Employing a dynamic lifecycle condition index (CI) to accommodate for changes in the expected service life of an item based on observance of the item and select extrinsic factors

• US 7,769,568 B2
• Filed: 09/08/2005
• Issued: 08/03/2010
• Est. Priority Date: 07/09/2004
• Status: Active Grant

• Current Assignee: United States Government Humphreys Engineer Center
• Original Assignee: The United States Of America As Represented By The Secretary Of The Army
• Inventors: Marrano, Lance R., Grussing, Michael N., Uzarski, Donald R.
• Primary Examiner(s): Shah; Kamini S
• Assistant Examiner(s): GEBRESILASSIE, KIBROM K
• Application Number:
  US11/223,251
• Publication Number:
  US 20070239368A1
• Time in Patent Office: 1,790 Days
• Field of Search: 702/34, 702/33, 702/184, 703/2
• US Class Current: 703/2
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/0635 Risk analysis of enterprise or organisation activities
• CPC Class Codes: G06Q 10/20 Administration of product repair or maintenance

Abstract

Initial assumptions related to the service life of a particular item, such as a component section of a building, are mathematically modeled to construct an initial lifecycle condition relationship as condition index (CI) v. time. To update the model, empirical data may be input at any time. As modeled in an engineering management system, for example, inspections are performed on the item to verify actual condition with that predicted. Quantitative inspection data are then used to update the initial curve. As inspections are performed and data recorded, the curve is updated to accurately capture observed condition and provide realistic estimates of predicted condition, and expected service life. In select embodiments of the present invention, empirical data, such as that from inspections, are weighted, e.g., inspection data may be weighted based on type, level of detail, time in service, time since last inspection and the like.

Claims

1. A tool for use in managing operations and maintenance (O&

M) of facilities comprising multiple items, comprising;

computer readable media incorporating;

a mathematical model that employs pre-specified parameters to adjust a relationship of a condition index (CI) to time for at least one said item, said model automatically adjusting said relationship of said CI to time to reflect at least observed changes to said at least one item, said changes input as updates to categories of data;

data of a first type in at least one of said categories of data representing at least assumptions and comprising data points;

wherein at least one said assumption is related to said at least one item, said at least one assumption asserted as at least one limitation in said mathematical model;

data of a second type in at least one of said categories of data representing at least some empirical observations related to said at least one item,

wherein at least some of said updates utilize at least said data of a second type;

at least one specially configured computer in operable communication with said computer readable media, said at least one specially configured computer specially configured to at least execute said mathematical model; and

at least one display in operable communication with each said at least one specially configured computer,

wherein on at least said at least one display, said adjustable relationship of said CI to time is at least presented as one curve of CI versus time in which said at least one curve is adjusted to include the last said change input as a point on said curve, and

wherein use of said tool yields at least quantitative measures of the impact on at least O&

M lifecycle cost of most recent upgrades, repairs and preventive maintenance actions to at least said at least one item.

2. The tool of claim 1 in which said assumptions are used to define parameters to initially establish said CI, said assumptions at least comprising:

said CI is between 0 and 100;

said CI is 100 at the start of service life of each of said multiple items selected for review;

said CI approaches zero asymptotically;

said service life of each of said multiple items selected for review is defined at a pre-specified value of CI occurring at a pre-specified time; and

said CI decreases over time for each of said multiple items without at least an enhancement to each of said multiple items.

3. The tool of claim 2 said data of a second type selected from the group of said updates consisting of:

data on enhancements made to each of said multiple items selected for review, data from inspections made of each of said multiple items selected for review, historical data related to each of said multiple items selected for review by item type, historical data related to replaced ones of said multiple items selected for review of the same said item type, industrial data related to said item type of said multiple items selected for review, and combinations thereof.

4. The tool of claim 3 in which said enhancements are selected from the group consisting of:

corrective actions, repairs, upgrades, preventive maintenance, scheduled maintenance, and combinations thereof.

5. The tool of claim 2 in which said parameters are a, α

, and β

as used in the mathematical relationship defining a Weibull distribution,

C(t)=a×

e<sup>−

(t/β

)α</sup>

where;

C(t)=said condition index (CI) as a function of time,

t=normalized age (years) with respect to design life of said item,

a=parameter, initial steady state condition index,

β

=parameter, service life adjustment factor,

α

=parameter, accelerated deterioration factor,

wherein each said parameter for setting an initial said C(t) is derived from selected said assumptions.

6. The tool of claim 2 in which said pre-specified value of said CI is between about 60 to about 80 and said pre-specified time is equal to between about 60% to about 80% of said service life.

7. The tool of claim 6 in which said pre-specified value of said CI is about 70 and said pre-specified time in service is about 70% of said service life.

8. The tool of claim 3, said mathematical model further incorporating at least one factor to adjust for differences resulting from said updates to said data of a second type.

9. The tool of claim 8 said at least one factor comprising at least one weighting scale that can be applied to at least some of said data of a second type.

10. The tool of claim 9 further employing regression analysis to update a curve of said CI v. time by minimizing a “

sum of the squares”

residual error, wherein each said data point of said data of a second type is associated with one said weighting scale that modifies the value of said data point based on the residual error of said data point; and

wherein the smaller said residual error in said data point is, the higher is the value of said weighting scale, and

wherein said updated curve passes through a plot of the last recorded said condition index.

11. A method for employing mathematical models in a tool for use in managing facilities comprising multiple items, said tool residing on computer readable media, said method comprising:

executing at least one of said mathematical models by at least initially establishing a time-dependent relationship for a condition index (CI) for at least one of said multiple items;

employing said mathematical model to further automatically adjust said CI to reflect at least impacts to said time-dependent statistical relationship for a Cl, said impacts input as updates to data in pre-specified categories;

incorporating as said time-dependent relationship at least the equation

C(t)=a×

e<sup>−

(t/β

)α</sup>

in at least one of said mathematical models,

where;

C(t)=said condition index (CI) as a function of time,

T=normalized age (years) with respect to design life of said item,

a=parameter, initial steady state condition index,

β

=parameter, service life adjustment factor,

α

=parameter, accelerated deterioration factor; and

executing said mathematical model incorporating said equation on at least one specially configured computer, inputs for developing said parameters comprising;

data of a first type in at least one of said pre-specified categories of data comprising data points;

at least one quantitatively described assumption related to said at least one of multiple items; and

data of a second type in at least one of said pre-specified categories of data comprising at least some empirical data collected about said at least one of multiple items,

wherein at least some of said data of a second type are used in said updates; and

displaying results of using said tool,

wherein said results are presented as at least one curve of said adjustable CI versus time in which said at least one curve is adjusted to include the last said impact input as a point on said curve, and

wherein employing said method yields at least quantitative measures of the impact on at least O&

M lifecycle cost of at least most recent upgrades, repairs and preventive maintenance actions to at least said multiple items.

12. An engineering management system (EMS), said EMS incorporating therein a management tool to manage facilities comprising multiple items, said tool comprising:

computer readable media incorporating;

a mathematical model that establishes an adjustable time-dependent relationship of a lifecycle condition index (CI) for at least one of said multiple items, said model employed to further automatically adjust said CI to reflect at least observed impacts to said adjustable time-dependent relationship, said impacts input as updates to data in pre-specified categories;

data of a first type in at least one of said pre-specified categories, said data of a first type comprising data points;

at least one assumption related to said at least one of said multiple items, said at least one assumption asserted as at least one limitation in said mathematical model;

data of a second type in at least one of said pre-specified categories, said data of a second type representing at least some empirical observations related to said multiple items,

wherein results of inspections and industry data are input as said empirical observations data for said EMS

empirical observations, and

wherein some said updates utilize at least said data of a second type;

at least one specially configured computer in operable communication with said computer-readable media, said at least one specially configured computer specially configured to execute said mathematical model; and

at least one display in operable communication with said at least one specially configured computer,

wherein on said at least one display said adjustable CI is at least presented as at least one curve of said adjustable CI versus time in which said at least one curve is adjusted to include the last said impact input as a point on said curve, and wherein use of said tool yields at least quantitative measures of the impact on at least O&

M lifecycle cost of at least most recent upgrades, repairs and preventive maintenance actions to said multiple items.

13. A method for executing a mathematical model to yield at least one quantitative measure to aid operation and maintenance (O&

M) personnel in managing facilities incorporating multiple items, comprising;

1) selecting at least one of said multiple items;

2) establishing at least one mathematical relationship between pre-specified parameters related to a condition index (CI) of said multiple items and time to yield at least one first said quantitative measure,

3) choosing one said at least one mathematical relationship, including at least one of said pre-specified parameters, for incorporating in said mathematical model,

wherein at least one said mathematical relationship comprises at least one type of statistical distribution;

4) choosing at least one assumption related to said multiple items and said chosen mathematical relationship for incorporation in said mathematical model,

wherein said at least one assumption is provided from at least one category of data as data of a first type comprising data points;

5) applying said at least one assumption in said mathematical model;

6) executing said mathematical model on at least one specially configured computer;

7) gathering data of a second type containing at least empirical data related to the condition of said at least one item as defined by said at least one mathematical relationship to input updates to said at least one parameter of said mathematical model,

wherein said empirical data are from at least one category of data of a second type comprising at least some empirical data collected about at least one of said multiple items;

8) entering said data of a second type in said mathematical model;

9) re-executing said mathematical model on said at least one specially configured computer; and

10) repeating steps 7-9 as said updates of said data of a second type are input, resulting in automatic adjustment to said CI during the service life of said at least one item,

wherein results from executing said mathematical model are presented as at least one curve of said CI versus time in which said at least one curve is adjusted to include the last said change input as a point on said curve for said multiple items, and wherein employing said mathematical model yields at least quantitative measures of the impact on O&

M lifecycle cost of at least most recent upgrades, repairs and preventive maintenance actions to said multiple items.

14. The method of claim 13 further comprising gathering some said empirical data related to characteristics of said multiple items selected from the group consisting of:

manufacturer'"'"'s bulletins, upgrades, refurbishment, improvements, preventive maintenance, repair, replacement and combinations thereof, for inputting to steps 3, 4 and 7.

15. At least one specially configured computer for use in managing facilities comprising multiple items, said computer in operable communication with computer readable media, said media incorporating:

a mathematical model that establishes an adjustable time-dependent mathematical relationship for a condition index (CI) for at least one of said multiple items, said mathematical model automatically adjusting said CI to reflect at least observed impacts to said time-dependent mathematical relationship, said impacts input as updates to data in pre-specified categories, said mathematical model incorporating at least one type of statistical distribution in said mathematical model and at least one factor to adjust for differences in at least said empirical data upon said updating of said mathematical model;

data of a first type in said pre-specified categories of data, said data of a first type comprising data points;

at least one assumption related to said multiple items, said at least one assumption asserted as at least one limitation in said mathematical model;

data of a second type in said pre-specified categories , said data of a second type representing at least some empirical observations related to said multiple items,

wherein some said updates utilize said data of a second type; and

at least one display in operable communication with said specially configured computer,

wherein on at least said display, said adjustable CI is presented as at least one curve of said adjustable CI versus time in which said curve is adjusted to include the last said impact input as a point on said curve, and

wherein use of said data of said first and said second type on said computer readable media with said specially configured computer yields at least quantitative measures of the impact on O&

M lifecycle cost of at least most recent upgrades, repairs and preventive maintenance actions said multiple items.

16. The specially configured computer of claim 15, said at least one factor comprising at least one weighting scale that is applied to at least some of said empirical data.

17. The specially configured computer of claim 16 further incorporating a regression analysis algorithm to fit a lifecycle condition index curve by minimizing a “

sum of the squares”

residual error,

wherein each said data point is associated with said at least one weighting scale that modifies any residual error associated with said data point; and

wherein the more accurate said data point is, the higher the value of said weighting scale in defining an adjusted lifecycle condition index curve.