The Statistical Process Flashcards
what questions should be asked as part of the statistical process control
probabilistic modelling
1. is my process out of specification
2. is the process changing with time
3. do I pass or fail the batch
4. on what basis do I pass or fail the batch
5. have I taken enough samples
6. what are the consequences of making an erroneous judgement
What is the Six Sigma
DMAIC
- Define goals
- measure using standards
- analyse causality
- improve results
- control and correct
What is the design for six sigma
DMADV
- define
- measure
- analyse
- design
- verify
What does process statistics measure
measure of dispersion and location
- QC charts
how are the attributes of process statistics measured
routine inspections
- P/F charts
- C and U type charts
how are the variables of process statistics measured
process character
- X and R charts
- Shewhart type charts
- moving average type charts
what is UCL/LCL
upper/lower control level
- is centreline +- 3
what is UWL/LWL
Upper/lower warning level
- is centreline +-2
what is UAL/LAL
upper/lower action level
- is centreline +- 1
what are the general features of charts based on variables
- a routine process uses the 3 sigma control levels, US standard quality range
- UCL/LCL
what is a type I error
false negative- producer disposes of acceptable goods
what is a type II error
false positive- producer accepts dangerous substandard goods
what is the moving average chart based on variables
- average represents 4 measurements and is constantly updated for new complete sets of data
- uses a weight constant for a weighting of data and defining limits
what is meant by uniformly weighted moving average
larger values of weight actually guard against small shifts
what is meant by exponentially weighted moving average
small values of weight are used to guard against small shifts