(19)
(11)EP 2 510 443 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
01.05.2019 Bulletin 2019/18

(21)Application number: 10836427.4

(22)Date of filing:  29.11.2010
(51)International Patent Classification (IPC): 
G05B 23/02(2006.01)
G06N 5/04(2006.01)
G05B 9/02(2006.01)
(86)International application number:
PCT/US2010/058200
(87)International publication number:
WO 2011/071709 (16.06.2011 Gazette  2011/24)

(54)

AUTOMATION MANAGEMENT SYSTEM AND METHOD

AUTOMATISIERUNGSVERWALTUNGSSYSTEM UND -VERFAHREN

SYSTÈME ET PROCÉDÉ DE GESTION D'AUTOMATISATION


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 09.12.2009 US 267940 P
26.11.2010 US 954747

(43)Date of publication of application:
17.10.2012 Bulletin 2012/42

(73)Proprietor: Beet, LLC
Plymouth, MI 48178 (US)

(72)Inventors:
  • WANG, David
    Southfield, Michigan 48033 (US)
  • RED, Daisy
    Southfield, Michigan 48033 (US)
  • NAUSLEY, Ivan
    Southfield, Michigan 48033 (US)

(74)Representative: McDonough, Jonathan et al
Hepworth Browne 7 Wragley House
Valley Road Hebden Bridge HX7 7BN
Valley Road Hebden Bridge HX7 7BN (GB)


(56)References cited: : 
US-A- 5 148 363
US-A- 6 023 667
US-A1- 2005 039 086
US-A- 5 949 676
US-A1- 2003 061 546
US-A1- 2005 055 257
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    BACKGROUND



    [0001] In industrial automation facilities, optimal operation of the facility can partially depend on the equipment monitoring system employed therein. Equipment monitoring systems can collect data from, for example, computational devices (programmable logic controllers, programmable automation controller, etc.) in equipment located throughout the facility. The collected data can assist in, for example, monitoring equipment health, predicting when equipment failure will occur and/or informing operators when equipment failure has occurred. A computer interface, such as a human-machine interface (HMI), may be interconnected with the computational devices to facilitate programming or control thereof, to monitor the computational device, or to provide other such functionality.

    [0002] Current approaches of data collection include, for example, utilizing an interface (e.g. OPC server, open connectivity, server) with the HMI or other computer to poll data from the computational devices according to a preset time interval. Such an approach may be less than optimal since a large amount of data is collected and stored even when there may be no activity or change in information. Further, since many times the OPC server is configured based on the configuration of the computational device, modifications to the software present on the computational device can prevent accurate data collection.

    [0003] Other current approaches of data collection include, for example, programming the computational device to transmit data upon detection of a problem using the common industrial protocol (CIP). Although less data will typically be collected and stored using this approach as compared to the above-described polling approach, this is largely a reactive system. Accordingly, for example, data is captured only if the computational device detects an error. Further, simply collecting this error data may not permit an operator to engage in playback (i.e. replicate equipment performance) because complete performance data may not be have been transmitted by the computational device.
    US5148363 (Sakamoto et al) discloses a breakdown diagnosing method which prevents erroneous diagnosis of abnormalities in an operating system, particularly a production line. The fact that an abnormal state of the equipment is not accompanied by a malfunction can be detected. In the event the operating system equipment is actually in the abnormal operating state, an abnormal block and an abnormal operating step in the abnormal operating block of operations executed by the equipment can be promptly and correctly specified.

    SUMMARY



    [0004] Embodiments of a method for monitoring performance of equipment located in an automation environment as defined in claim 1 and of an apparatus for monitoring performance of equipment located in an automation environment as defined in claim 11 are disclosed herein.

    [0005] These and other embodiments are disclosed in additional detail hereafter.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0006] The description herein makes reference to the accompanying drawings wherein like reference numerals refer to like parts throughout the several views, and wherein:

    FIG. 1 is schematic diagram of an automation management system according to one embodiment of the present invention;

    FIG. 2 is a timing diagram of an exemplary design cycle as used in the automation management system of FIG. 1;

    FIGS. 3A-3D are performance diagrams of exemplary actions of the exemplary cycle of FIG. 2; and

    FIG. 4A is a performance data diagram for a cycle as used in the automation management system of FIG. 1;

    FIG. 4B is a machine level performance diagram using the cycle performance data of FIG. 4A;

    FIG. 5A is a performance data diagram for another cycle as used in the automation management system of FIG. 1;

    FIG. 5B is a machine level performance diagram using the cycle performance data of FIG. 4A;

    FIG. 6A is a performance data diagram for another cycle as used in the automation management system of FIG. 1;

    FIG. 6B is a machine level performance diagram using the cycle performance data of FIG. 4A; and

    FIG. 7 is an exemplary flowchart diagram of a prediction routine used in the automation management system of FIG. 1.


    DETAILED DESCRIPTION



    [0007] Referring to FIG. 1, an automation management system 10 includes a PLC 12 and a PC 14. The data that is collected from the PLC 12 can be based on, for example, the programming of the PLC 12, as will be discussed in more detail below. PC 14 can include a data access server, such as an OPC (OLE, Object Linking and Embedding, for Process Control) server, to retrieve data from PLC 12 and to convert the hardware communication protocol used by PLC 12 into the server protocol (e.g. OPC protocol). Although in this embodiment, the data access server is, by way of example, an OPC server other suitable data access servers are available having custom and/or standardized data formats/protocols. The data retrieved by the OPC server can optionally be stored in a database (not shown).

    [0008] Both PLC 12 and PC 14 can have suitable components such as a processor, memory, input/output modules, and programming instructions loaded thereon as desired or required. PLC 12 and PC 14 can be in transmit/receive data through a wired or wireless communication protocol such as RS232, Ethernet, SCADA (Supervisory Control and Data Acquisition). Other suitable communication protocols are available.

    [0009] Although one PLC 12 is illustrated in FIG. 1, more than one PLC 12 may be in communication with PC 14, or other PCs. PLC 12 can be connected to and control any equipment including machines such as clamps or drills. To ease the reader's understanding of the embodiments, the description will refer to machines although the embodiments can be used with equipment other than the machines.

    [0010] The machines can be part of any system including but non-limited to machining, packaging, automated assembly or material handling. PLC 12 is not limited to being connected to a machine and/or can be connected to any other suitable device. Other devices may be used in lieu or in addition to PLC 12 such as PACs (Process Automation Controllers) or DCS (Distributed Control Systems) or any other suitable computational device.

    [0011] The PC 14 can also include a client application to obtain the data from or send commands to PLC 12 through the OPC server. Also, in other embodiments, the client application can be connected to several OPC servers or two or more OPC servers can be connected to share data. Further, for example, PC 14 can include a graphical display representing information such as the status of the machines on the plant floor. In other embodiments, the HMI can also be located as a separate device from the PC 14.

    [0012] Any other device can be used in lieu of or in addition to PC 14. For example, some non-limiting examples include Personal Digital Assistants (PDAs), hand held computers, palm top computers, smart phones, game consoles or any other information processing devices.

    [0013] It is noted that the architecture depicted in FIG. 1 and related description is merely an example and the automation management system 10 described herein can be used with virtually any architecture. For example, modules (e.g. OPC server, client application, etc.) can be distributed across one or more hardware components as desired or required.

    [0014] One use of PLC 12 is to take a machine through a repetitive sequence of one or more operations or tasks. The completion of the repetitive sequence of tasks can be denoted as a cycle. Each task can have, for example, an optimal design start time and design end time in the sequence and resulting duration time ("reference values"). These optimal design times can be based on, for example, the manufacturer's specification or an equipment user's study of when and how long a certain tasks should be executed. The design times can be determined by any other method.

    [0015] Referring to FIG. 2, an exemplary design cycle 30 is illustrated from time t1-t20. The design cycle includes nine tasks 32a-i (collectively referred to as tasks 32). During development, each task 32 is designed to begin operation at a specific start time and designed to end operation at a specific end time. When PLC 12 is operating, PLC 12 can be programmed to collect this start time and end time information and made available to PC 14. For example, the start time and end time information can be sent to PC 14 or PC 14 can periodically poll PLC 12. Accordingly, rather than collecting unnecessary information from PLC 12 (e.g. downtime data), PLC 12 can collect and transmit information that can be used to appropriately monitor and determine a reactive, preventive and/or predictive plan for the machine(s) being controlled. This information can be sent at the end of the cycle 30, during the cycle 30 upon request by the PC 14 or at any other time as desired or required. Thus, as illustrated for cycle 30, advance pin 1 task 32a starts at t0 and ends at t1 and has a duration of t1-t0, advance pin 2 task 32b starts at t0 and ends at t1 and has a duration of t1-t0, close clamp 1 task 32c starts at t1 and ends at t2 and has a duration of t2-t1, close clamp 2 task 32d starts at t1 and ends at t2 and has a duration of t2-t1, weld task 32e starts at t2 and ends at t18 and has a duration of t18-t2, open clamp 1 task 32f starts at t18 and ends at t19 and has a duration of t19-t18, open clamp 2 task 32g starts at t18 and ends at t19 and has a duration of t19-t18, return pin 1 task 32h starts at t19 and ends at t20 and has a duration of t20-t19 and return pin 2 task 32i starts at t19 and ends at t20 and has a duration of t20-t19. This cycle is merely exemplary, and other design cycles are available with different types of tasks, different numbers of tasks and different start and end times for the tasks.

    [0016] PLC 12 generally receives input data, output data and/or other data ("series of signals") from the machines they control. Each task 32 can be defined as a series of one or more input and/or output states. The tasks 32 can be defined, for example, by a programmer during software development for the PLC 12. Thus, for example, PLC 12 can determine whether advance pin 1 task 32a has been complete by examining whether certain inputs and/or outputs or any other condition (e.g. timer) are set to their correct states or values. It is to be understood that the term "states" does not limit embodiments of the invention to digital input and/or signals. In other embodiments, analog inputs or outputs, or a combination of digital and analog inputs or outputs, can be collected from the machines and can be used to define each task 32.

    [0017] Referring to FIGS. 3A-3D, exemplary performance graphs 50, 60, 70 and 80 of task 32a, task 32b, task 32c and tasks 32d are shown, respectively, over one hundred cycles. As illustrated in FIG. 3A, advance pin 1 task 32a is designed to complete its operation within, for example, one second, as indicated by base line 52. The actual operation advance pin 1 task 32a is indicated by a series of plot points 54. Each plot point 54 can represent the actual completion time or duration ("timing value") for task 32a during that particular cycle. As can be seen from the performance graph, the completion time of task 32a is gradually increasing. This can, for example, provide notice to a user that an input and/or an output associated with task 32a is or may in the future experiencing a failure. Accordingly, if desired or required, appropriate maintenance procedures can be undertaken. Although the timing value illustrated in FIGS. 3A-3D, other timing values are available in lieu of or in addition to duration. For example, other timing values include a start time or end time for the task.

    [0018] Similar to that described above in connection with reference to performance graphs 50, advance pin 2 task 32b is designed to complete its operation within one second, as indicated by base line 62 and the actual operation of task 32b is indicated by a series of plot points 64. Close clamp 1 task 32c is designed to complete its operation within one second, as indicated by base line 72 and the actual operation of task 32c is indicated by a series of plot points 74. Close clamp 2 task 32d is designed to complete its operation within one second, as indicated by base line 82 and the actual operation of task 32d is indicated by a series of plot points 84. Since the series of plots points 64, 74 and 84 do not, for example, consistently deviate from the base lines 62, 72 and 82, the user(s) can, for example, ascertain that the tasks are 32b-d are operating normally.

    [0019] As stated above, the performance graphs 50, 60, 70 and 80 are merely exemplary. Other performance graphs may contain other data, depicted using another graph type, or combined into one graph. Further, although 100 cycles are shown, performance graphs can contain any number of cycles.

    [0020] Referring to FIGS. 4A, 5A and 6A are performance data diagrams 100, 120 and 140, respectively. The performance data diagram 100 includes information for a first cycle 102, the performance data diagram 120 includes information for a twentieth cycle 122 and performance data diagram 140 includes information for a hundredth cycle 142. It is to be understood that these cycles are selected from the 100 cycles previously shown but are in no way intended to limit the scope of the embodiments disclosed herein. Rather, selection of these cycles is intended to assist the reader's understanding of embodiments. Other cycles can contain different data.

    [0021] The cycles 102, 122 and 142, as discussed previously, can include tasks 32. For each task 32, performance data can include design cycle data 104, learned cycle data 106, current cycle data 108 ("timing value"), current versus design cycle data 110, current versus learned cycle data 112, accumulated current versus design cycle data 114 and accumulated current versus learned cycle data 114.

    [0022] Design cycle data 104 can include data pertaining to design times for the certain machine performing the tasks and is not based on the series of signals collected from the particular machine being monitored. Thus, as discussed previously, each task may have an expected start time, end time and duration based on for example, a manufacturer's specification or as set by a user. For example, if a manufacturer's specification indicates that the weld task 32e should be performed in 16 seconds, the design cycle time can be 16 seconds. Design cycle times 104 can be determined by other methods. The design cycle times 104 are preferably the same throughout the execution of the tasks 32, although in some embodiments the design cycle can change.

    [0023] Learned cycle time 106 can include data pertaining to a reference time for a certain machine. In other words, learned cycle time 106 is a reference value based on the series of signals collected from the machine. For example, a user can cause the machine to execute a cycle of tasks 32a-i, in order to teach the system the reference data for that particular machine. These learned cycle times can be recorded, for example, during setup of the machine, maintenance of the machine or at any other suitable time. Once the machine has the design cycle time 104 and the learned cycle time 106, the machine can begin operation and can begin collecting current cycle times for each task 32. Current cycle time 108 can be the duration of the time needed to complete each task (i.e. difference between start time and end time). Thus, for example, as shown in FIG. 4A, during the first cycle 102, the welding process lasted 16.073 seconds. During the twentieth cycle 122, the welding process lasted 15.987 seconds. During the hundredth cycle 142, the welding process lasted 15.937 seconds. As discussed previously, the start time and end time and/or the duration for each task 32, can be collected and sent to the PC 14.

    [0024] Once the current cycle times 108 have been collected, the current versus design time can be calculated for each task 32. Current versus design time 110 can be the difference between the design cycle times 104 and the current cycle times 108. Similarly, current versus learned time 112 can be the difference between the learned cycle times 106 and the current cycle times 108. The current versus design time 110 and current versus learned time 112 calculations can be made by PC 14 or by any other module. For example, if the HMI is a separate module than the PC 14, the HMI can perform the calculations.

    [0025] Once the current versus design time 110 has been calculated, a determination can be made as to whether the task 32 has been executed within a threshold value. Thus, for example, if the current versus design time 110 is less than or equal to 10% of the design cycle time 104, then the current cycle has been executed within the acceptable threshold value. The acceptable threshold can indicate that the task is operating normally and a normal operation indicator can be generated (e.g. at PC 14) as will be discussed in more detail below. If the current versus design time 110 is between 10% and 25% of the design cycle time 104, then the current cycle has been executed within a cautionary threshold. The cautionary threshold can indicate that an action may be needed on some input or output associated with that certain task. Similar to the normal operation indicator, a cautionary indicator can be generated that indicates that the current cycle is within the cautionary threshold. If the current versus design time 110 is greater than 25% of the design cycle time 104, then the current cycle has been executed within a warning threshold. The warning threshold can indicate that an action may be needed on some input or output associated with that certain task and a warning indicator can be generated as will be discussed in more detail below. In other embodiments, any number of threshold ranges can be used with different range values. Further, in other embodiments, the current versus learned time 112 instead of the current versus design time 110 can be used to determine whether the task 32 is operating within a predetermined threshold. Other embodiments may contain different calculations to ascertain whether the execution time of the tasks is acceptable. For example, rather than using the design cycle time 104, the learned cycle time can be used to ascertain whether the task has been executed within an acceptable threshold.

    [0026] Once a determination has been made in regard to whether the task 32 has been executed within an acceptable threshold for a particular cycle, this information can be displayed to the user(s). For example, if the tasks 32 have been executed in the acceptable threshold, that particular task can be highlighted, for example, in green ("normal operation indicator"). Similarly, for example, if the tasks 32 have been executed within the cautionary threshold, that particular task can be highlighted in yellow ("cautionary indicator") and if the tasks 32 have been executed within the warning threshold, that particular task can be highlighted in red ("warning indicator"). In other embodiment, the indicators are audible rather than visually displayed to the user. Other techniques for generating indicators are also available.

    [0027] Referring to FIG. 4A, for example, all tasks 32a-i have been executed in the acceptable threshold. Referring to FIG. 5A, for example, tasks 32a-h have been executed in the acceptable threshold and task 32i has been executed within the cautionary threshold. Referring to FIG. 6A, tasks 32b-h have been executed within the acceptable threshold and tasks 32a and 32i have been executed within the warning threshold.

    [0028] Once the current versus design times 110 have been calculated, the accumulated current versus design time 114 ("accumulated variance value") can be calculated for each task 32. The accumulated current versus design time 114 can be the running total or the sum of the current versus design time 110 across some or all cycles (e.g. 100 cycles) for a particular machine run. The machine run may be terminated based on, for example, user(s) intervention, machine breakdown etc. Similarly, once the current versus learned times 106 have been calculated, the accumulated current versus learned time 116 ("accumulated variance value") can be calculated for each task 32. Again, the accumulated current versus design time 116 can be the running total or sum of the current versus learned time 116 across all cycles (e.g. 100 cycles) for a particular machine run. The accumulated current versus learned time 114 and the accumulated current versus design time 116 can be calculated at PC 14, or on any other computational device (e.g. PLC 12).

    [0029] Both the accumulated current versus design times 114 and the accumulated current versus learned times 116 can be graphed on a machine level performance diagram for each cycle and for each task 32 as shown in FIGS. 4B, 5B and 6B. Series1 214 represents the accumulated current versus design time 114 for each task 32 and Series2 216 represents the accumulated current versus learned time 116 for each task. The machine level performance can be displayed, for example, a HMI. As the accumulated current versus design time 114 and/or the accumulated current versus learned time 116 increase, it can alert the user(s) that the particular task may be experiencing a problem that may require an action. For example, as can be seen from FIG. 6B, task 32a has an accumulated current versus design time 114 of 24.356 seconds and an accumulated current versus learned time of 24.656 seconds, which may indicate that advance pin2 task 32a is experiencing a problem. A similar observation can be made in regard to return pin2 task 32i.

    [0030] FIG. 7 is a prediction routine 300 that can be used in automation management system 10. Beginning at block 302, a user enters one or more threshold values for at least one task. For example, a user enters one or more reference values (e.g. design cycle time 104) at block 304. Control then moves to block 306 to collect data or a series of signals from, for example, a machine being controlled PLC 12. PLC 12 can, in turn, send timing values to PC 14 related to the performance of one or more tasks (e.g. tasks 32a-i). In other embodiments, the series of signals can be collected directly from the machine being controlled.

    [0031] At block 308, control moves to determine the variances or differences (e.g. current versus design time 110) between the reference value and one or more of the timing values collected from PLC 12 over a period of time. Design cycle data and/or learned cycle data can be used to determine the differences. The period of time may be any suitable value. As one example, the period of time is 2 seconds. At block 310, control moves to sum the variance to generate an accumulated variance value (e.g. accumulated current versus design time 114).

    [0032] As the variances are being summed, control moves to block 312 to perform trend pattern detection to predict a potential failure of the machine. As one example, trend pattern detection determines a slope of the accumulated variance value. Control then moves to decision block 314 to determine if a pattern has been detected. A pattern can be detected if the slope of the accumulated variance value is either generally increasing or generally decreasing over the period of time rather than being generally constant. When the slope is generally constant, the accumulated variance value does not deviate significantly from zero. In other words, the value of the differences (determined at block 308) are at or close to zero.

    [0033] One example of a detected pattern, is shown in FIG. 3A where the slope of the accumulated variance value is illustrated as generally increasing. In contrast, FIGS. 3B-D show the slopes of their respective accumulated variance values as generally constant. Other suitable techniques of pattern detection are also available that do not include calculation of slope for the accumulated variance value

    [0034] If a pattern has been detected, control can move to step 316 to report the prediction of the potential failure. The prediction can be reported by generating the predictive failure indicator. Otherwise, if no pattern is detected (i.e. the slope is generally constant) the predictive failure indicator is not generated. The predictive failure indicator can be audible or visually displayed to the user at PC 14. Other suitable techniques for generating the predictive failure are also available.

    [0035] The information collected from PLC 12 can be used in all aspects of reactive, predictive and preventive maintenance. In other embodiments, the operation of the tasks 32 can be monitored using any other statistical process control method or any other suitable method. Further, the information related to the executed tasks can be displayed in any other manner using control charts or any other suitable graphical display such as a Gantt chart.

    [0036] Further, embodiments of the present invention are not limited to use in industrial factories, packaging plants etc. For example, embodiments of the present invention can be suitably incorporated into a monitoring and maintenance system of a rollercoaster or any other system that contains a computational device.


    Claims

    1. A method for monitoring performance of equipment located in an automation environment, comprising:

    identifying a cycle (30) associated with the equipment, the cycle (30) defined by a sequence of a plurality of tasks (32);

    collecting, over a plurality of repetitions of the sequence, a series of signals for each of the plurality of tasks (32) in the cycle (30);

    associating, for each of at least some of the plurality of repetitions, one or more signals in the series to define a completion time (54) for each of the plurality of tasks (32);

    comparing, for each of the plurality of tasks (32), each of the completion times (54) to a reference value indicative of an expected completion time of a given task (32);

    generating, for each of the plurality of tasks (32), an accumulated variance value (114) based on the comparisons;

    selectively generating a predictive failure indication based on the generated accumulated variance values (114); and

    if a predictive failure indication is generated, identifying which of the plurality of tasks (32) in the cycle (30) resulted in the predictive failure indication.


     
    2. A method as claimed in claim 1, wherein comparing, for each of the plurality of tasks (32), each of the completion times (54) to the reference value comprises:
    determining, for each of the plurality of tasks (32), a difference between each of the completion times (54) and the reference value.
     
    3. A method as claimed in claim 2, wherein generating the accumulated variance value (114) comprises:
    summing, for each of the plurality of tasks (32), the differences corresponding to at least some of the completion times (54).
     
    4. A method as claimed in claim 1, further comprising:
    determining, for each of the plurality of tasks (32), a slope of the accumulated variance values (114) over a period of time.
     
    5. A method as claimed in claim 4, wherein selectively generating the predictive failure indication further comprises:

    if the slope is at least one of increasing or decreasing over the period of time, generating the predictive failure indicator; and

    if the slope is substantially constant over the period of time, not generating the predictive failure indicator.


     
    6. A method as claimed in claim 1, wherein the reference value for a given task (32) is based on data other than the series of signals.
     
    7. A method as claimed in claim 6, wherein the reference value for a given task (32) is further based on a predefined specification.
     
    8. A method as claimed in claim 1, wherein the reference value for a given task (32) is based on the series of signals.
     
    9. A method as claimed in claim 1, wherein the predictive failure indication is at least one of audible or visual.
     
    10. A method as claimed in claim 1, wherein the series of signals are at least one of analog data or digital data.
     
    11. An apparatus for monitoring performance of equipment located in an automation environment, comprising:

    a memory; and

    a processor configured to execute instructions stored in the memory to:

    identify a cycle (30) associated with the equipment, the cycle (30) defined by a sequence of a plurality of tasks (32);

    collect, over a plurality of repetitions of the sequence, a series of signals for each of the plurality of tasks (32) in the cycle (30);

    associate, for each of at least some of the plurality of repetitions, one or more signals in the series to define a completion time (54) for each of the plurality of tasks (32);

    compare, for each of the plurality of tasks (32), each of the completion times (54) to a reference value indicative of an expected completion time of a given task (32);

    generate, for each of the plurality of tasks (32), an accumulated variance value (114) based on the comparisons;

    selectively generate a predictive failure indication based on the generated accumulated variance values (114); and

    if a predictive failure indication is generated, identify which of the plurality of tasks (32) in the cycle (30) resulted in the predictive failure indication.


     
    12. An apparatus as claimed in claim 11, wherein the processor is further configured to:
    determine, for each of the plurality of tasks (32), a slope of the accumulated variance value (114) over a period of time.
     
    13. An apparatus as claimed in claim 12 wherein the processor is configured to selectively generate the predictive failure indication including:

    if the slope is at least one of increasing or decreasing over the period of time, generate the predictive failure indicator; and

    if the slope is substantially constant over the period of time, not generate the predictive failure indicator.


     


    Ansprüche

    1. Verfahren zur Equipment-Leistungsüberwachung in einer Automatisierungsumgebung, umfassend:

    Identifizieren eines Zyklus (30), der dem Equipment zugeordnet ist, wobei der Zyklus (30) durch eine Sequenz einer Vielzahl von Aufgaben (32) definiert ist;

    Sammeln, über mehrere Wiederholungen der Sequenz, einer Serie von Signalen für jede der mehreren Aufgaben (32) im Zyklus;

    Zuordnen, für jede der wenigsten der Vielzahl von Wiederholungen, eines oder mehrere Signale in der Serie, um eine Fertigstellungszeit (54) für jede der Vielzahl von Aufgaben (32) zu definieren;

    Vergleichen, für jede der Vielzahl von Aufgaben (32), jeder der Fertigstellungszeiten (54) mit einem Referenzwert, der eine erwartete Fertigstellungszeit einer vorgegebenen Aufgabe (32) angibt;

    Erzeugen, für jede der Vielzahl von Aufgaben (32), eines akkumulierten Varianzwertes (114) basierend auf den Vergleichen;

    selektives Erzeugen einer vorausschauenden Fehleranzeige basierend auf den erzeugten akkumulierten Varianzwerten (114); und

    wenn eine vorausschauende Fehleranzeige erzeugt worden ist, Identifizieren, welche der Vielzahl von Aufgaben (32) im Zyklus (30) zu der vorausschauenden Fehleranzeige geführt hat.


     
    2. Verfahren nach Anspruch 1, wobei das Vergleichen, für jede der Vielzahl von Aufgaben (32), jeder der Fertigstellungszeiten (54) mit dem Referenzwert, umfasst:
    Bestimmen, für jede der Vielzahl von Aufgaben (32), einer Differenz zwischen jeder der Fertigstellungszeiten (54) und dem Referenzwert.
     
    3. Verfahren nach Anspruch 2, wobei das Erzeugen des akkumulierten Varianzwertes (114) umfasst:
    Summieren, für jede der Vielzahl von Aufgaben (32), der Differenzen, die mindestens einigen der Fertigstellungszeiten entsprechen (54).
     
    4. Verfahren nach Anspruch 1, weiter umfassend:
    Bestimmen, für jede der Vielzahl von Aufgaben (32), eines Gradienten der akkumulierten Varianzwerte (114) über einen Zeitraum.
     
    5. Verfahren nach Anspruch 4, wobei das selektive Erzeugen einer vorausschauenden Fehleranzeige weiter umfasst:

    wenn der Gradient über den Zeitraum mindestens entweder steigend oder fallend ist, Erzeugen der vorausschauenden Fehleranzeige; und

    wenn der Gradient über den Zeitraum im Wesentlichen konstant ist, Nicht-Erzeugen der vorausschauenden Fehleranzeige.


     
    6. Verfahren nach Anspruch 1, wobei der Referenzwert für eine vorgegebene Aufgabe (32) auf anderen Daten als auf denen der Serie von Signalen basiert.
     
    7. Verfahren nach Anspruch 6, wobei der Referenzwert für eine vorgegebene Aufgabe (32) ausserdem auf einer vordefinierten Spezifikation basiert.
     
    8. Verfahren nach Anspruch 1, wobei der Referenzwert für eine vorgegebene Aufgabe (32) auf den Serien von Signalen basiert.
     
    9. Verfahren nach Anspruch 1, wobei die vorausschauende Fehleranzeige mindestens entweder hörbar oder optisch ist.
     
    10. Verfahren nach Anspruch 1, wobei die Serien von Signalen sind mindestens entweder analoge Daten oder digitale Daten.
     
    11. Vorrichtung zur Equipment-Leistungsüberwachung in einer Automatisierungsumgebung, umfassend:

    einen Speicher; und

    einen Prozessor, der konfiguriert ist, um im Speicher gespeicherte Anweisungen auszuführen, um:

    einen Zyklus (30), der dem Equipment zugeordnet ist, zu identifizieren, wobei der Zyklus (30) durch eine Sequenz einer Vielzahl von Aufgaben (32) definiert ist;

    um, über mehrere Wiederholungen der Sequenz, eine Serie von Signalen, für jede der mehreren Aufgaben (32) im Zyklus (30), zu sammein;

    um, für jede der wenigsten einen Vielzahl von Wiederholungen, ein oder mehrere Signale in der Serie, zuzuordnen, um eine Fertigstellungszeit (54) für jede der Vielzahl von Aufgaben (32) zu definieren;

    um, für jede der Vielzahl von Aufgaben (32), jede der Fertigstellungszeiten (54) mit einem Referenzwert, der eine erwartete Fertigstellungszeit für eine vorgegebene Aufgabe (32) angibt, zu vergleichen;

    um, für jede der Vielzahl von Aufgaben (32), einen, auf den Vergleichen basierenden, akkumulierten Varianzwertes (114) zu erzeugen;

    um, eine, auf den erzeugten akkumulierten Varianzwerten (114) basierende, vorausschauende Fehleranzeige selektiv zu erzeugen; und

    um, wenn eine vorausschauende Fehleranzeige erzeugt worden ist, zu identifizieren, welche der Vielzahl von Aufgaben (32) im Zyklus (30) zu der vorausschauenden Fehleranzeige geführt hat.


     
    12. Vorrichtung nach Anspruch 11, wobei der Prozessor ausserdem konfiguriert ist,
    um:
    für jede der Vielzahl von Aufgaben (32), einen Gradienten des, über einen Zeitraum akkumulierten, Varianzwertes (114), zu bestimmen.
     
    13. Vorrichtung nach Anspruch 12, wobei der Prozessor konfiguriert ist, um die vorausschauende Fehleranzeige selektiv zu erzeugen, einschliesslich:

    um, wenn der Gradient über den Zeitraum mindestens entweder steigend oder fallend ist, die vorausschauende Fehleranzeige zu erzeugen; und

    um, wenn der Gradient über den Zeitraum im Wesentlichen konstant ist, die vorausschauende Fehleranzeige nicht zu erzeugen.


     


    Revendications

    1. Procédé de surveillance de la performance d'un équipement situé dans un environnement d'automatisation comprenant :

    l'identification d'un cycle (30) associé à l'équipement, le cycle (30) étant défini par une séquence d'une pluralité de tâches (32) ;

    la collecte, au cours d'une pluralité de répétitions de la séquence, d'une série de signaux pour chacune de la pluralité de tâches (32) au sein du cycle (30) ;

    l'association, pour chacune d'au moins certaines répétitions de la pluralité des répétitions, d'un ou plusieurs signaux dans la série afin de définir un temps d'accomplissement (54) pour chacune des tâches de la pluralité de tâches (32) ;

    la comparaison, pour chacune des tâches de la pluralité de tâches (32), de chaque temps d'accomplissement (54) par rapport à une valeur de référence indiquant un temps d'accomplissement attendu d'une tâche donnée (32) ;

    la génération, pour chacune des tâches de la pluralité de tâches (32), d'une valeur de variance accumulée (114) sur la base des comparaisons ;

    la génération sélective d'une indication d'échec prédictive sur la base des valeurs générées de variance accumulée (114) ; et

    si une indication d'échec prédictive est générée, l'identification de la tâche parmi la pluralité de tâches (32) au sein du cycle (30) a conduit à l'indication d'échec prédictive.


     
    2. Procédé selon la revendication 1, selon lequel la comparaison, pour chacune des tâches de la pluralité de tâches (32), de chaque temps d'accomplissement (54) par rapport à une valeur de référence comprend :
    le calcul, pour chacune des tâches de la pluralité de tâches (32), d'une différence entre chacun des temps d'accomplissement (54) et la valeur de référence.
     
    3. Procédé selon la revendication 2, selon lequel la génération de la valeur de variance accumulée (114) comprend :
    le calcul de la somme, pour chacune des tâches de la pluralité de tâches (32), des différences correspondant à au moins certains des temps d'accomplissement (54).
     
    4. Procédé selon la revendication 1, comprenant en outre :
    le calcul, pour chacune des tâches de la pluralité de tâches (32), d'une pente des valeurs de variance accumulées (114) sur un intervalle de temps.
     
    5. Procédé selon la revendication 4, selon lequel la génération sélective des indications d'échec prédictives comprend en outre :

    si la pente est au moins une pente croissante ou décroissante sur l'intervalle de temps, la génération de l'indication d'échec prédictive ; et

    si la pente est sensiblement constante sur l'intervalle de temps, l'absence de génération de l'indication d'échec prédictive.


     
    6. Procédé selon la revendication 1, selon lequel la valeur de référence d'une tâche donnée (32) est basée sur des données autres que la série de signaux.
     
    7. Procédé selon la revendication 6, selon lequel la valeur de référence d'une tâche donnée (32) est en outre basée sur une spécification prédéfinie.
     
    8. Procédé selon la revendication 1, selon lequel la valeur de référence d'une tâche donnée (32) est basée sur les séries de signaux.
     
    9. Procédé selon la revendication 1, selon lequel l'indication d'échec prédictive est moins une indication choisie parmi une indication audible ou visuelle.
     
    10. Procédé selon la revendication 1, selon lequel la série de signaux est au moins une série de signaux parmi des données analogiques ou des données numériques.
     
    11. Appareillage pour la surveillance de la performance d'un équipement situé dans un environnement d'automatisation comprenant :

    une mémoire ; et

    un processeur configuré pour exécuter des instructions stockées dans la mémoire afin de :

    identifier un cycle (30) associé avec l'équipement, le cycle (30) étant défini par une séquence d'une pluralité de tâches (32) ;

    collecter, au cours d'une pluralité de répétitions de la séquence, une série de signaux pour chacune de la pluralité de tâches (32) au sein du cycle (30) ;

    associer, pour chacune d'au moins certaines répétitions de la pluralité des répétitions, un ou plusieurs signaux dans la série afin de définir un temps d'accomplissement (54) pour chacune des tâches de la pluralité de tâches (32) ;

    comparer, pour chacune des tâches de la pluralité de tâches (32), chacun des temps d'accomplissement (54) par rapport à une valeur de référence indiquant un temps d'accomplissement attendu d'une tâche donnée (32) ;

    générer, pour chacune des tâches de la pluralité de tâches (32), d'une valeur de variance accumulée (114) sur la base des comparaisons ;

    générer de manière sélective une indication d'échec prédictive sur la base des valeurs générées de variance accumulée (114) ; et

    si une indication d'échec prédictive est générée, identifier la tâche parmi la pluralité de tâches (32) au sein du cycle (30) ayant conduit à l'indication d'échec prédictive.


     
    12. Appareillage selon la revendication 11, selon lequel le processeur est en outre configuré pour: calculer, pour chacune des tâches de la pluralité de tâches (32), une pente des valeurs de variance accumulées (114) sur un intervalle de temps.
     
    13. Appareillage selon la revendication 12 selon lequel le processeur est en outre configuré pour générer de manière sélective l'indication d'échec prédictive comportant :

    si la pente est au moins une pente croissante ou décroissante sur l'intervalle de temps, la génération de l'indication d'échec prédictive ; et

    si la pente est sensiblement constante sur l'intervalle de temps, l'absence de génération de l'indication d'échec prédictive.


     




    Drawing























    Cited references

    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description