Ch. 27: Standard data The reuse of previous times. Advantages For example, predict cost of automotive repairs. Advantages Ahead of production The operation does not have to be observed. Allows estimates to be made for bids, method decisions, and scheduling. Cost Time study is expensive. Standard data allows you to use a table or an equation. Consistency Values come from a bigger database. Random errors tend to cancel over many studies. Consistency is more important than accuracy. ISE 311 - 25
Random vs constant errors ISE 311 - 25
Disadvantages of standard data Imagining the task The analyst must be very familiar with the task. Analysts may forget rarely done elements. Database cost Developing the database costs money. There are training and maintenance costs. ISE 311 - 25
Motions vs. elements Decision is about level of detail. MTM times are at motion level. An element system has a collection of individual motions. Elements can come from an analysis, time studies, curve fitting, or a combination. ISE 311 - 25
Constant vs. variable Each element can be considered either constant or variable. Constant elements either occur or don’t occur. Constant elements tend to have large random error. Variable elements depend on specifics of the situation. Variable elements have smaller random error. ISE 311 - 25
Developing the standard Plan the work. Classify the data. Group the elements. Analyze the job. Develop the standard. ISE 311 - 25
Curve fitting To analyze experimental data: Plot the data. Guess several approximate curve shapes. Use a computer to determine the constants for the shapes. Select which equation you want to use. ISE 311 - 25
Statistical Concepts Least-squares equation Standard error Coefficient of variation Coefficient of determination Coefficient of correlation Residual ISE 311 - 25
Curve Shapes Y independent of X Y = A Determine that Y is independent of X by looking at the SE. 0 2 4 6 8 10 10 8 6 4 2 [x] [y] y=4 ISE 311 - 25
Curve Shapes Y depends on X, 1 variable Examples Others: (2) Y = AeBX (1) Y = A + BX (2) Y = AeBX (3) Y = X / (A + BX) Others: Y = AXB Y = A + BXn Y = A + BX + CX2 ISE 311 - 25
Curve Shapes Y depends on X, multiple variables Y = A + BX + CZ Results in a family of curves ISE 311 - 25
Example: Walk normal times (min) From table 27.2, pg. 526 5 m 10 m 15 m 20 m .0553 .1105 .1654 .2205 .0590 .1170 .1751 .0550 .1660 .2090 .0521 .1045 .1680 .2200 .0541 .1080 .1625 .2080 .0595 .1200 .1800 .1980 ISE 311 - 25
Walk Data Scatterplot ISE 311 - 25
Equations for Walk Data Set Walk time h =.0054 + .01D r2 = .986 σ = .0073 h Walk time h = –.01 + .014D –.00013D2 r2 = .989 σ = .0067 h Walk time h = –.13 + .11 loge D r2 = .966 σ = .012 h 1/Walk time h = .24 – .96 (1/D) r2 = .881 σ = .021 h-1 ISE 311 - 25