DE102012103830A1 - Method for preventing deadlock of a pair of robots in multi-robot system, involves avoiding deadlock condition during execution of programs by automatically determining and executing deadlock-free motion statement - Google Patents

Abstract

The method involves identifying portions of common workspace (10) occupied by robots (6,8), during concurrent execution of programs. The interference region is identified and analyzed, and occurrence of deadlock condition of two robots is identified. The deadlock condition is avoided during execution of programs by automatically determining and executing the deadlock-free motion statement prior to execution of any motion of the robots that results in deadlock condition. An independent claim is included for computer readable media storing program for preventing deadlock of a pair of robots.

Description

CROSS-REFERENCE TO APPROPRIATE REGISTRATION

This application is a request for partial further handling of pending US Patent Application Serial No. 21 May 2008 filed. 12/124 430.

This application claims the benefit of US Provisional Patent Application Serial No. 5, filed May 5, 2011. 61/482 808.

FIELD OF THE INVENTION

The present invention generally relates to a system for controlling a plurality of robots and a method for controlling the avoidance of movement disturbances for the plurality of robots.

BACKGROUND OF THE INVENTION

Moving objects in space is a necessary task in a typical manufacturing environment. Robotics has been increasingly used to effect the necessary movement. However, if several objects are moved, there is a potential for interfering with these objects. A disturbance exists when the at least two objects share the same space at the same time, such as when the objects have the same coordinates with respect to a common frame of reference.

Because modern industrial robots move at considerable speeds, interference between robots can result in collision and unwanted damage to them and the workpieces carried by the robots. The collisions can lead to expensive downtime in the manufacturing process. Consequently, it is desirable to avoid such collisions.

Prior art systems and methods have been used in attempting to minimize interfering influences and collisions. However, there are several deficiencies in the systems and methods of the prior art. It is typical that a tool working point (TCP) is checked only with respect to a predetermined disturbance area or "static space". With multiple robots, it is difficult to directly or effectively prevent their collision or interference. Further, it is difficult to specify a trouble space with respect to a resting coordinate system for a plurality of moving robots. Each disturbance space is not only a function of the moving distance of the robot, but also a function of the movement speed. Difficulties also exist in attempting to handle a blockage situation when two or more robots request to move simultaneously to a common space.

Prior art systems also attempt to prevent a TCP from colliding with a robot in a fixed space with respect to its world coordinate system. If multiple robots (with multiple controllers) share a common room or "trouble room" during a task execution, each control unit must wait until there is no robot in the common spaces. Then, the control unit can output the motion control commands for the robot to move. This process is also called a "wait-and-go process," which generally increases the time of a work cycle. However, it is difficult to effectively set a trouble space as compared with a fixed coordinate system because the trouble space is not only the function of the moving distance but also the moving speed of the robot. When more than one robot requests to move simultaneously to a common space, a situation of mutual blocking is created in which none of the robots can move because they are waiting for each other.

In prior art systems, it is also attempted to model the robot by bullets and cylinders. These systems dictate a future location of the robot during real-time motion. Because these systems do not determine the space occupied by the robot over time, these systems must constantly compare during robot movement. These systems compare element by element the models of all robots in the processing chamber. This comparison is computationally very costly and the cost potentially increases as the number of robots and elements used to model a robot and machine equipments increases. Because the comparison is made in real time when an imminent collision is detected, these systems must normally stop all robots affected by the impending collision, and an automatic programmed operation must be set. These systems become more difficult when the robots are on different control units because they require large amounts of information to be transferred between controllers in real time. Prior art systems have also attempted to prevent input disturbance from interfering with Receipt operation to use. There is no need for memory programmable input-output control in the present invention.

One known system and method is disclosed in co-pending International Application No. PCT / US2007 / 066638 of the assignee, which is incorporated herein by reference in its entirety. The system and method include a "dynamic space inspection system" in which the efficiency of a robotic operation is maximized and a potential for interfering or colliding multiple robots is minimized. Robots controlled by each control unit only operate on a user-defined, dynamic space, thus avoiding a collision. However, the dynamic space checking system generally protects TCP only against a user-defined, straight-line space.

Another known method for preventing robotic collisions is described in Pollack et al. other members U.S. Patent No. 5,150,452 reported. The method includes generating a collision image template containing a desired robot motion. The home position of the desired robot is removed from a "world image template" by combining the robotic image template and the world image template in a logical exclusive OR operation, and then combining the collision image template and the world image template in a local exclusive OR operation, combining the Collision image template and the world image template in a logical inclusive OR operation byte by byte. A crash is indicated by a difference in any bit position of the inclusive and exclusive OR joins. The method provides a two-dimensional xy projection and a one-dimensional height for collision detection, but does not allow real-time three-dimensional collision detection.

Another known method for detecting a collision between a robot and one or more obstacles before it occurs is described in Greenspan et al. other members U.S. Patent No. 5,347,459 described. The robot is modeled by bullets in a voxelized workspace. Each voxel within the workspace is assigned a value that corresponds to its distance from the nearest obstacle. A collision is determined to be imminent if the voxel value at the center of one sphere is less than the radius of another sphere in voxels. However, the method only protects a single robotic arm. The robot is also modeled only by bullets, which results in insufficient protection of critical process paths of the robot.

There is a continuing need for a system and method for controlling the prevention of motion disturbance for a plurality of robots. It is desirable that the system and method provide three-dimensional real-time collision detection, advance robotic motion transmission to the robotic system, reservation of programmed pathways without collision and protection of critical process paths.

Conditions of mutual blocking occur when the programs or tasks are in such a state that continued, sequential execution of any one of the programs or tasks results in interference between one or more robots associated with the programs or tasks. and that there is no program or task that can progress in a sequential manner without mutual interference.

The Chaffee et al. granted U.S. Patent No. 7,114,157 describes a method for preventing mutual blocking by detecting aids in a set process. While mutual blocking can be avoided by this method, it does not provide a capability to allow a pass if the set deadlock-free sequence has not been determined. Furthermore, it does not allow any operation that can be free of deadlock if the requested operation is outside of the prescribed, deadlock-free, flow.

US patent application, publication no. 2009/0326711 for Chang et al. describes a method of using automatic ranges to avoid deadlock either automatically or manually. Chang proposes a method of using priority values to prevent mutual blocking. Chang does not suggest a method of preventing mutual blocking that does not use priority values.

SUMMARY OF THE INVENTION

In accordance with the present disclosure, surprisingly, a system and method for controlling motion disorder prevention for a plurality of robots has been found, the system and method providing real-time three-dimensional collision detection, robot motion in advance to robotic system, reservation of programmed paths without collision and the protection of crucial process paths without using the mechanism of an input-output acknowledge operation.

In one embodiment, there is a method for preventing mutual blocking of a pair of robots having a common work area in which each of the robots is controlled by an associated program, wherein if the programs are executed simultaneously, each of the robots has at least a portion of the shared work area during part of the execution of the associated program, the method comprising the steps of: while concurrently executing the programs, identifying portions of the shared work area occupied by the robots; Marking at least one fault area in which the portions of the common work area overlap; Analyzing the at least one fault area and identifying where at least one condition of mutual blockage of the two robots may occur; and avoiding the at least one condition of mutual blocking during execution of the programs by automatically determining and executing at least one instruction of mutual blocking free movement prior to performing any movement of the robots, resulting in the at least one mutual blocking condition.

In another embodiment, there are computer readable media containing instructions executable by a computer to perform a method of preventing mutual blocking of a pair of robots having a shared workspace, each of the robots being controlled by an associated program in which each of the robots occupies at least a portion of the shared work area during an execution portion of the associated program when the programs are concurrently executed, the method comprising: while concurrently executing the programs, the instructions, the portions of the common, by the robots marked occupied work area; the instructions identifying at least one region of disturbance in which the portions of the common work area overlap; the instructions that analyze the at least one fault area and identify where at least one condition of mutual locking of the two robots may occur; and the instructions that prevent the at least one mutual blocking condition from executing the programs by automatically determining and executing at least one mutual blocking free movement instruction prior to executing any movement of the robots, resulting in the at least one mutual blocking condition ,

In another embodiment, there is a method for preventing mutual blocking of a plurality of robots having a common work area, each of the robots being controlled by an associated program, wherein, when the programs are executed simultaneously, each of the robots comprises at least a portion of the robot shared work area during a portion of the execution of the associated program, the method comprising the steps of: while simultaneously executing the programs, identifying portions of the shared work area occupied by the robots; Characterizing at least one fault area, wherein the portions of the common work area overlap for at least two of the robots; Analyzing the at least one fault area and identifying where at least one condition of mutual blockage of the at least two robots may occur; and avoiding the at least one condition of mutual blocking during execution of the programs by automatically determining and executing at least one mutual blocking free movement instruction prior to performing any movement of the at least two robots, resulting in the at least one condition of mutual blocking ,

The system and method of the present invention do not require a previously known free-from-blockage process. This system and method will automatically determine the potential conditions of deadlock and recognize a possibility to avoid these conditions. This system and method does not require priority values to prevent mutual blocking. This system and method avoids interference conditions while avoiding the possibility of conditions of mutual blockage.

DESCRIPTION OF THE DRAWINGS

The above-mentioned as well as other advantages of the present invention will become apparent to those skilled in the art upon consideration of the following detailed description of a preferred embodiment given the accompanying drawings, in which:

1 shows an exemplary robotic system according to the present disclosure having a first robot and a second robot operating in a work area;

2 FIG. 10 is a process flow diagram illustrating a method of fault checking of an automatic area according to the present disclosure; FIG.

3 shows a voxelated model of the respective first and second robot, which in 1 are shown;

4 is an isometric representation of the voxelated model used in 3 is illustrated;

5 shows an enlarged view of each first and second, in 1 a robot having a voxelated model formed by voxelized balls and cylinders respectively overlaid on the first and second robots;

6 shows the in 5 illustrated voxelated spheres and voxelated cylinders formed from a plurality of voxels;

7 Figure 4 shows a mutual blocking free matrix and associated program listings according to the invention;

8th shows a fault matrix and associated program listings according to the invention;

9 Figure 12 shows paths and associated tables that combine multiple areas of mutual blocking into a single, unblock-free area;

10 shows several execution sequences of a robot program without and with mutual blockage prevention according to the invention;

11 Figure 3 is the flowchart of a mutual blockage avoidance procedure according to the invention; and

12 Figure 3 is a flow chart for determining potential areas of mutual blockage according to the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The following description is exemplary only and is not intended to limit the present disclosure, application, or uses. It should be understood that throughout the drawings, like reference numerals indicate like or corresponding parts and features. With regard to the disclosed methods, the illustrated steps are exemplary and thus not necessarily or decisively.

1 is a schematic diagram showing a robot system 2 to control avoidance of collisions between multiple robots. As a non-limiting example, the robot system 2 either a body shell robot system, a water jet cutting robot system, a laser welding robot system, an arc welding robot system, and a paint robot system having at least two robots. Other robot systems 2 with a variety of robots can also be used as needed.

The robot system 2 includes a workspace 4 which forms a cladding within which a first robot 6 and a second robot 8th are constructed so that they are in operation. The first robot 6 and the second robot 8th are designed so that they selectively at least one within the work area 4 arranged common space 10 occupy. Although only a first robot 6 and a second robot 8th should be clear that the robot system 2 have more than two robots without departing from the scope and spirit of the disclosure.

The first robot 6 can during a movement of the first robot 6 along a first programmed route a first section 12 of the common room 10 taking. The second robot 8th can during a movement of the second robot 8th along a second programmed route a second section 14 of the common room 10 taking. The first and second sections 12 . 14 are also known as "Auto Areas" or "Auto Zones". Each programmed route includes one or more automatic ranges, and each automatic range is preprocessed and derived from one or more motion instructions. The first and second programmed routes may be controlled by one or more movement instructions in an order of operations, for example, with multiple movement instructions. A skilled craftsman should understand that the first and second sections 12 . 14 What can be covered in the absence of a method of interference prevention for controlling the movements of the first and second robot 6 . 8th leads to the probability of a robot collision.

The first and second robots 6 . 8th be through at least one control unit 16 . 18 controlled. In the in 1 illustrated embodiment, the first robot 6 through the first control unit 16 and the second robot 18 through the second control unit 18 controlled. The at least one control unit 16 . 18 is designed for electrical transmission with an electrical power source (not shown). The control unit 16 . 18 can be the order of instructions, such as one in the control unit 16 . 18 computer program. In other embodiments, the instruction sequence may be on a computer-readable medium or in communication with the controller 16 . 18 be located memory.

The robot system 2 may further include other components known per se, such as network media (not shown) configured to interconnect the various system components, a programmable logic device (not shown) and / or in electrical communication with the at least one controller 16 . 18 located teaching device 20 , In a specific embodiment, the teaching device 20 include a monitor and is designed to be the first robot as needed 6 , the second robot 8th , the first section 12 of the common room 10 and the second section 14 of the common room 10 for viewing by an operator of the robot system 2 graphically. The teaching device 20 may include means for initiating a sequence of instructions to impact the robot.

An exemplary set of instructions in accordance with the present disclosure is in 2 shown. The instruction sequence includes a fault checking procedure 200 for automatic areas. The fault test procedure 200 for automatic ranges includes first an introductory step 202 for providing the at least one common room 10 within the workspace 4 , The introductory step 202 closes a first determination step 204 and a second determination step 206 at. The first determination step 204 includes determining the first section 12 of the common room 10 that during the movement of the first robot 6 occupied along the first programmed route. The second determination step 206 includes determining the second portion 14 of the common room 10 who is in the movement of the second robot 18 along the second programmed route. It should be clear that the first section 12 and the second section 14 can be automatically determined, ie, without requiring a user-defined operation area for each robot, as is known in the art. The first paragraph 12 and the second section 14 are then in a comparison step 208 compared to see if there is any overlap between them 210 exist.

As a non-limiting example, the first determining step 204 and the second determination step 206 during an initial operation of the robot system 2 with the first robot 6 and the second robot 8th essentially in real time. The initial operation can be done with the first robot 6 and the second robot 18 in a locked mode to avoid their potential collisions. In another example, the first determination step 204 and the second determination step 206 by performing runs of the first and second programmed links in unconnected mode, for example the first section 12 and the second section 14 of the common room 10 to be identified by the robots 6 . 8th during real-time operation. The first paragraph 12 and the second section 14 during the first and second determination steps 204 . 206 Further, for example, they may be retained on the memory and on subsequent operations of the first robot 6 and the second robot 18 to be used again. It should be understood that the programmed tracks or tracks of the robot are preprocessed and stored in this manner to counteract a collision thereof.

In response to the presence of the overlap 210 become first and second movement steps 212 . 214 for the first robot 6 and the second robot 8th selected. If it is between the first section 12 and the second section 14 an overlap 210 There will be a collision of the robot 6 . 8th considered probable, and it becomes the first move step 212 selected. The first movement step 212 involves moving only one of the first robot 6 along the first programmed route and the second robot 8th along the second programmed route. If between the first section 12 and the second section 14 no overlap 210 is present, will be a collision of the robot 6 . 8th considered unlikely, and it becomes the second move step 214 selected. The second movement step 214 involves moving the first robot 6 along the first programmed route and moving the second robot 8th along the second programmed route.

The first and second determination steps 204 . 206 can be handled by any means known in the art. In a specific embodiment according to the present disclosure, the first determination step 204 and the second determination step 206 the step representing the common space 10 , the first section 12 and / or the second section 14 with at least one voxel model 300 include, at least one voxel 302 has, as it is in 3 to 6 is shown clearly. The voxels 302 are volume elements that represent a value on a regular grid in a three-dimensional space. The comparison step 208 can the step Determining whether, for example, between a first section 12 performing, first voxel model 300 and one the second section 14 representing the second voxel model 300 the overlap 210 is available. The voxel model 300 is generally due to the multitude of voxels 302 represented a shape of various components such as a base, an arm and a tool of the first robot 6 and the second robot 8th come closer.

Regarding 3 and 4 can be an explanatory voxel model 300 include a plurality of voxels arranged at coordinates along an X-axis, a Y-axis, and a Z-axis. The voxel model 300 is designed so that it is a three-dimensional volume of the common space 10 illustrated by the first robot 6 and / or the second robot 8th is taken. The voxel model 300 may be in the form of a data file that is generated and, for example, on the control unit 16 . 18 or stored on another computer readable medium or memory. In certain embodiments, the voxel model is 300 dynamic and includes a variety of voxel models 300 that are associated with a sequence of robot movements. The dynamic voxel model 300 can be used to create the three-dimensional volume of the common space 10 to illustrate during the movements of the first robot 6 and / or the second robot 8th is taken. Each automatic section becomes an accumulation / overlay of multiple snapshots of voxel models 300 derived, each snapshot being taken at one or more ITP intervals

With reference to now 5 to 6 can the voxel models 300 of the first robot 6 and the second robot 8th at least one voxelated ball 500 and a voxelated cylinder 502 include that of the first and second robots 6 . 8th represent occupied three-dimensional volume. As will be further described herein, the voxelized spheres 500 and the voxelized cylinders 502 the various components such as a base and an arm of the first robot 6 and the second robot 8th represent. The voxelated balls and cylinders 500 . 502 generally contain a multitude of voxels 302 , which are arranged relative to each other, to the general shapes and limitations of the first robot 6 and the second robot 8th to get closer.

The ordinary person should understand that the voxelated balls 500 and cylinders 502 can be generated by any suitable means. For example, the voxelated ball 500 By approximating that first an exemplary sphere, the components of the robot 6 . 8th is projected onto an XY plane. It is then a smallest cylinder "A", which encloses the ball marked. The same sphere is then projected onto a YZ plane and the smallest cylinder "B" enclosing the sphere is marked. The same sphere is then projected onto a ZX plane and the smallest cylinder "C" enclosing the sphere is marked. The voxelated ball 500 is then approximated by marking the intersections of the cylinders "A", "B" and "C".

In another embodiment, the voxelated ball 500 be approximated by first finding out a smallest box, which is the component of the robot 6 . 8th enclosing ball. The box has a volume occupation "A". The volume occupancy "B" of the voxel, which is inside the box "A" and outside the sphere, will be marked afterwards. The voxelated ball 500 is then approximated by subtracting the volume "B" from the volume "A".

The voxelated ball 500 may alternatively be approximated by identifying an intersection of the exemplary sphere and the XY plane. The intersection contains the center of the sphere along the Z-axis and forms a sectioned, circular plate "A1". The circular plate "A1" is assigned an arbitrary height, and then the voxel occupation of the cut-through circular plate "A1" is marked. Further circular plates "A2", "A3" ... "A (n)" are located by shifting the intersection of the exemplary sphere and the XY plane along the Z-axis, also identifying their volume occupancies. The voxelated ball 500 is then approximated by integrating each of the voxel fillings identified for the circular plates of the example sphere.

The voxelled cylinder 502 can be approximated by marking the voxel occupation "A" of a left hemisphere / whole sphere, identifying a voxel occupation "B" of a right hemisphere / whole sphere, and identifying a voxel occupation "C" of several circular plates between the left hemisphere and the right hemisphere. Then, a circular base plate is created which is perpendicular to a baseline. The baseline connects the two hemispheres at the ends of the cylinder. Subsequent circular planes can be derived by translating the circular base plate along a Z-axis. Alternatively, a voxel population of the baseline may be calculated and later the baseline shifted to fill the entire area between the two hemispheres. The voxelled cylinder 502 is later approximated by integrating each of the voxel populations "A", "B" and "C".

An alternative method of approximating the voxelized cylinder 502 first includes identifying a smallest box that may enclose an exemplary cylinder representing the robot component. The box has a volume occupation "A". Then, a voxel volume occupation "B" located inside the box and outside the cylinder is marked. The voxelled cylinder 502 is later approximated by subtracting the voxel occupation "B" from the voxel volume occupation "A". It should be understood by those of ordinary skill in the art that other means may be used as appropriate to approximate the components of the robot 6 . 8th performing, voxelated balls 500 and cylinders 502 are suitable.

Accordingly, voxelation can also be determined from any arbitrary CAD surface or volume, which may represent robots, arm dressouts, and tooling.

Voxelization is a very efficient method of representing the space occupied by a robot motion segment. The voxelization process has a good deal of instruction processing time to produce the voxelated space, but the runtime component is very efficient. Once a space is voxulated, the maximum storage requirement of a voxelated space is fixed, no matter how complex or how much the common space is occupied by the robot during the travel distance. The runtime test for interference between voxelated spaces is very efficient. Although the preferred embodiment provides for displaying automatic areas with perturbation inspection through voxelled areas, any method of volume or surface representation of the space occupied by a robot during a travel distance may be used.

In other embodiments, collision is avoided during a programmed movement, a jogged motion, or other movement initiated motion. They also protect a moving robot from colliding with a stationary robot or with other stationary objects or other fixed areas in the processing chamber. For programmed motion, the voxelization process or other modeling process for fault checking of automatic regions may be computer-independent or during a test run, and the voxelated data may be efficiently stored for later retrieval. This process can also occur in real time when a new moving distance is determined, such as from a new program, robot jogging, or other means-triggered movement. For these cases, the flow of instructions may be determined by using a teaching device or by the other means that initiate the movement. It is not allowed that the actual movement starts until the process is completed and there is no interference with this Fault Check Automatic Area and other occupied Fault Check Automatic areas.

In further embodiments, the first section 12 and / or the second section 14 of the common room 10 a tooling space and a dressout assembly (eg wire feed unit for arc welding tool) occupied by tooling attached to the first robot 6 and / or the second robot 8th is attached. The step of determining the coverage 210 can also compare the first section 12 and / or the second section 14 with a third section of the common room 10 which is occupied by a non-robotic obstacle (not shown). The obstacle can be specified manually by a set of parameters, as needed to define the size, shape and location of the common space 10 be used. In another example, the method may 200 further comprising the step of determining the third portion of the common space occupied by the obstacle 10 include. It should be understood that the third section of the common space occupied by the non-robotic obstacle may also be voxelized as needed 302 in a similar voxel model 300 can be represented.

The procedure 200 The present disclosure may further include the step of determining an unoccupied portion of the common space 10 include after the first robot 6 has moved along the first programmed distance and / or the second robot 8th has moved along the second programmed route. The unoccupied section of the common room 10 can then be released for use by another robot, which joins as needed another programmed route or train. In a further embodiment, the method 200 the step of transmitting the coordinates of the first section 12 and second section 14 of the common room 10 on the at least one control unit 16 . 18 for further use according to the method 200 of the present disclosure. The control unit 16 . 18 may then be for handling at least one of the first and second determination steps 204 . 206 , as described here, and for converting the coordinates into the voxel models 300 that the first section 12 and the second section 14 be used.

The skilled artisan should recognize that when determining whether between the first section 12 and the second section 14 of the common room 10 the overlap is present, the respective movements of the first robot 6 and the second robot 8th can be determined by priority values. As a non-limiting example, the priority values of the first robot 6 and the second robot 8th be based on a Durchrück system. In another example, one of the first and second robots 6 . 8th always a higher priority value than the other of the first and second robots 6 . 8th exhibit. The higher priority value may be a predetermined user preference for priority of the robots 6 . 8th in the robotic system 2 be based on. In another example, the priority values may be based on the availability of unoccupied sections within the shared space 10 to be selected. A skilled craftsman can give the desired priority values to the first robot when needed 6 and the second robot 8th assign.

In a further embodiment, the method comprises 200 and the step of analyzing the plurality of programmed paths for the presence of the overlap 210 which may result in a state of mutual blockage. A functional sequence of the plurality of programmed routes can then be adjusted as needed to counteract the occurrence of the state of mutual blocking. The functional sequence can be set manually or automatically as needed, for example on the basis of priority values.

It should be clear that the fault checking system 2 for automatic areas and the procedure 200 Advantageously, the present disclosure provides a competitive advantage over conventional multi-arm robotic systems. The system 2 and procedures 200 cause the avoidance of a three-dimensional collision in real time. The system 2 requires minimal programming because it counteracts the need to specify or show disturbance ranges. It also reduces production downtime.

It should also be appreciated by one of ordinary skill in the art that the use of voxelated spheres 500 and cylinders 502 typically reduces the processing requirements associated with creating robotic models to the minimum. Movements of the first robot 6 and second robot 8th be in advance to the system 2 , for example, by a computer-independent generation of the voxel models 300 , transfer. Likewise, programmed paths or paths of the first robot 6 and the second robot 8th reserved without collision. A process path is a sequence of movements during which a user process such as cutting, welding, grinding, laying, painting, dispensing or other similar processes may take place. Generally, it is important that the entire process path is completed without interruption as soon as a process path has begun. It should also be clear that crucial process paths of the first robot 6 and second robot 8th similar to the present method 200 to be protected.

The system 2 Supports the avoidance of collisions of multiple robotic arms within the same control unit 16 . 18 and as needed via multiple control units 16 . 18 , system 2 and procedures 200 further provide a simplified robotic configuration and address the need for an I / O handshake protocol, as is handled with certain prior art systems and methods. The fault checking system 2 for automatic areas and the procedure 200 simplify the configuration of multi-arm failure checks and counteract mutual blockage of multiple robotic arms.

The prevention of interference and the prevention of mutual blocking are closely related. Disruption occurs when two robots try to occupy the same physical space. One method of interference prevention involves checking the space taken or taken at the instantaneous move command from all robots, and stopping one or more robots before the disturbance occurs. This type of interference prevention has the major advantage that it is very susceptible to deadlock when one or more of the suspended robots is currently occupying the space needed or needed by another robot and the stopped robot can not progress because it is blocked by another robot.

Another method of interference prevention is to manually set up fault areas by entering into the robot programs through program entry and exit area commands. These commands would effect communication with a programmable logic controller or other robots that would make it easy for only one robotic arm to occupy a particular area at a time. For a single area this is of little importance. The basic problem would be one of priority; who would get the area first if several robots are waiting for it to become available. For several areas, however, it is very easy for mutual blocking to occur. For example, it is easy to see for adjacent areas if each robot currently occupies one area and the next movement is within the area occupied by the other robot. In this case, each robot will constantly wait for the other robot to abandon the desired area. There will be a state of mutual blockage

The system and method according to this invention excludes the state of deadlock in the above cases by setting the instructions of a blockage-free movement prior to performing the motions having potential conditions of mutual blockage. The determination of instructions of a blockage-free movement may be made independently of the computer, outside of normal execution, or during normal production execution. If enough CPU processing time is available, setting during normal production execution provides the most flexibility to respond to dynamic conditions such as changes in the timing of input / output or the timing of external events or operations. In order to minimize the load on the CPU, the definition is made independent of the computer, whereby many permutations of program sequences can be analyzed and an optimized execution sequence can be found.

The system and method of the present invention does not require a deadlock-free process previously known. System and procedures will automatically determine the potential conditions of mutual blocking and identify a way to avoid these conditions. The system and method do not require priority values to prevent deadlock. The system and method prevent conditions of disturbance and at the same time the possibility that mutual blocking conditions exist.

In 7 is a matrix free of "mutual blocking" 400 of row numbers for a listing 401 "Original Program A" to be traversed by a first robot and a listing 403 "Original program B", which is to be traversed by a second robot, shown. Hatched fans in the matrix 400 represent a rising, outwardly curved shell, which forms an area free from mutual blocking. A plurality of the hatched shelves represent program line pairs that have interference during movement on a route from line "i-1" to line "i" of one program and line "j-1" to line "j" in the other program , The remaining compartments must complete the rising, outwardly curved shell. The fans in the matrix 400 may be denoted in a form (i, j), where "i" is the line number of program "A" and "j" is the line number of program "B". Consequently, the perturbators (3.6, 3.7, 4.5, 4.6, 4.7, 5.4, 5.5, 5.6, 5.7) and the remaining slots are (3.4 ; 3.5; 4.4).

For example, if program "A" is on line 3 and program "B" is on line 5, then the condition of mutual blocking exists because if "A" attempts to move to line 4, then a fault with " B "will be present on line 5; and if "B" tries to move to line 6, then there will be an interference with "A" on line 3. Therefore, no robot can progress. By completing the area of mutual blocking from the "A" lines 3 to 5 and "B" lines 7 to 4, only one program, "A" or "B", can occupy this area at the same time. This is indicated as Area [1] in the sample programs that are listed 402 "New Program A" and listing 404 "New program B" are shown.

8th shows an example in which a "disturbance matrix" 405 can be subdivided into three areas free of mutual blocking: area [1]; Area [2] and area [3]. The disturbance compartments are (3.3, 3.4, 3.5, 3.6, 3.7, 4.5, 4.6, 4.7, 5.4, 5.5, 5.6, 5, 7, 6, 6, 6, 7). Area [1] comprises the compartments (3,3; 3,4; 3,5; 3,6; 3,7), area [2] comprises the compartments (4,5; 4,6; 4,7; 5) , 4, 5.5, 5.6, 5.7) and area [3] comprises the compartments (6, 6, 6, 7). A rising, outwardly bulging envelope is created by the pocket (4, 4), which is a virtual perturbation, so that the three outwardly curved envelopes can be created and the entire face is arched outward. The three areas are subsequently listed 406 "New Program A" and a listing 407 Included "New Program B". There is no case that with any combination of pair pairs free of interference, at least one program can progress to the next row without interference.

9 Figure 12 shows the example of a pair of programs having multiple perturbation areas and two areas of mutual blockage where mutual blockage may be present. A "Program A" writes a route 409 from line 3 to line 15. A "program B" 410 writes a line from line 9 to line 21. The programs have a first section 412 mutual blocking and a second area 411 mutual blocking as it does in the ranges and in a table 413 is shown. A simple method of establishing a single area of mutual blockage is to combine all perturbation areas and intermediate non-perturbation areas into a single area as shown in the table 414 where the region will naturally be free of mutual blockage.

10 shows five processes of the robots 6 . 8th , each of which "Program A" and "Program B" go through. In the first sequence on the left-hand side, the robot goes through 6 the "Program A" without the robot 8th emotional. In the next sequence to the right the robot goes through 8th the "program B" without the robot 6 emotional. In the middle sequence the robots go through 6 . 8th the respective programs at the same time, and a collision occurs. The next sequence to the right shows the robots 6 . 8th that undergo the respective programs with a feature of prior art interference prevention causing mutual blocking simultaneously. The process on the far right shows the robots 6 . 8th which simultaneously pass through the respective programs with mutual blocking prevention according to the invention, so that both programs are completed without interference.

The system and method described above will automatically add and / or modify program instructions to both prevent collisions and to pass potential deadlocks. In some cases, the automatic wait command that is issued may cause the total cycle time for a given task to increase. If increased cycle time is unacceptable, the system and method of the present invention may recommend or automatically generate recurring routes that will allow the robots to cycle through the programs without interruption. In other words, if there is a section of the routes where two robots will occupy the same space at the same time, the system and method will automatically recommend a new route sequence (welding the points in a different order) so that this condition is excluded. This feature of re-running the program instructions has reserves for use in car paint applications.

If non-binding execution instructions occur in the program, then the analysis may contain all combinations of execution execution. Or, for simplicity, the program segments included in the non-binding branches could be treated as independent. Typically, the robot programmer will show such non-binding areas so that the entry and exit distances to the area will not interfere with other robots and thus the independent treatment may be convenient. However, if the entry or exit routes to the non-binding areas interfere with other robots, then the multiple combinations of non-binding and binding areas should be analyzed to ensure no interference and no deadlock.

One of the advantages of this method is that the same mechanism that can preclude conditions of mutual blocking can also preclude interference during normal program motion. Thus, the real-time inspection of a robot-to-robot disturbance during a normal program movement can be excluded. This can save significant CPU utilization, allowing the use of processing time for other purposes such as a faster interpolation rate or shorter interpolation time. In fault conditions or if the program is not running, a reduced CPU usage of the fault check, as disclosed in pending US patent application Ser. No. No. 12 / 124,430 entitled "Fault check of a multi-arm robotic system via three-dimensional automatic ranges".

The area occupied by a program may be a function of physical space, time, program flow, line count, rate of completion of a program or line, or state change related to the robot, program, or system. The perturbation area and the area of mutual blockage are correlated by the same function that defines the occupied space.

The way to avoid mutual blocking can be handled in several ways. The easiest way is to broaden the perturbation areas so that they have both mutual blocking and interference according to 9 prevent. Extending the perturbation areas so that there is only one perturbation area for the entire program will guarantee that the system is free of interlocking, but some type of flow control and / or multiple perturbation areas can be used if cycle time is important. Another possibility is to provide an execution process outside the fault areas so that mutual blocking is prevented. This can be done simply by having a first robot wait for a second robot to move to a particular section program execution at which conditions of mutual blockage may exist before the first robot progresses.

There are several ways in which the occupied space can be determined. The simplest way is that the occupied space is purely the cumulative space that is consistently occupied by a robot during program execution. If several robots have their occupied space determined in this way, then the disturbance area is simply the space which is the intersection of the cumulative occupied space for each robot. Allowing only one robot to enter this space at a time prevents both interference and deadlock.

Another possibility is to determine the occupied space as a function of time. At any given moment, there is a room occupied by each robot. By representing the occupied space as integrating time over a particular interval, the occupied space can be represented as a series of time-correlated discrete spaces. The disturbance areas are the intersections of the robots at the same time or time interval. A disturbance can be avoided by allowing only one robot to be in a fault area at a time. Because the relative time changes due to scheduling, the time base must be set to correlate the new time after scheduling has been applied.

For the above time-based method, the time can be represented as a time interval number. In this way, the total elapsed time becomes less important and the disturbance area can be represented as the matrix of interval numbers. If there is a disturbance in a specific interval state, then this state must not occur because a collision may take place in this interval state.

Similarly, the program flow can be represented as line numbers, the proportion of the sequence, program state, etc. For each such flowchart, there may be a defined fault representation as a function of the selected parameter. The goal to avoid interference is to avoid the states of collection of functions where a disturbance exists. For functions that may be represented as sequential discrete intervals or sequential states, this may be referred to as a disturbance matrix according to FIG 8th being represented. The fault matrix contains the fault area. Of course, a matrix is just one way of displaying data, but the matrix can be a more visual representation. In general, the matrix is sparse so any sparse matrix technique would apply to the rendering.

Determination of states such as input-output states or a process or other states may be used to define the regions. As long as the states have a sequential meaning or can be defined to represent a sequential activity associated with robot position and occupied space, they can be used for the parameter representation of the regions of disturbance and mutual blockage.

Once the perturbation matrix has been determined, one method of preventing interlocking is to avoid any "trap" of monotonically increasing a purging count where perturbation would block forward-following. By making the perturbation matrix increasingly convex, both mutual blocking and perturbation are avoided. The increasing convex perturbation matrix may be referred to as the mutual block matrix. The mutual blockage matrix contains the area of mutual blocking as in 7 is shown. Like the disorder, the matrix of mutual interference is just one way to represent the data.

There are many ways to avoid mutual blocking and interference once the matrix of mutual blockage has been determined. The simplest means is to insert instructions in the execution timing control program, thus avoiding the area of mutual blockage of the interlocking matrix. The addition of instructions to the program has the advantage that the exact nature of the scheduling of the execution can be determined simply by looking at the program. Other means allow the timing to be determined outside the current program modification.

The parametric definition of the regions of perturbation and mutual blocking provides a general method for avoiding perturbation and mutual blocking. This representation of the areas allows for a diverse program control by various factors.

To reduce CPU usage, the current programs may also be updated to include scope and expiration information, so that all necessary information to avoid mutual blocking conditions is included in the programs. This eliminates the need for a real-time check of the Program sequences, disruptions and potential conditions of mutual blocking.

11 Figure 3 is the flow chart of a process for preventing blockage according to the system and method of the invention. In one step 420 become the robot control units 16 . 18 taught the programs. In one step 421 a occupied room is set. Then, trouble areas become one step 422 certainly. In one step 423 potential areas of mutual blocking are defined. In one step 424 requirements of program execution are determined to avoid mutual blocking. Finally, in one step 425 provided a mechanism for preventing mutual blocking, as described above.

The step 423 is in 12 shown in more detail. In one step 426 Both the occupied room and all the trouble areas are identified. In one step 427 every combination of fault areas is checked. In a decision step 428 a check is made for "condition of mutual blockage". If the result is "NO", the step becomes 423 in one step 429 completed. If the result is "YES", a step is effected 430 a means to prevent mutual blocking from occurring and the method returns to the step 427 back.

While certain typical embodiments and details have been shown for the purpose of illustrating the invention, it will be apparent to those skilled in the art that various changes may be made without departing from the spirit of the disclosure which is further described in the following appended claims.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 5150452 [0010]
• US 5347459 [0011]
• US 7114157 [0014]

Claims (15)

A method for preventing mutual blocking of a pair of robots having a common work area, each of the robots being controlled by an associated program, wherein when the programs are executed simultaneously, each of the robots covers at least a portion of the common work area during a part of the execution occupied by the associated program, the method comprising the steps of: while concurrently executing the programs, identifying portions of the shared workspace occupied by the robots; Identifying at least one fault area in which the portions of the common work area overlap; Analyzing the at least one fault area and identifying where at least one condition of mutual blockage of the two robots may occur; and Avoiding the at least one condition of mutual blocking during execution of the programs by automatically determining and executing at least one instruction of mutual blocking free movement prior to performing any movement of the robots resulting in the at least one condition of mutual blocking. The method of claim 1, comprising performing determination and execution during normal production execution of the programs. The method of claim 1, wherein the at least one instruction of the mutual blocking free movement is added to at least one of the programs. The method of claim 1, wherein the at least one mutual block free motion instruction is provided by modifying at least one operation in at least one of the programs. The method of claim 1, wherein the at least one instruction of mutual blocking free movement is provided by re-sequencing operations in at least one of the programs. The method of claim 1, wherein the at least one instruction of mutual blocking free movement is provided by controlling execution of at least one of the programs independently of operations of the program in the at least one program. The method of claim 1, including displaying the occupied portions of the shared work area with voxel models and comparing the voxel models to characterize the at least one perturbation area. Computer-readable media containing operations executable by a computer to perform a method of preventing mutual locking of a pair of robots having a common work area, each of the robots being controlled by an associated program, wherein if the programs simultaneously, each of the robots occupies at least a portion of the shared work area during an execution portion of the associated program, the method comprising: while concurrently executing the programs, the operations identifying portions of the shared workspace occupied by the robots; the instructions identifying at least one region of disturbance in which the portions of the common work area overlap; the instructions that analyze and identify the at least one region of disturbance where at least one condition of mutual blockage of the two robots may occur; and the instructions that prevent the at least one condition of mutual blocking during execution of the programs by automatically determining and executing at least one instruction of mutual blocking free movement prior to performing any movement of the robots resulting in the at least one condition of mutual blocking. The method of claim 8, including the steps of determining and executing during normal production execution of the programs. The method of claim 8, wherein the at least one instruction is added to a block-free motion to at least one of the programs. The method of claim 8, wherein the at least one mutual block free motion instruction is provided by modifying at least one operation in at least one of the programs. A method as claimed in claim 8, wherein the at least one instruction of mutual blocking free movement is provided by re-sequencing operations in at least one of the programs. The method of claim 8, wherein the at least one mutual-blocking-movement instruction is provided by controlling execution of at least one of the programs independently of program operations in the at least one program. The method of claim 8, comprising the steps of presenting the occupied portions of the shared workspace with voxel models and comparing the voxel models to characterize the at least one perturbation region. A method for preventing mutual blocking of a plurality of robots having a common work area, each of the robots being controlled by an associated program, wherein when the programs are executed simultaneously, each of the robots maintains at least a portion of the shared work area during an execution section of the associated work area Program, the procedure comprising the steps of: while concurrently executing the programs, identifying portions of the shared workspace occupied by the robots; Characterizing at least one fault area in which the portions of the common work area overlap for at least two of the robots; Analyzing the at least one fault area and identifying where at least one mutual lock condition of the at least two robots may occur; and Avoiding the at least one condition of mutual blocking during execution of the programs by automatically determining and executing at least one instruction of mutual blocking free movement prior to performing any movement of the at least two robots resulting in the at least one mutual blocking condition.

Images