EP3374136A1

EP3374136A1 - Method and computer program for correcting errors in a manipulator system

Info

Publication number: EP3374136A1
Application number: EP16794938.7A
Authority: EP
Inventors: Andreas Sedlmayr
Original assignee: KUKA Deutschland GmbH
Current assignee: KUKA Deutschland GmbH
Priority date: 2015-11-11
Filing date: 2016-11-09
Publication date: 2018-09-19
Also published as: WO2017080650A1; CN108349081B; CN108349081A; KR102671535B1; KR20180081775A; DE102015222168A1; US11065766B2; DE102015222168B4; US20180345497A1

Abstract

The invention relates to a method for correcting errors in a manipulator system, wherein the manipulator system comprises at least one manipulator and is controlled by means of at least one manipulator program, wherein the method comprises the following method steps: • providing at least one manipulator program, wherein the manipulator program comprises several operations; • combining at least two of the operations to form at least one operation structure; • defining at least one placement point (AP1, AP2), wherein the at least one placement point (AP1, AP2) forms the start and/or the end of an operation structure (310); • providing at least one reaction structure (320) and assigning the reaction structure (320) to an operation structure (310), wherein the at least one reaction structure (320) contains reaction operations (R1 to Rn), upon the execution of which, the manipulator program controls the manipulator system such that it is passed into a system state which corresponds to a placement point (AP1, AP2); • executing the manipulator program and, if an error occurs, • executing the reaction structure (320) such that the manipulator system is transferred into a system state which corresponds to a placement point (AP1, AP2).

Description

METHOD AND COMPUTER PROGRAM FOR CORRECTING ERRORS OF A MANIPULATOR SYSTEM

FIELD OF THE INVENTION The invention relates to a method and a computer program for correcting errors of a manipulator system. Such manipulator systems typically comprise at least one manipulator and are by means of a

Controlled manipulator program. If an error occurs in the manipulator system, then the manipulator system can be converted into a system state, which a

Resuming the manipulator program allowed.

background

Typical manipulator systems include at least one manipulator that is configured to physically interact with its environment. For example, such a manipulator may be an industrial robot having at least three movable, freely programmable axes and having an end effector, such as a gripper or a machining tool. such

Manipulator systems are used, for example, in automobile production.

Manipulator systems are typically controlled by means of a

Manipulator program, which determines the system behavior application-specific. The requirement for the programming of manipulator programs has changed, for example, through the use of additional sensor technology, which enables new tasks in robotics. For example, new functionalities can be programmed through the use of force and / or torque sensors, such as those used in the industrial robot LBR iiwa from KUKA AG. These include, among other things, search trips with sensors, force-controlled tool operation, MRK (human-robot cooperation) capability, sensitive gripping etc.

Manipulators implement these functions by means of programmed operations. A manipulator program typically includes several operations. The sequence of several operations forms a process. During the

Running a process may cause an error that prevents it from continuing on schedule. The reason for the occurrence of an error may be, for example

Programming errors (e.g., due to incorrectly set parameter values) or occurrence

CONFIRMATION COPY an unforeseen event to which the manipulator system can not respond on its own or requires an interruption of the planned process.

If an error occurs, it is basically possible to continue with a new, changed plan or to undo the process. Some processes can be completely undone and redone. Other processes are irreversible processes and can not be undone; however, knowing the nature of the defect, the process begun and aborted can be "repaired." However, it is often necessary to have the manipulator program

abort and restart, generating high restart times and costs.

An example of an irreversible process is welding by means of a

Manipulator. If the welding process is interrupted, the process can not simply be repeated, since the already welded seam could be damaged by the renewed (over) welding. However, the process could be continued from the point of interruption after elimination of an error that has occurred.

Consequently, methods are needed which allow for an (automated) continuation of started processes, a retry and / or a plan change in the process. If such methods are to be solved with conventional programming techniques, the complexity is additionally increased, whereby the number of

Programming error increases.

The programming of manipulator programs is usually done today

Domain-specific programming languages based on BASIC or PASCAL. These programming languages are usually executed by an interpreter. A large part of the programming is the debugging and process optimization, which following the programming of the regular process (i.e.

Manipulator program) takes place. Debugging and correcting require significant time and human resources, which results in high costs. Furthermore, threatening high damage, if an unrecognized error, for example, leads to an interruption of a series production. In addition, manipulator programs can be programmed using a general high-level language (eg Java). The program code is typically compiled and executed by a programming language bytecode interpreter (Java VM). However, the real-time capability of this kind of interpreter is limited, so the System time of the manipulator program and the manipulator system may differ. As a result, runtime errors can not be responded in real time. Furthermore, for example, the Java VM is typically not equipped for the special requirements of manipulator programming. In particular, among other things, the execution of the manipulator program from a certain point in the expiration of

Manipulator program (program entry), the jumping to a certain point in the flow of the manipulator program during the execution of the

Manipulator program as well as stopping and resuming at a defined point in the flow of the manipulator program is not supported. This makes it difficult to correct errors.

In order to reduce the likelihood of errors, automobile manufacturers and system integrators have their own guidelines (so-called coding standards, for example "VW standard", "Daimler standard", etc.) on the basis of the specific (manipulator)

Programming language developed. For example, such guidelines specify to systematically deposit an error handling / response for typical application errors. Thus, the programmer is given as program code to be written.

Such error handling or error responses should allow, after correcting the error, to continue the manipulator program as planned. Command structures used for this purpose, such as "GOTO" commands, "IF-ELSE" commands or other branch and branch commands, however, are not very clear, which leads to an increase in the complexity of the manipulator programs. Furthermore, at the

conventional error handling is often presupposed that any performed operation of the manipulator program must be reset by itself and in the correct order. This is not possible by means of conventional interpreters, such as Java VM. In particular, irreversible processes are not suitable for automated reverse processing of the individual operations.

The object of the present invention is to provide a method and a computer program as well as a device which can at least partially eliminate the disadvantages described above. Detailed description of the invention

The object of the invention is achieved by a method for correcting errors of a manipulator system according to claim 1, a computer program according to claim 14 and a device according to claim 16. In particular, the object is achieved by a method for correcting errors of a manipulator system, wherein the manipulator system at least one

Manipulator and is controlled by means of at least one manipulator program, the method comprising the following steps:

Providing at least one manipulator program, wherein the

Manipulator program includes multiple operations;

• Grouping at least two of the operations into at least one

Operation structure;

Defining at least one touchdown point, wherein the at least one

Touchdown point forms the beginning and / or end of an operation structure; Provide at least one reaction structure and assign the

Reaction structure to an operation structure, wherein the at least one

Reaction structure Reaction operations, in whose execution the manipulator program controls the manipulator system so that it into a

System state is performed, which corresponds to a touchdown point;

· Execute the manipulator program and if an error occurs

• Execute the reaction structure so that the manipulator system into a

System state that corresponds to a touchdown point.

Manipulator programs are used to control the manipulator system. For example, a manipulator program can specify movements and working steps of a manipulator of the manipulator system. For example, it is possible for a manipulator to be moved according to a defined trajectory in order to grasp an object. Subsequently, the object can be moved according to a second defined trajectory and stored at another position. Likewise, manipulator programs may include instructions related to mounting parts, welding, riveting, or other tasks.

The manipulator program for this purpose has a plurality of operations, wherein a plurality of operations are sequenced to a process to control the manipulator system and, for example, to perform the above tasks. there At least two of the operations become an operational structure

summarized. Typically, related operations, such as interdependent operations or functionally related operations, are combined into one operation structure. This makes the manipulator program clearer and easier to master.

If several operation structures are executed in succession, the previous operation structure is completed in each case. However, a single first operation may start a second operation that extends beyond the execution time of the first operation. In this case, the context of the operation structure must be preserved until the last operation of the operation structure is completed. Thus, the surgical structures do not affect each other. Consequently, parallel operation structures lead to the same results as serially executed operation structures. Furthermore, a change of state of a performed operation structure is permanent.

The beginning and / or the end of an operation structure can form a touchdown point. From a touchdown point, the manipulator program can be continued regardless of its execution history. The same results are always achieved.

If an error occurs that is based on a programming error or if it is due to an unforeseen event, then by means of a reaction structure such an attachment point can be reached and the manipulator program can continue from this. The touchdown point does not have to correspond to the last touchdown point. Likewise, from a touchdown point, a different operation structure can be executed, as the one in which the error has occurred. Thus, a started process can be corrected or continued with a new plan without having to abort or restart the manipulator program.

Furthermore, by the direct allocation of operational structure and

Reaction structure allows a clear manipulator program, which errors in programming can be reduced. Furthermore, the

Troubleshooting simplified. Preferably, the operation structure is consistent, so that the integrity conditions of the manipulator system are met at a touchdown point and / or before and / or after executing the operation structure. This ensures that a manipulator program, regardless of its execution history of a

Touchdown point can be continued. The same results are always achieved.

If an error occurs during the execution of the manipulator program, which is caused, for example, by a programming error or by an unforeseen event in the manipulator system, then a reaction structure can be carried out which brings the manipulator system into a system state corresponding to a touchdown point. Not only is the software "rewound", but the manipulator system is actually transferred to a system state that corresponds to the touchdown point.

The reaction structure preferably comprises reaction operations which can undo individual operations if the respective operation structure describes a reversible process. If the respective operation structure describes an irreversible process, then reaction operations may be contained in the reaction structure which allow it to continue with a modified plan, i. which differ from a pure reverse execution performed operations.

For example, when placing a rivet by means of a manipulator, the rivet may jam in a bore. In this case correct riveting is not possible. If the manipulator system detects this error, the operation structure "riveting" can be broken off and an associated reaction structure initiated.A possible reaction structure in this example could include opening the riveting tool and then commanding the jammed rivet out of the hole a new rivet will be taken, and the operations structure "riveting" will be performed again.

Other reaction structures are also possible. For example, depending on the number of failed attempts to execute an operation structure, a response structure may command a different approach. To stay in the previous example, after repeatedly performing the

Operation Structure "Riveting" which leads to the same error "Jammed Rivet" another procedure can be commanded and an operator called. This could be prompted to manually remove the jammed rivet.

An operation structure and an associated reaction structure are preferably implemented in a common semantic module. Thus, in the

Manipulator program a fixed assignment of operations structure and

Reaction structure ensured, whereby the complexity is reduced.

In addition to undoing an operation structure, a reaction structure can also cancel the operation structure and / or continue the operation

Command manipulator program from another touchdown point.

Preferably, the system state of the manipulator system at a touchdown point and / or before and / or after executing an operation structure is not consistent in all system parameters. However, there must be consistency in the essential parameters, so that resuming or repeating the

Manipulator program is possible from this Aufsetzpunkt. Such non-consistent system parameters may be due, for example, to irreversible processes. For example, if a welding process using the

Controlled manipulator program, so the parameter "weld length"

irreversible because, for example, a certain part of the weld has already been welded until the fault has occurred. If an error occurs, the

Reaction structure does not produce the same system state as before

Execution of the welding process. Therefore, in this case, the reaction structure must provide a corresponding reaction operation, which makes it possible to eliminate the causative error of the welding process (eg cleaning of the welding process)

Welding tool) and then continue at the point of the weld at which the weld was interrupted. It is also possible that in this case, the reaction structure provides an alternative approach, after which the interrupted weld is not finished welded, but to be reworked in a manual step. The manipulator program can be continued elsewhere in this case. Preferably, the execution of an operation structure upon the occurrence of a

Error interrupted and then continued with the execution of a reaction structure. Thus, an operation structure of the manipulator program can be repeated immediately after the occurrence of an error or

Manipulator program continue from another touchdown point. This allows to minimize downtime of the manipulator system and

preferably automated, d. H. without an operator intervention, continue. Thus, the productivity of the manipulator system can be increased.

Particularly preferably, not every operation of an operation structure is assigned its own reaction operation to a corresponding reaction structure. Likewise, not every operation structure must have its own reaction structure.

Preferably, an operation structure is assigned a reaction structure, wherein the reaction structure is preferably assigned to at least one further operation structure. Thus, with a reduced number of reaction structures, it is possible to react to errors of different operational structures. Preferably, a reaction structure is assigned to a plurality of operation structures if the operation structures have a common touchdown point.

Preferably, the reaction structure includes at least one reaction operation whose execution leads the manipulator system to a system state corresponding to the touchdown point, which touchdown point forms the beginning of the operation structure in which the failure occurred. Such reaction structures allow the manipulator system to be "rewound." Thus, the operation structure in which the error occurred can be repeated, and the return to the touchdown point does not have to correspond to a direct inversion of the operation structure, but can be performed in an alternative way For example, a movement of a manipulator are reversed and the direct return path is blocked, so the manipulator can be traced on an alternative trajectory to the starting point of the movement.

Preferably, the at least one reaction operation undoes an operation of the operation structure. Thus, individual operations can be reversed immediately. This is especially true for reversible processes

advantageous. For example, a movement performed by the manipulator can be immediately reduced.

Preferably, an operation of the manipulator program is defined by at least one parameter, wherein the at least one parameter is defined by a

Reaction operation of the reaction structure is changeable. A parameter may be a numerical parameter, such as a manipulator speed, or an instruction parameter representing the sequence of operations of a

Operation structure or the selection of the subsequent operation structure affected. Thus, a response operation may cause the operation structure to continue with changed parameters (eg, slower manipulator speed) or repeat with changed parameters. It is also possible to achieve an alternative branching of the manipulator program. Preferably, the reaction structure includes at least one reaction operation whose execution continues the interrupted operation structure with at least one changed parameter or executes the interrupted operation structure again. Thus, it is possible to have operational structures in which an error has occurred again or, if necessary, with changed, preferably optimized ones

To execute parameters.

The reaction structure preferably includes at least one reaction operation whose execution interrupts the interrupted operation structure until operator input has taken place. This makes it possible, if the occurred error can not be corrected automatically by the manipulator system, to request the intervention of an operator. For example, the operator may be requested to check a certain system state, to remove or replace defective parts or workpieces from the manipulator system, and / or the like.

If the operator confirms to the manipulator system a successful execution of the task, it is possible to continue with the reaction structure or an operation structure.

Preferably, depending on the type of error that has occurred,

carried out different reaction operations of the reaction structure. This is advantageous because it allows you to respond individually to different types of errors. Preferably, different reaction structures are performed depending on the error that has occurred. For this purpose, an operational structure is preferably assigned a plurality of reaction structures, which are executed as a function of the error that has occurred.

For example, if an operation structure is aborted as a

If the security mechanism of the manipulator system intervenes, then the operation structure may simply be repeated if the reason for the intervention of the security device was a temporary one. On the other hand, if an error occurs that can not be remedied automatically, an operator may

be summoned. It is also possible that an error requires continuing at another touchdown point of the manipulator program. For example, a Skipping operation structure and continue the manipulator program elsewhere (see above example, according to which a broken weld is not finished). Thus, an erroneously executed operation of an operation structure and / or an erroneous executed

Operations structure follow a corresponding reaction structure of the manipulator program. This allows error-dependent, differential error correction and continuation of the manipulator program from a suitable touchdown point.

Furthermore, parameters can preferably be changed differently depending on the error that has occurred. For example, repeating a

Operation structure with reduced speed of the manipulator are performed when a first error occurs or with a changed force threshold of force monitoring of the manipulator when a second error occurs. Other changed parameters are also conceivable.

Preferably, the at least one operation structure is linked in a touchdown point with at least one further operation and / or operation structure, wherein after reaching the touchdown point, one of the linked operations and / or operation structures is executed. In this case, the touchdown on a

Operation structure (i.e., forward) or achieved via a reaction structure (i.e., backward). A combination of several operation structures with a touchdown point makes it possible to implement different program sequences by the branching of the manipulator program. Thus, after reaching a touchdown point with different

Operations structures / operations are continued. This allows flexible error correction. Likewise, an operation structure may preferably be executed with changed parameters, depending on the error that has occurred.

Preferably, a parameter is a numerical parameter or a

Statement parameters. Numeric parameters are parameters that are expressed in numerical values and correspond to physical quantities. These are, for example, speeds of the manipulator, force thresholds, momentary thresholds, waiting times and the like. Instruction parameters influence the course of the

Manipulator program and specify, for example, the execution order of the operation structures or individual operations.

Preferably, the manipulator program includes real-time operations and

Real-time reaction operations as well as non-real-time operations and non-real-time Reaction operations, where before running a real-time operation all

Real-time reaction operations and non-real-time reaction operations are completed. By determining non real time and real time operations, a deterministic behavior of the

Manipulator program can be achieved. Should a real-time operation or

Reaction operation, it is ensured that the time of the manipulator program and the actual system time match.

On the other hand, if non-real-time operations are executed, the time of the manipulator program and the actual system time may differ. Thus, for example, non-real-time operations or responses may be in the background.

The object is further achieved by a computer program, comprising

Program instructions which, when loaded on a computer and / or microcontroller, cause the computer and / or microcontroller to perform the described method. Such computer programs may be loaded on devices adapted to control manipulator systems. Preferably, an operation structure and an associated

Reaction structure implemented in the same manipulator program part and preferably form a semantic module. As a result, the surgical structures or reaction structures are closely linked, so that the programming can be made clear. Furthermore, preferably

Computer program implemented as part of the manipulator program. Also, the computer program may be a stand-alone computer program.

The object is further achieved by a device which is set up to execute the computer program. Such devices are for example

Control devices of the manipulator system.

Description of the figures

In the following, the invention will be described in detail with reference to FIGS. 1 to 5. 1 shows a method for troubleshooting according to the prior art;

Fig. 2 is a schematic representation of a Semantikbausteins a

Operation structure and a reaction structure; Fig. 3 is a schematic representation of an operational structure and a

Reaction structure of the manipulator program;

4 shows a schematic representation of a model of a manipulator program; and FIG. 5 is a schematic representation of an application of the manipulator program.

In particular, FIG. 1 shows a schematic representation of an error handling of a manipulator program 1 according to a known method. After the execution of a process step Si, S2, S3 it is checked whether the respective process step has been carried out successfully. If this is the case, the following process step S2 can be continued. If an error occurs, the manipulator program jumps to an error handling structure Fi, F2, F3, which terminates the manipulator program. After the error has been corrected, the manipulator program must be restarted. This error handling is inflexible and therefore only partially suitable for manipulator systems. If complicated error cases are also to be covered with conventional methods, the complexity of the manipulator program and thus the probability of errors for possible programming errors increases.

FIG. 2 shows a schematic representation 2 of a semantic block of a

Computer program for carrying out a method for correcting errors in a manipulator system. In this case, the semantic module 200 is divided into an operational structure 210 and a reaction structure 220. Occurs during the

If the operation structure 210 fails, the operation structure 210 is interrupted and the reaction structure 220 is continued.

The reaction structure 220 includes reaction operations 221 to 225. In a first reaction operation 221, the type of failure that has occurred has been determined and a decision has been made as to which of the reaction operations 222-224 to continue. For example, reaction operation 222 may include instructions that include a

Continuation of the operation structure 210 with a changed parameter set command. Reaction operation 223 may include, for example, instructions that at least partially undo the operations of the executed and interrupted operation structure 210. Reaction operation 224 may be a

Abort instruction so that it is not continued with operation structure 210, but with a further operation structure. Reaction operation 225 includes an instruction that specifies the touchdown point and the subsequent operation structure with which the manipulator program should continue. Such a touchdown point, for example, at the beginning of

Operations structure 210 may be defined so that the operation structure 210 may be executed again. The renewed execution of the operation structure 210 can optionally take place with a changed parameter set. Likewise, other touchdown points can be used to continue the manipulator program. Thus, there is a branching and diverse responses to mistakes are possible. Since both the operation structure and the reaction structure in the same

Semantikbaustein are implemented, the manipulator program remains clear and allows an individual adaptation of the operation structure and / or the

Response structure. Programming errors can be fixed quickly and due to the reduced complexity and increased clarity is the

Likelihood of programming errors being degraded.

FIG. 3 shows a schematic representation of an operation structure 310 and a reaction structure 320. The operation structure 310 comprises a plurality of operations Ol to On. These operations are preferably performed sequentially, as indicated by the straight, solid arrows. The operation structure 310 starts at a first touchdown point APi and ends at a second touchdown point AP2. In the attachment points, the manipulator system is preferably consistent. Thus, the operation structure can be independent of a touchdown point

previous execution history and provides the same results.

The operational structure 310 is assigned a reaction structure 320 which comprises a plurality of reaction operations Ri to Rn. In the illustrated case, if an error occurs in an operation Ol to On, with a corresponding reaction operation Ri to Rn, the relevant operation Ol to On is first undone. If this is successful, it can be decided whether the operation Ol to On, in which the error has occurred, should be executed again, or whether by means of further

Reaction operations Ri to Rn further previous operations Ol to On should be undone. The invention is not limited to the embodiment shown herein. In particular, several operations can be reversed by a reaction structure, or a different way than the actual "undo" operation to a previous one

Aufsetzpunkt be trodden. 4 shows a schematic model 4 of a manipulator program comprising a plurality of operations ΟΊ to 0'2i and a plurality of reaction operations R'i to R'30. Starting at touchdown point AP5, the operation ΟΊ can be performed. The predicted program sequence is shown by dashed arrows. The

Manipulator program then performs the operation 0'2. Here is the

Possibility of branching to one of the operations Ο'3-O's. In the example shown, the program flow leads to operation structure C3. Subsequently, with

Operations structure O'io, which is assigned the touchdown point AP7. In the following operation structure, the operations ΟΊι, ΟΊ3, ΟΊ4 and ΟΊ6 are successively executed, and an error 400 occurs in the execution of the operation ΟΊ6. The erroneous operation ΟΊ6 is aborted and with the

Reaction operation R'i continued.

The reaction operation R'i provides two possible approaches. The first leads to reaction operation R'io the second to reaction operation R'20, which increases

Touchdown point AP7 leads. From touchdown point AP7, operation O'io can be continued or returned via operation R'30 to workpoint AP6 associated with operation 0 ^* 2. Thus, different operations with different reaction operations can be reversed or, if processes are irreversible, other paths can be chosen by the manipulator program. A path denotes the sequence of operations and / or

Response operations.

Fig. 5 shows an exemplary example of an application of the method. According to the application, in representation A, a first part 510, which has a

Bayonet lock 511 has a second part 520 aligned. The second part 520 comprises a projection 521 which can engage with the bayonet lock 511. To assemble the two parts 510, 520, a manipulator first has to concentrically align the first part 510 relative to the second part 520. Once this has taken place, the first part 510 can be lowered (see illustration B), so that the projection 521 is inserted into the bayonet closure 511 of the part 510. Subsequently, the first part 510 is rotated by means of the manipulator (illustration C), so that the

Bayonet lock 511 with the projection 521 engages. If, for example, an error occurs when lowering the first part 510, the manipulator program can be made to start again by aligning the two parts 510, 520 by means of a reaction structure. If this fails, ie the error can not be corrected automatically, an operator can be called in, for example, the first one Part 510 correctly positions to the second part 520 or removes the first part 510. Consequently, the assembly process can be continued and an error occurring can be remedied without the manipulator program having to be aborted and restarted.

LIST OF REFERENCE NUMBERS

AP1-AP9 touchdown points

Egg termination step

Fi to F3 error structures

Ol to On operations

ΟΊ-Ο'21 operations

Ri to Rn reaction operations

R'i-R'30 reaction operations

Si to S3 execution steps

200 semantic building block

210; 310 operation structure

220; 320 reaction structure

221, 222, 223,

224, 225 reaction operations

400 errors

610 first part

620 second part

611 bayonet lock

621 advantage

Claims

Claims 1 to 16

Method for correcting errors of a manipulator system, wherein the manipulator system comprises at least one manipulator and is controlled by means of at least one manipulator program, and wherein the method comprises the following method steps:

Providing at least one manipulator program, wherein the

Manipulator program includes several operations (Ol to On);

Grouping at least two of the operations (Ol to On) into at least one operation structure (310);

• defining at least one touchdown point (APi, AP2), wherein the at least one touchdown point (APi, AP2) forms the beginning and / or the end of an operation structure (310);

Providing at least one reaction structure (320) and assigning the reaction structure (320) to an operation structure (310), wherein the at least one reaction structure (320) includes reaction operations (Ri to Rn), in the execution of which the manipulator program controls the manipulator system so that it is brought into a system state corresponding to a touchdown point (APi, AP2);

• executing the manipulator program and if an error occurs

• exports the reaction structure (320), so that the manipulator system is transferred to a system state that corresponds to a touchdown point (APi, AP2).

The method of claim 1, wherein an operational structure (310) is consistent such that the integrity constraints of the manipulator system on a

Touchdown point (APi, AP2) and / or before and / or after the execution of the operation structure (310).

Method according to one of the preceding claims, wherein the

System state of the manipulator system at a touchdown point (APi, AP2) and / or before and / or after executing an operation structure (310) is not consistent in all system parameters.

4. The method of claim 1, wherein the execution of an operation structure is interrupted on the occurrence of a fault, and is then continued with the execution of a reaction structure.

5. The method according to any one of the preceding claims, wherein the

Reaction structure (320) includes at least one reaction operation (Ri to Rn), the execution of which leads the manipulator system in a system state corresponding to the touchdown point (APi), which touchdown point (APi) forms the beginning of the operation structure (310) in which the error occurred is.

6. The method of claim 5, wherein the at least one response operation (Ri to Rn) reverses an operation (Ol to On) of the operation structure (310).

7. The method according to any one of the preceding claims, wherein an operation (Ol to On) of the manipulator program is defined by at least one parameter, and wherein the at least one parameter by a

Reaction operation (Ri to Rn) of the reaction structure (320) is variable.

8. The method according to any one of the preceding claims, wherein the

Reaction structure (320) includes at least one reaction operation (Ri to Rn), the execution of which continues the interrupted operation structure (310) with at least one changed parameter or executes the interrupted operation structure (310) again.

9. The method according to any one of the preceding claims, wherein the

Reaction structure (320) includes at least one reaction operation (Ri to Rn), whose execution interrupts the interrupted operation structure (310) until an operator input is made.

10. The method according to any one of the preceding claims, wherein depending on the nature of the error occurred different reaction operations (Ri to Rn) of the reaction structure (320) are performed.

11. The method according to any one of the preceding claims, wherein the at least one operation structure (310) in a touchdown point (APi, AP2) with at least one further operation (Ol to On) and / or operation structure (310) is linked and after reaching the touchdown point (APi, AP2) one of the linked operations (Ol to On) and / or operation structures (310) is executed.

12. The method according to any one of the preceding claims 7 to 11, wherein a

Parameter is a numeric parameter or an instruction parameter.

13. The method according to any one of the preceding claims, wherein the

Manipulator program involves real-time operations and real-time reaction operations as well as non-real-time operations and non-real-time reaction operations, and where all real-time reaction operations and non-real-time reaction operations are completed prior to the execution of a real-time operation.

A computer program comprising program instructions which, when loaded on a computer and / or microcontroller, cause the computer and / or microcontroller to perform the method of any one of claims 1 to 13.

The computer program of claim 14, wherein an operational structure (310) and an associated reaction structure (320) are therein

Manipulator program part are implemented and preferably one

Form semantic building block (100).

16. Apparatus adapted for executing a computer program

one of claims 14 or 15.