Engineered Software

Reliability & Maintenance Analyst
Data Input


Home

Life Data Analysis

Maintenance Optimization

Engineered Software Home Page

The File option of the main menu provides functions related to getting data in and out of the software. All options are standard with the exception of the New option and the Reporting option.  The 3 types of data are:

New

Select this option to enter data using the data entry screen. The user is prompted for the type of data to be entered - individual, grouped or field data. The "New" icon also performs this action. The default is to provide an entry grid for individual data, but if a grouped data grid is open, a new grouped entry grid will be provided. Data can be copied into the data entry screen from the clipboard with the "Paste" icon . A "check-mark" in the "Censoring" column will occur by copying a "-1" into the spreadsheet. The "Cut" icon or the "Copy" icon can be used to move data from the data entry grid to other applications.

When a cell is in edit mode the menu is disabled. Press the enter key, or click another cell to enable the menu.

Individual Data

The Individual Data entry grid, shown in the figure below, is used when the time to fail or the censoring time is known for all items in the data set. If the number of failures in a given time interval is known, but exact failure times are not known, use the Grouped Data entry grid. The default number of rows in the grid is 50,000, but up to 2 billion rows can be added using the "Specify Number of Rows" entry on the "Edit" menu. There are three columns in the Individual Data entry grid:

  1. Time - enter the time to fail or cycles to fail here.

  2. Quantity - enter the number of items that failed, or were censored, at the corresponding time here. This column is useful when many items are censored at the same time. For example, consider a test with 10,000 light bulbs. If the test is ended after 400 hours have elapsed, and only 5 light bulbs have failed, then 9,995 light bulbs were censored at 400 hours. There are two options for entering this data - enter 400 in the Time column 9,995 times with a Count of 1 and the Censored box checked, or enter 400 in the Time column one time with a Count of 9,995 and the Censored box checked. The value of this field must be equal to or greater than 1.

  3. Censoring - check this box if an item is removed from testing without failing.

images/rma_ind.jpg (59710 bytes)
Individual data entry grid.

 

There are three limits when using individual data:

  1. All failure times must be greater then zero,

  2. All entries in the "Quantity" column must be greater than or equal to 1 and must be integers, and

  3. The total number of data points must be less than 2 billion. There are several ways to exceed the limit of 2 billion data points. For example, there could be 4 rows of data with a quantity of 10,000 in each row for a total of 40,000 data points. It is best to use the grouped data option when data sets become large because of the time required to process large data sets. Any data set with over 1,000 points is a good candidate for grouped data.

When using the individual data grid there should not be more than 1 data point with the same failure time. This occurs when testing is monitored on an interval basis rather than continuously. It is best to use the grouped data procedures for this type of data collection. It is recommended to use the grouped data format if there are more than 1 million data points. Although the individual data grid allows 2 billion data points, execution time becomes large if the number of data points is greater than 1 million.

Grouped Data

The Grouped Data entry grid, shown in the figure below, is used if exact failure times are not known, but the number of failures in an interval is known. There are 3 columns in the Grouped Data entry grid:

  1. Start Time - enter the starting time of the interval here.

  2. End Time - enter the ending time of the interval here.

  3. Quantity - enter the number of failures in the time interval here.

images/rma_group.jpg (42359 bytes)
Grouped data entry grid.

 

The starting time of the first interval is assumed to be zero, and the starting time of each successive interval is set equal to the ending time of the previous interval. The value of the quantity must be greater than zero. The cumulative distribution function is not estimated from the last interval, it is only used to determine the total number of units tested. For example, consider the data shown in the figure below.

images/Image238.gif (3530 bytes)
Grouped data example.

 In this example, the test began with 30 items (3+5+1+21). After testing for 10 time units, 3 items had failed. After testing 20 time units, 5 additional units failed, for a total of 8 failures. After testing 30 time units, 1 additional unit failed, for a total of 9 failures, and testing was ceased. Since 30 units were tested, the additional 21 units are placed in the last cell. The "End Time" of 40 is arbitrary and has no meaning. The only restriction is that it is greater than the "Start Time" in the same row.

When using grouped data the only method of parameter estimation available is probability plotting. Also, there are no confidence limits on the probability plot. When grouped data is used, the sample sizes are typically so large that any confidence interval computed is very small.

Field Data

The Field Data entry grid, shown in the figure below, is used for reports give the percentage of failure per time period.  There are 2 options for the type of data input

  1. R/1000 - Data is entered in the format of failures per 1000 items in the field.

  2. C/100 - Data is entered in the format of failures per 100 items in the field.

There are 2 columns in the Field Data entry grid:

  1. Months in Service - enter the ending time of the interval here.  Months in service is the most common, but the units can also be in years or days.

  2. R/1000 or C1/00 - the title of this column changes with your selection of the data type.  Enter the cumulative number of failures per 100 or per 1000 depending on the data type.


Field data example.

In the screen shown above:

  • 0.03 failures had occurred on average for every 1000 items with at least 1 month of time in service.

  • 0.04 failures had occurred on average for every 1000 items with at least 2 months of time in service.

  • 0.07 failures had occurred on average for every 1000 items with at least 3 months of time in service.

  • 0.09 failures had occurred on average for every 1000 items with at least 4 months of time in service.

  • 0.15 failures had occurred on average for every 1000 items with at least 5 months of time in service.

Back to Top