阅读说明：本技术 诊断方法和设备 (Diagnostic method and apparatus ) 是由 M·阿查里亚 B·菲德勒 B·克拉珀 K·塞弗林 于 2018-07-13 设计创作，主要内容包括：一种用于在包括一个或多个机器人(1a,1b,2a,2b,...)的机器人系统中进行问题诊断的方法,所述方法包括以下步骤：a)从所述机器人系统的机器人接收(S1)第一问题消息,问题消息包括描述机器人经历的问题的一个或多个数据元素；b)从所述机器人系统的机器人接收(S1)后续问题消息；c)如果接收到所述后续消息与接收到紧接在前的消息之间经过的时间短于预定阈值(S2),则将所述后续消息添加到包括所述紧接在前的消息的消息集合中(S3)；d)如果经过的时间长于所述预定阈值(S2),则终止(S4)紧接在前的消息的消息集合而不添加后续消息,并建立(S5、S6)新的消息集合,其中,所述后续消息是第一消息；e)在终止的消息集合中,标识(S7)至少一个描述性数据元素,所述至少一个描述性数据元素在集合的消息中是频繁的或者在集合的第一消息中出现；以及f)将所述至少一个描述性数据元素作为名称分配(S8)给所述终止的消息集合。(A method for problem diagnosis in a robotic system comprising one or more robots (1 a, 1b, 2a, 2 b.), the method comprising the steps of: a) receiving (S1) a first question message from a robot of the robotic system, the question message comprising one or more data elements describing a question experienced by the robot; b) receiving (S1) a subsequent question message from a robot of the robotic system; c) adding the subsequent message to a set of messages including an immediately preceding message if an elapsed time between receipt of the subsequent message and receipt of the immediately preceding message is less than a predetermined threshold (S2) (S3); d) if the elapsed time is longer than the predetermined threshold (S2), terminating (S4) the message set of the immediately preceding message without adding a subsequent message, and establishing (S5, S6) a new message set, wherein the subsequent message is a first message; e) identifying (S7), in a terminated set of messages, at least one descriptive data element that is frequent in a message of the set or that occurs in a first message of the set; and f) assigning (S8) the at least one descriptive data element as a name to the terminated set of messages.)

1. A method for problem diagnosis in a robotic system comprising one or more robots (1 a, 1b, 2a, 2 b.), the method comprising the steps of:

a) receiving (S1) a first question message from a robot of the robotic system, the question message comprising one or more data elements describing a question experienced by the robot;

b) receiving (S1) a subsequent question message from a robot of the robotic system;

c) adding the subsequent message to a set of messages including an immediately preceding message if an elapsed time between receipt of the subsequent message and receipt of the immediately preceding message is less than a predetermined threshold (S2) (S3);

d) if the elapsed time is longer than the predetermined threshold (S2), terminating (S4) the message set of the immediately preceding message without adding the subsequent message, and establishing (S5, S6) a new message set, wherein the subsequent message is the first message;

e) identifying (S7), in a terminated set of messages, at least one descriptive data element that is frequent in the messages of the set or that occurs in a first message of the set; and

f) assigning (S8) the at least one descriptive data element as a name to the terminated set of messages.

2. The method of claim 1, further comprising the steps of: outputting (S9) a list of names of the terminated message sets.

3. The method of claim 2, wherein the outputting step (S9) includes: -outputting said names on a screen (15) and associating a hyperlink (25) with each name, by means of which hyperlink (25) the message with the set of said names can be accessed.

4. Method according to one of the preceding claims, further comprising the step of: those sets whose names include the same descriptive data elements are identified in the plurality of terminated sets of messages, and symbols (26) associated with the sets are displayed on the screen (15) such that the distance between symbols of sets having the same descriptive data elements is less than the distance between symbols of sets having different descriptive data elements.

5. The method of claim 4, wherein the symbol is displayed in a two-axis diagram, wherein one of the axes represents time and the distance is a distance measured perpendicular to the one axis.

6. An apparatus for diagnosing problems in a robotic system comprising one or more robots (1 a, 1b, 2a, 2 b.), the apparatus (10) comprising:

-an interface (11) for receiving problem messages from the robots (1 a, 1b, 2a, 2b,) of the system;

-processing means (13) for: adding a subsequent message to a set of messages including an immediately preceding message if an elapsed time between receipt of the subsequent message and receipt of the immediately preceding message is less than a predetermined threshold, or terminating the set of messages of the immediately preceding message without adding the subsequent message and establishing a new set of messages if the elapsed time is longer than the predetermined threshold, wherein the subsequent message is the first message;

for identifying at least one descriptive data element in a terminated set of messages, said at least one descriptive data element being frequent in said messages of said set; and

for assigning the at least one descriptive data element as a name to the terminated set of messages.

7. The apparatus of claim 6, further comprising: a screen (15) for outputting thereon the names (21) assigned to the terminated message sets or symbols (26) associated with the terminated message sets; and an indicating instrument (16) for selecting one of said names (21) or a symbol (26) output on said screen.

Technical Field

The present invention relates to a method for efficiently diagnosing the origin of technical problems in a robotic system comprising one or more robots.

Background

Diagnosing the origin of technical problems in a multi-robot system, such as a manufacturing plant, where robots perform successive steps in the manufacture of a product, can be a very difficult and time consuming task, since a problem that manifests itself in a manufacturing step that is not performed correctly by a given robot may have its cause in a previous manufacturing step that was performed incorrectly without being noticed. Software errors may also cause the robot to malfunction, and defects in the robot are not visible.

To facilitate identification and removal of faults, it is known to collect problem messages from the robots of such systems, so that when a system should actually fail, it is likely that the cause of such failure can be determined from the problem messages generated by the robots. In practice, however, there is a high probability that these problem messages will mostly involve problems unrelated to actual faults, so that the more robots and other message issuing components are present, the larger the number of messages becomes, and the messages must be studied in order to find the cause of a particular technical problem.

Disclosure of Invention

It is an object of the present invention to provide a method and an apparatus which makes it easier and faster to find the cause of a technical problem in a robot system.

This object is achieved by a method for problem diagnosis in a robotic system comprising one or more robots, the method comprising the steps of:

a) receiving a first question message from a robot of the robotic system, the question message including one or more data elements describing a question experienced by the robot;

b) receiving a follow-up question message from a robot of the robotic system;

c) adding the subsequent message to a set of messages comprising an immediately preceding message if the time elapsed between receipt of the subsequent message and receipt of the immediately preceding message is less than a predetermined threshold;

d) if the elapsed time is longer than a predetermined threshold, terminating the message set of the immediately preceding message without adding a subsequent message, and establishing a new message set, wherein the subsequent message is a first message;

e) identifying, in a terminated set of messages, at least one descriptive data element that is frequent in messages of the set; and

f) assigning the at least one descriptive data element as a name to the terminating set of messages.

This method reflects the inventors' findings as follows: problem messages tend to aggregate over time, and problem messages that are generated closely one after another tend to have a common cause. Thus, if such messages generated closely one after another are grouped into a set of messages, and if it is assumed that the messages of such set have a common cause, then searching for the cause of the problem message is facilitated. If the text of the question message is selected appropriately, the common terms in a set of question messages may indicate a common cause, so that if such common terms form a name assigned to such a set, the name may identify the cause of the question or at least give the practitioner a valuable hint of what might be the cause of the question.

Alternatively, a descriptive data element may be selected from the first message of the terminating set of messages for naming the set, which may be based on the following assumptions: the first question message of a set is probably most directly caused by a question that also causes-directly or indirectly-a subsequent message of the set and, therefore, the first message is most likely to clearly name the cause of the wrong message.

If only the names of the sets are output in the list, the amount of data that a human practitioner must examine in order to find or guess the cause of the problem is greatly reduced relative to the conventional case where the practitioner must examine the problem message in this way, so that the practitioner can make a reasonable guess at least in a short time for the cause of the problem.

Confirmation of guesses may still be a detailed check of the problem message. Such a check can be greatly facilitated if the outputting step includes outputting the names on a screen and associating each name with a hyperlink through which messages with a set of names can be accessed.

On a larger time scale, the identification of a recurring (recurring) problem can be facilitated by the step of identifying, among a plurality of terminated sets of messages, those sets whose names include the same descriptive data elements, and displaying on a screen symbols associated with said sets, so that the distance between symbols of sets having the same descriptive data elements is smaller than the distance between symbols of sets having different descriptive data elements. In this way, any symbol cluster on the screen can be easily identified as likely to be linked for a common reason.

In a preferred embodiment, the symbols are displayed in a two-axis diagram (two-axis diagram), where one of the axes represents time and the distance is the distance measured perpendicular to the one axis. Thus, symbols having a common cause are easy to identify because they tend to form a line parallel to the one axis.

The object is also achieved by a device for diagnosing problems in a robotic system, in particular for performing a method as described above, the system comprising one or more robots, the device comprising:

-an interface for receiving problem messages from robots of the system;

-processing means for: adding a subsequent message to a set of messages including an immediately preceding message if an elapsed time between receipt of the subsequent message and receipt of the immediately preceding message is less than a predetermined threshold, or terminating the set of messages of the immediately preceding message without adding the subsequent message and establishing a new set of messages if the elapsed time is longer than the predetermined threshold, wherein the subsequent message is a first message;

-means for identifying in a terminated set of messages at least one descriptive data element which is frequent in the messages of the set; and

for assigning the at least one descriptive data element as a name to the terminating set of messages.

To facilitate access to the problem message, the apparatus should further comprise: a screen for outputting thereon a name assigned to the termination message set or a symbol associated with the termination message set; and a pointing instrument for selecting one of a name or a symbol output on the screen. Such a pointing instrument may be, for example, a mouse by means of which a user can place a cursor displayed on the screen on a particular set of messages to be opened, or a touch screen superimposed on the screen, which enables the user to select a symbol on the screen by touching it, for example with a finger.

Drawings

Further features and advantages of the invention will become apparent from the subsequent description of embodiments of the invention with reference to the accompanying drawings.

FIG. 1 is a schematic view of a robotic system in which the present invention may be applied;

FIG. 2 is a flow chart of a method of the present invention;

FIG. 3 is a schematic screenshot of a display screen of the apparatus for diagnosing problems shown in FIG. 1; and

fig. 4 is a second schematic screenshot.

Detailed Description

Fig. 1 shows a robot system and a diagnosis apparatus embodying the present invention, which is applied to the robot system. The robot system includes robots such as la, 2a, 3a,. that sequentially perform successive processing steps on the same workpiece 5, and robots such as la, lb,. that perform the same processing steps in parallel on different workpieces 5. Each robot includes sensors for detecting one or more operating parameters, the values of which may indicate a possible failure of the system, and for generating a problem message if one of the operating parameters is out of an expected range.

The diagnostic device 10 comprises an interface 11 for receiving problem messages from sensors of the robotic system, a timer 12, a central processor 13, a storage means 14 for storing the received problem messages and timing information associated with each problem message, a display screen 15 and an indicating instrument 16. The pointing device 16 may be a mouse, as shown in FIG. 1, a touch sensitive layer on the display screen 15, or the like.

Fig. 2 is a flow chart of a method performed by the diagnostic device 10. The method may be performed online when a problem message is received from a sensor, or offline for processing a problem message retrieved from the storage device 14. In the online mode, when a problem message is received from the sensor in step S1, the processor 13 associates it with a time stamp from the timer 12. In the offline mode, the problem message and the timestamp associated therewith retrieve the earlier message from the storage 14. In step S2, the processor 13 compares the time stamp with the time stamp of the immediately preceding question message. If the timestamps differ by less than a predetermined threshold (that is, a predetermined threshold between 1 and 10 minutes), then they are determined to belong to the same message set, and the identifiers of the message set are stored in the storage device 14 in association with the question message (S3). If they differ by more than a threshold, then the set to which the immediately preceding problem message belongs is assumed to be complete; accordingly, the message set is terminated (S4), a new message set identifier is created (S5) and stored in the storage device 14 in association with the question message (S6). In this way, all problem messages that have been collected up to this moment are assigned to one message set at any moment.

The question messages stored in the storage means 14 preferably each comprise an identifier of the robot or the sensor from which they originate and a code associated with a particular type of question.

In step S7, the processor 13 identifies the most frequent code (or a small number of the most frequent codes) and the most frequent robot identifier in each set of messages.

In step S8, each set of messages is assigned a name that includes the most frequent code (or combination of the most frequent codes). If there is one of the problem messages of the message set that appears significantly more frequently than the other problem messages, then that robot identifier (e.g., "2 a" in the first row of the list of FIG. 3) may be included in the name. If the message set includes only one question message, the name of the message set may be the question message itself.

Ideally, the code is selected such that the name given to the set of messages in step S8 directly identifies the cause of the problem message. In practice, a user trying to find out the cause will in many cases wish to anticipate problem messages in detail and face the problem of identifying, among a large number of stored messages, those messages that are actually relevant to the problem he has observed and is trying to solve.

When the user inputs a request to show question messages, he is shown a list of message sets in step S9, each identified by the name to which it was assigned in step S8. Since this name may not be unique (indeed, if the problem recurs, the set of messages associated with it should have the same name), a number or other unique identifier should be associated with each set of messages. Such a list may appear on the display screen 15, for example as shown in fig. 3: each message set is assigned a row on the display 15 containing the number 20 and name 21 of the message set and the time at which it was generated, i.e. the timestamp 22 of its first question message and preferably also the timestamp 23 of the last message.

Even if the user is not looking for the cause of an existing problem, but is simply performing a routine check, the compact list of message sets may provide him with a quick overview in which system operation may not be optimal and in which problems may arise in the future.

The user may manually view a list of message sets using, for example, the scroll bar 24 and pointing instrument 16, for moving up and down in the list and for selecting a given message therein; preferably, he can also enter filter instructions so that only those sets of messages are shown in the list that meet the filter criteria, e.g. the name of the filter criteria contains a reference to a particular problem observed by the user, or a reference to a particular robot that observed no normal operation.

At least all those rows of the list associated with a message set comprising more than one message comprise a hyperlink 25, which hyperlink 25 can be selected and activated by the user using the indication means 16. When the hyperlink 25 is activated, the question message of the message set is displayed, for example, by: a new window is opened on the screen 15 or by inserting question messages 26 under the row of their respective message sets in the window shown in figure 3.

Fig. 4 shows an alternative display mode by which an overview of a very large number of question messages can be obtained in a short time. Here, a two-dimensional graph is displayed on the screen 15, one of the two axes being a time scale, and the other displaying the name of the message set. Each set of messages is represented as a symbol 26, e.g., a point or line in the figure, whose position in two coordinates is defined by its name and the timestamp of its question message. Symbols may be assigned different colors that indicate the number or frequency of problem messages in the set, or indicate their criticality, i.e., the likelihood that they will cause damage if left unattended.

The symbol 26 may be selected and activated using the pointing instrument 16, the activation causing a list of question messages forming the selected set of messages to be displayed on the screen 15.

Here, two sets of messages named "aaaaaaaa" in fig. 3 appear in the same row; thus, the repeatability problems associated with a set of messages having the same name can be immediately discerned. Furthermore, causal relationships between issues of different message sets are easily identified, as they tend to form a regular pattern in the diagram of fig. 4.

In the diagram of fig. 4, it is not only helpful to place the same named set of messages in the same row parallel to the time axis, but it is also helpful to place messages with similar names in the immediately adjacent rows. This similarity may be due to a name assignment problem common to several robots, or due to several problems encountered by a robot whose identifier appears in the message set name. In either case, there is a high probability of problems having a common cause, which becomes more noticeable to the user due to the juxtaposition of rows.

In the method described above, there may be no time overlap between the two message sets, because even if problem messages originate from different robots, they are concentrated into the same message set if the time between them is short enough. This can make it difficult to identify the cause of a problem in very large systems where two or more simultaneous causes may trigger a problem message. In that case, the system may be divided into a plurality of subsystems and the method may be applied separately to each subsystem, for example, robots 1a, 1b, 1a, 2a, which perform the same task in the system of fig. 1, or robots 1a, 2a associated with the same assembly line 4 a. In practice it is advisable to use different divisions in the subsystems at the same time, because if the division is done by an assembly line it is easy to identify that problem messages originating from robots 1a, 2a and not from robots 1b, 2b may have a cause in the assembly line 4a, whereas if the division is done by an assembly phase, problem messages generated by several robots 2a, 2b of the same phase are easily identified as relating to the common supply of these robots.

List of reference numerals

1a, 1b

2a, 2b

3a, 3b

4a, 4b

5 workpiece

10 diagnostic device

11 interface

12 timer

13 processor

14 storage device

15 display screen

16 indicating instrument

20 number

21 name of

22 time stamp

23 time stamp

24 scroll bar

25 hyperlink

26 symbol