robot 12 are determined within the workspace 22 of the worker 20 in particular, at which points a predetermined movement speed is exceeded by the lightweight robot 12 or a predetermined mass of an element to be moved by means of the robot 12 (in this case the tool 16 plus the workpiece 14) is exceeded, wherein the previously determined axial movement patterns and the specified workspace 22 are taken into account. On the basis of the axial movement patterns, it is furthermore possible to determine potential crushing or pinning points, in particular within the workspace 22 of the worker 20, where respective minimum distances between the workstation 10 and the lightweight robot 12 as well as its tool geometries or the workpiece geometries are not maintained.

(15) Furthermore, the tool and workpiece geometries relevant to a collision (i.e., sharp edges and chamfered corners, for example), as well as any collision detection and avoidance strategies that may be in place or in use, can and should be taken into account.

(16) Critical path points are identified as such and selected in the event that particularly high speeds, large masses, and large momentums resulting therefrom arise.

(17) The geometries on the tool or workpiece that are relevant to a collision, whether pinning or impacts are involved, and which body parts are at risk must also be factored into the calculation of the critical path points.

(18) In this process, zones that do not lie within the workspace 22 of the worker 20 can be masked or not considered. In a particularly precise embodiment, the movement of the robot 12 can be broken down into individual points at intervals of a few milliseconds. Each single resulting point can then be selected and made safe accordingly.

(19) The lightweight robot 12 and another industrial robot 24 are shown in FIG. 4. The industrial robot 24 serves as a flexible collision station, by means of which any number of impact scenarios can be simulated with variable tool loads. In other words, a test setup consisting of the industrial robot 24 equipped with a force measurement sensor system is used with the lightweight robot 12. In this process, the industrial robot 24 can be operated in the space in a flexible manner and set in such a way that it can precisely simulate respective impact directions and corresponding resistances in the case of an impact. The setup shown here is a physical construction, but in principle could be integrated in a digital process.

(20) The simulation described above is dependent on the trajectory, in other words the movement path of the robot, the tools/workpieces being manipulated, the type of robot used, and the collision detection and avoidance strategies employed.

(21) For the path points of the lightweight robot 12 determined as critical beforehand, or if necessary for all points of the path, respective collisions are thus simulated by means of the industrial robot 24, which is coupled with the lightweight robot 12. In this process, respective collision vectors are determined for the appropriate collision points corresponding to the critical path points. In doing so, the industrial robot 24 embodies an impact of the lightweight robot 12 against a wall.

(22) In this simulation of collisions, the robot doing the impacting can be interchanged if investigations need to be performed for a certain type of robot, a certain software status, or for a certain hardware status.

(23) FIG. 5 shows a perspective view of a pendulum 26, which has a load cell 28. After performing respective impacts by means of the test setup shown in FIG. 4, which comprises the industrial robot 24 and the lightweight robot 12, respective impacts are translated by the pendulum 26 shown here to the load cell 28. The mass of the pendulum 26 can be varied so that the actual tool mass can be simulated exactly. The pendulum can always be deflected to a sufficient degree so that it acts on the load cell 28 with the same speed that was simulated previously by means of the industrial robot 24, as shown in FIG. 4.

(24) The collisions simulated by means of the industrial robot 24 are therefore iteratively repeated by means of the pendulum 26 and the load cell 28 with different operating speeds (i.e., different deflections of the pendulum 26) until corresponding collision forces, collision pressures, and surface pressures no longer reach the respective threshold values for each of the critical path points.

(25) Different force-time progressions are plotted in a schematic diagram in FIG. 6. The line 30 shows an example force-time progression that would arise from an actual impact of the lightweight robot 12 on the load cell 28. Also shown are other force-path progressions 32, 34 arising from the use of the pendulum 26. Different biofidelic load limits will result, depending on what kind of collisions are simulated with, for example, different body parts of the worker 20. The deflection of the pendulum 26 is iteratively repeated for the simulation of the respective collisions of each of the path points ranked as critical until biofidelic load limits are no longer reached in each case.

(26) Ultimately it is thus possible to determine actual collision forces and collision pressures for respective path points ranked as critical, or every point on the path of the lightweight robot 12 in minimum distances. Furthermore, permissible operating speeds for each of the path points of the lightweight robot 12 deemed as critical are determined.

(27) These are determined on the basis of all relevant geometries (corners, edges, etc.) on the tool and the workpiece.

(28) Information on the respective speeds, masses, and the momentums resulting therefrom is thus available for each of the critical path points. Accordingly, forces and surface pressures arising in each case can be evaluated and then checked for compliance with current standards. Should this not be the case, corresponding operating speeds can be varied until these standards are fulfilled. Thresholds (pain thresholds, injury thresholds) introduced at a later date could also be taken into account. The collision strategies, parameters, and zero space and orientation positions could likewise be varied and thus optimally adjusted in this process step.

(29) In an advantageous configuration, impact effects can be simulated by using suitable impact elements (foam materials or the like) and critical path points can then be padded and documented as optimization recommendations. Functions for correcting free body impact to restrained impact (see IFF Study, TS 15066), etc., can also be considered.

(30) The determination of critical path points described above requires expert knowledge to some extent. If such knowledge is not available or if the most reliable approach possible is desired, a sequential testing of a path is recommended, in which the process is broken down into incremental partial points and each of these points is tested. The advantages of this method lie in comprehensive testing, which can then also be documented, as well as in complete transparency. Furthermore, because the fully automated configuration does not give rise to any expense or effort, this testing does not entail any significant additional expense or effort versus the testing of just the path points deemed critical.

(31) There are diverse application possibilities for the methodology described above: on the one hand, it has applications in the tendering and planning phase of a plant, it can serve as a support for the developers of such plants, as an aid in plant project planning, for example to carry out a cycle time optimization at an early stage in the plant planning.

(32) Furthermore, the methodology has applications in the startup, acceptance, and release as well as in the CE approval of a plant. In the event that adjustments and optimizations are being made to existing plants, such adjustments and optimizations can then be reevaluated with the method. During the operation time of a plant, maintenance measures, corrective measures, and further process optimization measures can be carried out in this manner. The methodology described herein can be used in an advantageous manner in the event of plant modifications such as expansions due to new variants or series, new technologies, materials, tools, or the like.

(33) This methodology can furthermore be used when normative framework conditions change, for instance if the permissible collision forces or collision pressures or body models change.