Intelligent condition-monitoring and fault diagnostic system for predictive maintenance

• US 7,882,394 B2
• Filed: 07/11/2006
• Issued: 02/01/2011
• Est. Priority Date: 07/11/2005
• Status: Active Grant

• Current Assignee: Brooks Automation Ltd. (United Kingdom) (Azenta, Inc.)
• Original Assignee: Brooks Automation Incorporated (Azenta, Inc.)
• Inventors: Krishnasamy, Jay, Prochazka, Jan, Hosek, Martin
• Primary Examiner(s): Baderman; Scott T
• Assistant Examiner(s): KO, CHAE M
• Application Number:
  US11/485,143
• Publication Number:
  US 20070067678A1
• Time in Patent Office: 1,666 Days
• Field of Search: 714/26, 714/47, 702/184, 702/183, 702/185
• US Class Current: 714/26
• CPC Class Codes: G05B 23/0235 based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold
• CPC Class Codes: G06F 11/008 Reliability or availability analysis
• CPC Class Codes: G07C 3/00 Registering or indicating the condition or the working of machines or other apparatus, other than vehicles
• CPC Class Codes: G16H 40/40 for the management of medical equipment or devices, e.g. scheduling maintenance or upgrades

Abstract

A system for condition monitoring and fault diagnosis includes a data collection function that acquires time histories of selected variables for one or more of the components, a pre-processing function that calculates specified characteristics of the time histories, an analysis function for evaluating the characteristics to produce one or more hypotheses of a condition of the one or more components, and a reasoning function for determining the condition of the one or more components from the one or more hypotheses.

Claims

1. A system for condition monitoring and fault diagnosis comprising:

a first controller having a processor configured to perform a data collection function that acquires time histories of selected variables for one or more of the components according to specified sampling parameters;

a second controller having a processor configured to perform a pre-processing function that calculates specified characteristics of each of the time histories individually;

a third controller having a processor configured to perform an analysis function for evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components; and

a fourth controller having a processor configured to perform a reasoning function for determining the condition of the one or more components from the one or more hypotheses;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

2. The system of claim 1, further comprising a function for utilizing the determined condition of the one or more components to effect preventative maintenance or service for the components.

3. The system of claim 1, wherein the reasoning function determines a condition of another system in which the one or more components operate from the one or more hypotheses.

4. The system of claim 1, wherein the specified sampling parameters include a sampling period, trigger mode, and number of samples to be recorded for the selected variables.

5. The system of claim 1, wherein each of the first, second, third, and fourth controllers reside in the one or more components.

6. The system of claim 1, wherein the pre-processing function calculates specified characteristics of the time histories by performing mathematical operations on the time histories.

7. The system of claim 1, wherein the analysis function produces the one or more hypotheses by applying one or more algorithms and analysis techniques to the calculated specified characteristics of the time histories.

8. The system of claim 1, wherein the reasoning function includes an expert diagnostic system for determining a condition of the one or more components from the one or more hypotheses.

9. A method of component condition monitoring and fault diagnosis comprising:

acquiring time histories of selected variables for one or more of the components according to specified sampling parameters;

calculating specified characteristics of each of the time histories individually;

evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components;

determining the condition of the one or more components from the one or more hypotheses; and

utilizing the determined condition of the one or more components to determine if preventative maintenance or service is required for the components;

wherein multiple instances of processors configured to perform acquiring time histories, calculating specified characteristics, evaluating the specified characteristics, determining the condition of the one or more components and utilizing the determined condition reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

10. The method of claim 9 including determining a condition of a system in which the one or more components operate from the one or more hypotheses.

11. The method of claim 9, wherein the specified sampling parameters include a sampling period, trigger mode, and number of samples to be recorded for the selected variables.

12. The method of claim 9, wherein acquiring time histories is performed by a controller of the one or more components.

13. The method of claim 9, wherein calculating specified characteristics of the time histories includes performing mathematical operations on the time histories.

14. The method of claim 9, wherein evaluating the characteristics to produce one or more hypotheses of a condition of the one or more components includes applying one or more algorithms and analysis techniques to the calculated specified characteristics of the time histories.

15. The method of claim 9, wherein determining the condition of the one or more components from the one or more hypotheses includes utilizing an expert diagnostic system.

16. A computer readable storage medium encoded with a computer program for component condition monitoring and fault diagnosis, that when run on a computer causes the computer to:

acquire time histories of selected variables for one or more of the components according to specified sampling parameters;

calculate specified characteristics of each of the time histories individually;

evaluate the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components;

determine the condition of the one or more components from the one or more hypotheses, and

utilize the determined condition of the one or more components to determine if preventative maintenance or service is required for the components.

17. The computer program product of claim 16, that when run on a computer causes the computer to utilize the determined condition of the one or more components to effect preventative maintenance or service for the components.

18. The computer program product of claim 16, that when run on a computer causes the computer to determine a condition of a system in which the one or more components operate from the one or more hypotheses.

19. The computer program product of claim 16, wherein the specified sampling parameters include a sampling period, trigger mode, and number of samples to be recorded for the selected variables.

20. The computer program product of claim 16, wherein calculating specified characteristics of the time histories includes performing mathematical operations on the time histories.

21. The computer program product of claim 16, wherein evaluating the characteristics to produce one or more hypotheses of a condition of the one or more components includes applying one or more algorithms and analysis techniques to the calculated specified characteristics of the time histories.

22. The computer program product of claim 16, wherein determining the condition of the one or more components from the one or more hypotheses includes utilizing an expert diagnostic system.

23. A system for component condition monitoring and fault diagnosis comprising:

a first controller having a processor configured to perform a data collection function that acquires time histories of selected variables for one or more components according to specified sampling parameters;

a second controller having a processor configured to perform a pre-processing function that calculates specified characteristics of each of the time histories individually;

a third controller having a processor configured to perform an analysis function for evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components;

a fourth controller having a processor configured to perform a reasoning function for determining the condition of the one or more components from the one or more hypotheses; and

a fifth controller having a processor configured to perform a manager function that determines the selected variables acquired by the data collection function, triggers data processing in the pre-processing function for calculating the specified characteristics, initiates evaluation of the characteristics by the analysis function to yield the hypotheses, and triggers derivation of the component conditions by the reasoning function;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

24. The system of claim 23, wherein the manager function utilizes the determined condition of the one or more components to effect preventative maintenance or service for the components.

25. The system of claim 23, wherein the reasoning function determines a condition of a system in which the one or more components operate from the one or more hypotheses.

26. The system of claim 23, wherein the specified sampling parameters include a sampling period, trigger mode, and number of samples to be recorded for the selected variables under control of the manager function.

27. The system of claim 23, wherein the pre-processing function calculates specified characteristics of the time histories by performing mathematical operations on the time histories, wherein the mathematical operations are determined by the manager function based on the type of component.

28. The system of claim 23, wherein the analysis function produces the one or more hypotheses by applying one or more algorithms and analysis techniques to calculated specified characteristics of time histories of a group of components operating together as a mechatronic device.

29. The system of claim 28, wherein the reasoning function includes an expert diagnostic system for determining a condition of the mechatronic device from the one or more hypotheses.

30. A system for monitoring component conditions of a semiconductor production tool comprising:

a component controller including a processor having a data collection function that acquires time histories of selected variables of one or more of the components according to specified sampling parameters;

a mechatronic device controller including a processor having a pre-processing function that calculates specified characteristics of each of the time histories individually for those of the one or more components operating together as a mechatronic device;

a group controller for a group of mechatronic devices, the group controller including a processor with an analysis function for evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components;

a system controller including a processor having a reasoning function for determining the condition of the one or more components from the one or more hypotheses, and a manager function that determines the selected variables acquired by the data collection function, triggers data processing in the pre-processing function for calculating the specified characteristics, initiates evaluation of the characteristics by the analysis function to yield the hypotheses, and triggers derivation of the component conditions by the reasoning function;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

31. The system of claim 30, wherein the manager function utilizes the determined condition of the one or more components to effect preventative maintenance or service for the components.

32. The system of claim 30, wherein the reasoning function determines a condition of a system in which the one or more components operate from the one or more hypotheses.

33. The system of claim 30, wherein the specified sampling parameters include a sampling period, trigger mode, and number of samples to be recorded for the selected variables under control of the manager function.

34. The system of claim 30, wherein the pre-processing function calculates specified characteristics of the time histories by performing mathematical operations on the time histories, wherein the mathematical operations are determined by the manager function based on the type of component and mechatronic device.

35. The system of claim 30, wherein the analysis function produces the one or more hypotheses by applying one or more algorithms and analysis techniques to calculated specified characteristics of time histories of the group of components operating together as the mechatronic device.

36. The system of claim 30, wherein the reasoning function includes an expert diagnostic system for determining a condition of the mechatronic device from the one or more hypotheses.

37. A continuous health monitoring system comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of one or more component energy dissipation values during component operations according to specified sampling parameters;

a second controller including a processor configured to perform a pre-processing function that computes metrics using an operational energy dissipation from each of the time histories individually and a baseline energy dissipation;

a third controller including a processor configured to perform an analysis function for evaluating whether the computed metrics exceed predetermined threshold values to produce one or more hypotheses of a condition of the one or more components; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the one or more components from the one or more hypotheses,

wherein the data collection, pre-processing, and analysis functions operate in parallel with the component operations and multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

38. The system of claim 37, wherein the data collection function acquires time histories of mechanical energy dissipation values.

39. The system of claim 37, wherein the data collection function acquires time histories of electrical energy dissipation values.

40. The system of claim 37, wherein the data collection function acquires time histories of energy dissipation values of a robotic joint.

41. The system of claim 37, wherein the data collection function acquires time histories for a predefined sequence of component moves.

42. The continuous health monitoring system of claim 37, wherein the pre-processing function computes a difference between the operational energy dissipation and the baseline energy dissipation as a first metric and an exponentially weighted moving average of the difference as a second metric.

43. The system of claim 37, wherein the pre-processing function computes the predetermined threshold values using a confidence coefficient for predicting a change in the metrics.

44. The system of claim 37, wherein the baseline energy dissipation used by the pre-processing function is acquired from data obtained from a selected move sequence.

45. The system of claim 37, wherein the baseline energy dissipation used by the pre-processing function is acquired from a component model.

46. A method of continuously monitoring system health comprising:

acquiring time histories of one or more component energy dissipation values during component operations according to specified sampling parameters;

computing metrics during the component operations using an operational energy dissipation from each of the time histories individually and a baseline energy dissipation;

in parallel with the component operations, evaluating whether the computed metrics exceed predetermined threshold values to produce one or more hypotheses of a condition of the one or more components;

determining the condition of the one or more components from the one or more hypotheses; and

utilizing the determined condition of the one or more components to determine if preventative maintenance or service is required for the components;

wherein multiple instances of processors configured to perform acquiring time histories, computing metrics, evaluating the computed metrics, determining the condition of the one or more components and utilizing the determined condition reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

47. The method of claim 46, further comprising acquiring time histories of mechanical energy dissipation values.

48. The method of claim 46, further comprising acquiring time histories of electrical energy dissipation values.

49. The method of claim 46, further comprising acquiring time histories of energy dissipation values of a robotic joint.

50. The method of claim 46, further comprising acquiring time histories for a predefined sequence of component moves.

51. The method of claim 46, further comprising computing a difference between the operational energy dissipation and the baseline energy dissipation as a first metric and an exponentially weighted moving average of the difference as a second metric.

52. The method of claim 46, further comprising computing the predetermined threshold values using a confidence coefficient for predicting a change in the metrics.

53. The method of claim 46, further comprising acquiring the baseline energy dissipation from data obtained from a selected move sequence.

54. The method of claim 46, further comprising acquiring the baseline energy dissipation from data obtained from a component model.

55. A continuous health monitoring system comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of one or more values related to power consumption by a component during operation according to specified sampling parameters;

a second controller including a processor configured to perform a pre-processing function that computes metrics using an operational power consumption from each of the time histories individually and a power consumption baseline;

a third controller including a processor configured to perform an analysis function for evaluating whether the computed metrics exceed predetermined threshold values to produce one or more hypotheses of a condition of the component; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the component from the one or more hypotheses,

wherein the data collection, pre-processing, and analysis functions operate in parallel with the component operations and multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

56. The system of claim 55, wherein the component power consumption related values include component current consumption.

57. The system of claim 55, wherein the component power consumption related values include one or more of component position, velocity, or acceleration.

58. The system of claim 55, wherein the power consumption baseline used by the pre-processing function is acquired from a component model.

59. The system of claim 55, wherein the pre-processing function computes a difference between the operational power consumption and the baseline power consumption as a first metric and an exponentially weighted moving average of the difference as a second metric.

60. The system of claim 59, wherein the data collection function acquires the time histories for a predefined set of component locations and the pre-processing function computes a difference between the operational power consumption and the baseline power consumption at those predetermined locations as the first metric.

61. The system of claim 59, wherein the data collection function acquires the time histories for a predefined sequence of component moves and the pre-processing function computes an integral of the absolute value of the difference between the operational power consumption and the baseline power consumption over the predefined sequence of component moves as the first metric.

62. The system of claim 55, wherein the pre-processing function computes a fast Fourier transform on portions of the time histories and on portions of the baseline power consumption, and wherein the analysis function monitors peaks emerging or shifting in a frequency spectrum from the transform.

63. A method of continuously monitoring system health comprising:

acquiring time histories of one or more power consumption related values of a component during operation according to specified sampling parameters;

computing metrics during the component operations using an operational power consumption from each of the time histories individually and a power consumption baseline;

in parallel with the component operations, evaluating whether the computed metrics exceed predetermined threshold values to produce one or more hypotheses of a condition of the component;

determining the condition of the component from the one or more hypotheses; and

utilizing the determined condition of the component to determine if preventative maintenance or service is required for the component;

wherein multiple instances of processors configured to perform acquiring time histories, computing metrics, evaluating the computed metrics, determining the condition of the component and utilizing the determined condition of the component reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

64. The method of claim 63, wherein the power consumption related values include component current consumption.

65. The method of claim 63, wherein the power consumption related values include one or more of component position, velocity, or acceleration.

66. The method of claim 63, further comprising determining the baseline energy dissipation from a component model.

67. The method of claim 63, further comprising computing a difference between the operational power consumption and the baseline power consumption as a first metric and an exponentially weighted moving average of the difference as a second metric.

68. The method of claim 67, further comprising acquiring time histories for a predefined set of component locations and computing a difference between the operational power consumption and the baseline power consumption at those predetermined locations as the first metric.

69. The method of claim 67, further comprising acquiring time histories for a predefined sequence of component moves and computing an integral of the absolute value of the difference between the operational power consumption and the baseline power consumption over the predefined sequence of component moves as the first metric.

70. The method of claim 63, further comprising computing a fast Fourier transform on portions of the time histories and on portions of the baseline power consumption, and monitoring peaks emerging or shifting in a frequency spectrum from the transform.

71. A continuous health monitoring system comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of one or more variables of a vacuum component related to operation of a vacuum device during component operations according to specified sampling parameters;

a second controller including a processor configured to perform a pre-processing function that calculates specific characteristics of each of the time histories individually;

a third controller including a processor configured to perform an analysis function for evaluating the characteristics of each of the individual time histories to produce one or more hypotheses about a condition of the vacuum component; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the vacuum component from the one or more hypotheses,

wherein the data collection, pre-processing, and analysis functions operate in parallel with the component operations and multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

72. The system of claim 71, wherein the vacuum device includes an edge gripper.

73. The system of claim 71, wherein the vacuum device includes a suction gripper.

74. The system of claim 71, wherein the one or more vacuum component variables include vacuum pressure readings.

75. The system of claim 71, wherein the one or more vacuum component variables include vacuum pressure transition times.

76. The system of claim 71, wherein the one or more vacuum component variables include operation times of the vacuum component.

77. The system of claim 71, wherein the one or more vacuum component variables include positions of the vacuum component.

78. The system of claim 71, wherein the one or more vacuum component variables include states of a vacuum component.

79. The system of claim 71, wherein the specific characteristics of the time histories include changes in vacuum pressure readings.

80. The system of claim 71, wherein the specific characteristics of the time histories include changes in vacuum pressure transition times.

81. The system of claim 71, wherein the specific characteristics of the time histories include changes in operation times of the vacuum component.

82. The system of claim 71, wherein the specific characteristics of the time histories include changes in positions of the vacuum component.

83. The system of claim 71, wherein the specific characteristics of the time histories include state changes of the vacuum component.

84. The system of claim 71, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum leak in the component.

85. The system of claim 71, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum obstruction in the component.

86. The system of claim 71, wherein one or more hypotheses about a condition of the vacuum component includes a component part failure.

87. A method of continuously monitoring system health comprising:

acquiring time histories of one or more variables of a vacuum component related to operation of a vacuum device according to specified sampling parameters during component operations;

calculating specific characteristics of each of the time histories individually during the component operations;

in parallel with the component operations, evaluating the characteristics to produce one or more hypotheses about a condition of the vacuum component; and

determining the condition of the vacuum component from the one or more hypotheses

utilizing the determined condition of the vacuum component to determine if preventative maintenance or service is required for the vacuum component;

wherein multiple instances of processors configured to perform acquiring time histories, calculating specific characteristics of each of the time histories, evaluating the characteristics, determining the condition of the vacuum component and utilizing the determined condition of the vacuum component reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

88. The method of claim 87, wherein the vacuum device includes an edge gripper.

89. The method of claim 87, wherein the vacuum device includes a suction gripper.

90. The method of claim 87, wherein the one or more vacuum component variables include vacuum pressure readings.

91. The method of claim 87, wherein the one or more vacuum component variables include vacuum pressure transition times.

92. The method of claim 87, wherein the one or more vacuum component variables include operation times of the vacuum component.

93. The method of claim 87, wherein the one or more vacuum component variables include positions of the vacuum component.

94. The method of claim 87, wherein the one or more vacuum component variables include states of the vacuum component.

95. The method of claim 87, wherein the specific characteristics of the time histories include changes in vacuum pressure readings.

96. The method of claim 87, wherein the specific characteristics of the time histories include changes in vacuum pressure transition times.

97. The method of claim 87, wherein the specific characteristics of the time histories include changes in operation times of the vacuum component.

98. The method of claim 87, wherein the specific characteristics of the time histories include changes in positions of the vacuum component.

99. The method of claim 87, wherein the specific characteristics of the time histories include state changes of the vacuum component.

100. The method of claim 87, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum leak in the component.

101. The method of claim 87, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum obstruction in the component.

102. The method of claim 87, wherein one or more hypotheses about a condition of the vacuum component includes a part failure in the components.

103. A continuous health monitoring system comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories according to specified sampling parameters in the form of video data related to operations of a device;

a second controller including a processor configured to perform a pre-processing function that calculates specific characteristics of each of the time histories individually;

a third controller including a processor configured to perform an analysis function for evaluating the characteristics of each of the individual time histories to produce one or more hypotheses about a condition of the device; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the device from the one or more hypotheses,

wherein the data collection, pre-processing, and analysis functions operate in parallel with the device operations and multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

104. The system of claim 103, further comprising a video camera for collecting the video data.

105. The system of claim 104, wherein the video camera is positioned to collect the video data from within a workspace of the device.

106. The system of claim 104, wherein the video camera is carried by the device.

107. The system of claim 103, wherein the device operations include pick or place operations.

108. The system of claim 103, wherein the video data includes position data of one or more device components.

109. The system of claim 103, wherein the video data includes image data.

110. The system of claim 109, wherein the image data includes an image of one or more components of the device.

111. The system of claim 109, wherein the image data includes an image of an object carried by the device.

112. The system of claim 103, wherein the specific characteristics of the time histories include an operational repeatability of the device.

113. The system of claim 103, wherein the specific characteristics of the time histories include a positional repeatability of the device.

114. A method of continuously monitoring system health comprising:

acquiring time histories according to specified sampling parameters in the form of video data related to operations of a device while the device is operating;

calculating specific characteristics of each of the time histories individually while the device is operating;

evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses about a condition of the device while the device is operating;

determining the condition of the device from the one or more hypotheses; and

utilizing the determined condition of the device to determine if preventative maintenance or service is required, for the device;

wherein multiple instances of processors configured to perform acquiring time histories, calculating specific characteristics of each of the time histories, evaluating the specified characteristics, determining the condition of the device and utilizing the determined condition of the device reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

115. The method of claim 114, comprising collecting the video data from within a workspace of the device.

116. The method of claim 114, comprising collecting the video data from a video camera carried by the device.

117. The method of claim 114, wherein the device operations include pick or place operations.

118. The method of claim 114, wherein the video data includes position data of one or more device components.

119. The method of claim 114, wherein the video data includes image data.

120. The system of claim 119, wherein the image data includes an image of one or more components of the device.

121. The system of claim 119, wherein the image data includes an image of an object carried by the device.

122. The method of claim 114, wherein the specific characteristics of the time histories include an operational repeatability of the device.

123. The method of claim 114, wherein the specific characteristics of the time histories include a positional repeatability of the device.

124. A continuous health monitoring system comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of selected variables for one or more components according to specified sampling parameters during component operations including a difference between a predicted torque and a measured torque of the one or more components;

a second controller including a processor configured to perform a pre-processing function that calculates specified characteristics of each of the time histories individually;

a third controller including a processor configured to perform an analysis function for evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the one or more components from the one or more hypotheses,

wherein the data collection, pre-processing, and analysis functions operate in parallel with the component operations and multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

125. The system of claim 124, wherein

the pre-processing function operates to compute metrics using an operational torque from the time histories and a torque baseline;

and

the analysis function operates to evaluate whether the computed metrics exceed predetermined threshold values to produce one or more hypotheses of a condition of the one or more components.

126. The system of claim 125, wherein the torque baseline used by the pre-processing function is acquired from a component torque model.

127. The system of claim 124, wherein the selected variables include a difference between a predicted current and a measured current of the one or more components.

128. The system of claim 127, wherein

the pre-processing function operates to compute metrics using an operational current from the time histories and a current baseline;

and

the analysis function operates to evaluate whether the computed metrics exceed predetermined threshold values to produce one or more hypotheses of a condition of the one or more components.

129. The system of claim 127, wherein the current baseline used by the pre-processing function is acquired from a component current model.

130. The system of claim 124, wherein the selected variables include a tracking error of the one or more components.

131. The system of claim 130, wherein the data collection function acquires the time histories for predefined component locations and the pre-processing function computes an integral of the absolute value of a tracking error as at least one of the specific characteristics.

132. The system of claim 130, wherein the data collection function acquires the time histories over a predetermined sequence of component moves and the pre-processing function computes an integral of the square of the tracking error as at least one of the specific characteristics.

133. The system of claim 124, wherein at least one of the components includes a motor and the selected variables include at least an actual voltage, current, and velocity of the motor.

134. The system of claim 133, wherein the pre-processing function calculates a modeled current from the actual voltage and velocity, calculates a difference between the actual and modeled current, and determines a maximum variation between the actual and modeled current as at least one of the specific characteristics.

135. The system of claim 134, wherein the analysis function compares the maximum variation with an allowable threshold, and the reasoning layer utilizes results from the analysis layer over several instances of data collection to confirm a change in characteristics.

136. The system of claim 124, wherein at least one of the selected variables includes an incremental position of the one or more components recorded upon receiving an index pulse from an incremental encoder.

137. The system of claim 136, wherein the analysis function computes a difference between incremental positions at successive index pulses and determines if the difference is within a tolerance of a multiple of the number of incremental counts per revolution of the incremental encoder.

138. The system of claim 124, wherein at least one of the selected variables includes an absolute position of the one or more components recorded by an absolute encoder.

139. The system of claim 138, wherein the analysis function computes a difference between the absolute position and an actual position of the one or more components.

140. The system of claim 138, wherein the analysis function computes a difference between the absolute position and an average actual position of the one or more components if the absolute encoder is stationary.

141. The system of claim 124, wherein the one or more components includes a belt and at least one of the selected variables includes a tension of the belt.

142. The system of claim 124, wherein at least one of the selected variables includes a vibration measurement of the one or more components.

143. The system of claim 124, wherein at least one of the selected variables includes a vibration measurement of the one or more components and the pre-processing function calculates a power spectrum of the vibration measurement as one of the specified characteristics.

144. The system of claim 124, wherein at least one of the selected variables includes a temperature of the one or more components.

145. The system of claim 144, wherein the analysis function compares the temperature with a threshold as at least one of the specified characteristics.

146. The system of claim 145, wherein the reasoning function determines the condition of one or more cooling fans from the one or more hypotheses produced by the analysis function.

147. The system of claim 124, wherein at least one of the selected variables includes an output of an over-travel switch indicating that the one or more components have traveled beyond a prescribed limit.

148. The system of claim 124, wherein:

at least one of the selected variables includes an arrival time of a status message and an error count from the one or more components;

the analysis function operates to determine if an interval between successive status packets of the status message exceeds a certain threshold and to compute an error rate for the one or more components; and

the reasoning function identifies a network topology location causing the threshold to be exceeded.

149. The system of claim 124, wherein at least one of the selected variables includes a position repeatability of a substrate carried by the one or more components.

150. The system of claim 124, wherein at least one of the selected variables includes a result of running a virus scan by the one or more components.

151. The system of claim 124, wherein at least one of the selected variables includes a plenum pressure on a first side of a filter component and an environment pressure on a second side of the filter component, and wherein the reasoning function determines if the filter component is clogged from an analysis of the pressures.

152. The system of claim 124, wherein at least one of the selected variables includes an electrostatic charge in an environment of the one or more components.

153. A system for automatic fault diagnosis comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories according to specified sampling parameters of selected power consumption related values for one or more of components of a device in response to deterioration in the operation of the device;

a second controller including a processor configured to perform a pre-processing function that calculates specified characteristics of each of the time histories individually including a range and a minimum variance and requests additional time history acquisition if certain thresholds are not met for the range and minimum variance;

a third controller including a processor configured to perform an analysis function for evaluating the characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components; and

a fourth controller including a processor configured to perform a reasoning function for determining faults of the one or more components or of the device from the one or more hypotheses;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

154. The system of claim 153, wherein the component power consumption related values include component current consumption.

155. The system of claim 153, wherein the component power consumption related values include one or more of component position, velocity, or acceleration.

156. The system of claim 153, wherein the power consumption baseline used by the pre-processing function is acquired from a component model.

157. A method for automatic fault diagnosis comprising:

acquiring time histories of selected power consumption related values according to specified sampling parameters for one or more components of a device in response to deterioration in the operation of the device;

calculating specified characteristics of each of the time histories individually including a range and a minimum variance and requests additional time history acquisition if certain thresholds are not met for the range and minimum variance;

evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components; and

determining faults of the one or more components or of the device from the one or more hypotheses; and

utilizing the determined condition of the one or more components or the device to determine if preventative maintenance or service is required for the components or the device;

wherein multiple instances of processors configured to perform acquiring time histories, calculating specified characteristics of each of the time histories, evaluating the specified characteristics, determining faults and utilizing the determined condition reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

158. The method of claim 157, wherein the component power consumption related values include component current consumption.

159. The method of claim 157, wherein the component power consumption related values include one or more of component position, velocity, or acceleration.

160. The method of claim 157, wherein the power consumption baseline used by the pre-processing function is acquired from a component model.

161. A system for automatic fault diagnosis comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of selected variables according to specified sampling parameters for one or more of components of a device in response to deterioration in the operation of the device;

a second controller including a processor configured to perform a pre-processing function that calculates specified characteristics of each of the time histories individually;

a third controller including a processor configured to perform an analysis function for evaluating the specified characteristics of each of the individual time histories to produce one or more hypotheses of a condition of the one or more components; and

a fourth controller including a processor configured to perform a reasoning function for determining faults of the one or more components or of the device from the one or more hypotheses;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

162. The system of claim 161, wherein at least one of the selected variables includes an actual voltage, torque, velocity, and position of the one or more components during a constant velocity section of a move sequence.

163. The system of claim 162, wherein the pre-processing function operates to:

request additional time histories if the position variables lack a certain range and minimum variance;

compute nominal torque from a model and residual torque as a difference between the nominal torque and the actual torque;

compute a mean and variance of the residual torque; and

compute a Fourier transform of the residual torque with respect to the position variables.

164. The system of claim 163, wherein the analysis function operates to determine if the residual torque exceeds a threshold, report peak frequencies of the Fourier transform, and to determine a distance corresponding to an oscillation period of the residual torque.

165. The system of claim 164, wherein the reasoning function operates to determine the presence of a fault, and to analyze the Fourier transform peaks to determine the fault.

166. The system of claim 161, wherein the selected variables include a tracking error of the one or more components and the pre-processing function operates to compute a Fourier transform of the tracking error as one of the specified characteristics.

167. The system of claim 161, wherein the selected variables include a tracking error and a force of the one or more components.

168. The system of claim 161, wherein the selected variables include an output of a disturbance observer for the one or more components.

169. The system of claim 161, wherein the selected variables include a belt tension of the one or more components.

170. The system of claim 161, wherein the selected variables include a vibration spectrum measurement of the one or more components and the pre-processing function operates to calculate a Fourier transform of the vibration measurement to obtain a vibration power spectrum as one of the specified characteristics.

171. A system for on-demand fault diagnosis comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of motor excitations and responses of an encoder for a motor of a device on an on-demand basis according to specified sampling parameters;

a second controller including a processor configured to perform a pre-processing function that calculates a frequency response of the device from each of the time histories;

a third controller including a processor configured to perform an analysis function for identifying frequencies from the frequency response corresponding to transverse belt vibrations of a belt drive driven by the motor; and

a fourth controller including a processor configured to perform a reasoning function for comparing the identified frequencies with an allowable range for the belt drive to determine the condition of the belt drive;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

172. A method for diagnosing faults on demand comprising:

acquiring time histories of motor excitations and responses of an encoder for a motor of a device on an on-demand basis according to specified sampling parameters;

calculating a frequency response of the device from each of the time histories individually;

identifying frequencies from the frequency response corresponding to transverse belt vibrations of a belt drive driven by the motor;

comparing the identified frequencies with an allowable range for the belt drive to determine the condition of the belt drive; and

utilizing the determined condition of the belt drive to determine if preventative maintenance or service is required for the belt drive;

wherein multiple instances of processors configured to perform acquiring time histories, calculating a frequency response, identifying frequencies, comparing the identified frequencies and utilizing the determined condition reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

173. A system for on-demand fault diagnosis comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of one or more variables of a vacuum component of a vacuum device on an on-demand basis according to specified sampling parameters;

a second controller including a processor configured to perform a pre-processing function that calculates specific characteristics of each of the time histories individually;

a third controller including a processor configured to perform an analysis function for evaluating the specified characteristics of the individual time histories to produce one or more hypotheses about a condition of the vacuum component; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the vacuum component from the one or more hypotheses;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

174. The system of claim 173, wherein the one or more vacuum component variables include vacuum pressure readings.

175. The system of claim 173, wherein the one or more vacuum component variables include vacuum pressure transition times.

176. The system of claim 173, wherein the one or more vacuum component variables include operation times of the vacuum component.

177. The system of claim 173, wherein one or more hypotheses about a condition of the vacuum component include a vacuum leak in the component.

178. The system of claim 173, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum obstruction in the component.

179. The system of claim 173, wherein one or more hypotheses about a condition of the vacuum component include a part failure in the components.

180. A method of continuously monitoring system health comprising:

acquiring time histories of one or more variables of a vacuum component of a vacuum device on an on-demand basis according to specified sampling parameters;

calculating specific characteristics of each of the time histories;

evaluating the specified characteristics of the individual time histories to produce one or more hypotheses about a condition of the vacuum component;

determining the condition of the vacuum component from the one or more hypotheses; and

utilizing the determined condition of the vacuum component to determine if preventative maintenance or service is required for the vacuum component;

wherein multiple instances of processors configured to perform acquiring time histories, calculating specific characteristics of each of the time histories, evaluating the specified characteristics, determining the condition of the vacuum component and utilizing the determined condition of the vacuum component reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

181. The method of claim 180, wherein the one or more vacuum component variables include vacuum pressure readings.

182. The method of claim 180, wherein the one or more vacuum component variables include vacuum pressure transition times.

183. The method of claim 180, wherein the one or more vacuum component variables include operation times of a vacuum operated component.

184. The method of claim 180, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum leak in the component.

185. The method of claim 180, wherein one or more hypotheses about a condition of the vacuum component includes a vacuum obstruction in the component.

186. The method of claim 180, wherein one or more hypotheses about a condition of the vacuum component includes a part failure in the component.

187. A system for on-demand fault diagnosis comprising:

a first controller including a processor configured to perform a data collection function that acquires time histories of one or more variables related to a component of a device on an on-demand basis according to specified sampling parameters;

a second controller including a processor configured to perform a pre-processing function that calculates specific characteristics of each of the time histories;

a third controller including a processor configured to perform an analysis function for evaluating the specified characteristics of the individual time histories to produce one or more hypotheses about a condition of the component of the device; and

a fourth controller including a processor configured to perform a reasoning function for determining the condition of the component from the one or more hypotheses;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.

188. The system of claim 187, wherein the one or more variables include positions and torques of the component as the device follows a predetermined trajectory.

189. The system of claim 188, wherein the pre-processing, analysis, and reasoning functions operate to identify parameters of a rigid-body dynamic model of the device from the time histories.

190. The system of claim 189, wherein the pre-processing, analysis, and reasoning functions operate to compare parameters of the rigid-body dynamic model representing properties of a joint of the device with baseline parameters of the joint to identify problems.

191. The system of claim 187, wherein the one or more variables include a response of the device as the device moves to positions uniformly distributed on a grid and is excited by a swept-frequency harmonic signal.

192. The system of claim 187, wherein the component of the device includes a motor and the one or more variables include steady state voltage and current values of the motor as the motor is commanded to produce a known force.

193. The system of claim 192, wherein the analysis function operates to compute a winding resistance value for the motor, and to compute current and force residuals by substituting the steady state current values and the commanded force value into equations of a motor model simplified for static load conditions.

194. The system of claim 193, wherein the reasoning function operates to compare the current and force residuals with set thresholds to identify a violation of the motor model.

195. The system of claim 187, wherein the one or more variables include sine and cosine signal values of an encoder of the device as the device moves at a constant velocity.

196. The system of claim 195, wherein the pre-processing function operates to compute a magnitude of an imaginary vector whose Cartesian coordinates are the sine and cosine values.

197. The system of claim 196, wherein the analysis function operates to record maximum and minimum values of the sine and cosine signals and the imaginary vector.

198. The system of claim 197, wherein the reasoning function operates to compare the maximum and minimum values to pre-set thresholds to determine a condition of the encoder.

199. The system of claim 187, wherein the one or more variables include a dimension of a feature scanned by a mapper of the device.

200. The system of claim 199, wherein the pre-processing, analysis, and reasoning functions operates to compare the scanned dimension with a known dimension of the feature to determine a condition of the mapper.

201. The system of claim 187, wherein the one or more variables include a response of the device as the device moves to positions uniformly distributed on a grid and is excited by an impulse, step-frequency or swept-frequency harmonic signal.

202. The system of claim 201, wherein the pre-processing, analysis, and reasoning functions operate to determine a stability measure of a controller of the device from the response.

203. The system of claim 187, wherein the component includes a communication network connected to the device and the one or more variables include a topology of the network.

204. The system of claim 187, wherein the one or more variables include an actual position of the component that causes a light beam to be cut, and wherein the analysis function operates to detect a shift in the actual position.

205. The system of claim 204, wherein the light beam is produced by a beam mapper carried by the component and is cut by a fixed feature.

206. The system of claim 204, wherein the light beam is fixed with respect to a work cell of the device and is cut by the component.

207. The system of claim 187, wherein the one or more variables include a location and orientation of a station mapped by the device.

208. The system of claim 187, wherein at least one of the selected variables includes a result of running a virus scan by the component.

209. A hierarchical system for monitoring component conditions of a machine comprising:

a component controller including a processor having a data collection function that acquires time histories of selected variables of one or more components of the machine according to specified sampling parameters;

a device controller including a processor having a pre-processing function that calculates specified characteristics of each of the time histories for a plurality of the component controllers operating together;

a subsystem controller for a group of the device controllers, the subsystem controller including a processor having an analysis function for evaluating the specified characteristics of the individual time histories to produce one or more hypotheses of a condition of the one or more components; and

a system controller including a processor having a reasoning function for determining the condition of the one or more components from the one or more hypotheses, and having a manager function that determines the selected variables acquired by the data collection function, triggers data processing in the pre-processing function for calculating the specified characteristics, initiates evaluation of the characteristics by the analysis function to yield the hypotheses, and triggers derivation of the component conditions by the reasoning function;

wherein multiple instances of each function reside in progressively higher level controllers within a machine being monitored such that data required for health monitoring and fault diagnostic purposes are used at the level where intelligence to process that data is present.