qc tools

28
PROBLEM SOLVING TOOLS Lets us improve our understanding on…..

Upload: dilip-sahoo

Post on 24-Sep-2015

11 views

Category:

Documents


0 download

DESCRIPTION

problem solving method

TRANSCRIPT

  • PROBLEM SOLVING TOOLSLets us improve our understanding on..

  • FLOW CHARTBRAIN STORMINGCAUSE AND EFFECT DIAGRAMDATA COLLECTIONGRAPHS AND CHARTSSTRATIFICATIONPARETO ANALYSISHISTOGRAMSCATTER DIAGRAMBOX AND WHISKER

  • FLOW CHARTA flow diagram is a graphic representation of sequence of steps that we perform same output. The output may be a physical product, a service, information or a combination of the three.

  • ProcessDecision BoxDocument / RecordConnectorExtended ProcessManual OperationManual Input

  • Develop a common understanding of an overall process. Uncover potential problems, bottle necks unnecessary steps and rework loops in the process. Guide the potential solution without having to observe the process physically.

    FLOW CHART

  • InterpretationBetter understanding of the current process, as a whole, and common understanding.Estimation of cost of Quality.Develop plans for data gathering.Generate theories about root causes. Discuss ways to stratify the data for analysis, to identify root cause.Identify the need to streamline the process or changes in the process.Provide training on how the process should operate.

  • EXCERCISEDraw a Flow Chart for following process :- Rework of Non Conforming product. Disposition of Non Conforming In-coming material.

  • Cause EffectBasic to quality improvement is the need to identify the root cause of a problem. The cause effect diagram is an effective way to organize and display the various theories about what those root cause might be.

  • Visual Representation and Interrelationship among the possible factors and effect. (Aspect Impact Sheet)Strength

    Focuses the the attention of all people in the team on the specific problem at hand in a structured and systematic way.Encourage innovative thinking.C-E diagram has a tremendous capability of communicating to others.Cause Effect

  • Step how to construct :-Define clearly the effect.Draw a arrow (thicker) and place the effect at the arrow.Use brain storming or a rational step- by-step approach to identify the possible causes.Each of the major cause placed and joined with central spine at 70 deg. Angle.Add causes for each main area and subsidiary cause till you reach at root cause.Check the logical validity of each cause chain.Check the completeness.Cause Effect

  • EXCERCISEDraw a Cause and Effect Diagram for following effects :- Mixing of Arm rear brake stopper Reducing Water Generation.

  • Stratification

    Stratification is the separation of data into categories. Its most fragment use is during the diagnostic to identify categories contribute to the problem being solved.

    Stratification Variable :- The Characteristic used to separate the data is called a stratification variable.

  • Stratification

    Point out exactly the root cause.Which area needs further investigationInterpretation 1. Usually present the stratification results in bar chart .2. If no category stands out, look at another variable. Within each category of first variable.Pitfalls and Traps :-Conclude too much.Small difference / May be random error.

  • Exercise :-If a team is solving a problem for which no data exist, there is no point in making a list of variables or characteristics to be used in stratifying data.When a stratification does not indicate an obvious pattern, the exercise has been a waste of time.Stratification is a tool which applies well only to relatively simple situations with a limited number of variables.Stratification applies to situations dealing with methods and procedures as well as with technology.

  • When to Stratify :Identifying Projects COQ./COPQAnalyzing Symptoms.Testing theories. Steps :-1 Select the Stratification variable.2 Establish Category for each variable.3 Sort the observation.4 Display

  • Pareto Analysis Pareto Analysis is a ranked comparison of factor related to a quality problem. It helps a quality improvement project teams to identify and focus on the VITAL few factors.

    Vilfredo (1848-1923)M.O. LorenzVital few useful many.

  • Break Point Awkward Zone.

    Sheet1

    ItemNumber of ErrorsPercent of TotalCumulative Percent of Total

    G442929

    J382555

    M312175

    Q161186

    B8591

    D5395

    C3297

    A1197

    O1198

    R1199

    N1199

    L11100

    I00100

    E00100

    H00100

    K00100

    F00100

    P00100

    Total150

    Sheet1

    &A

    Page &P

    Percent of Total

    Cumulative Percent of Total

    Sheet2

    Sheet3

  • If a Pareto analysis does not shows a clear cut break point usually takes the following approach :-Identify the those few contributor which counts for about 60% of the total effect.Call these vital few and begin the diagnostic journey.Once you find a some path and improve repeat the Pareto, the contributor that were in awkward Zone may now be among the vital few.Repeat 1-3.

  • When to use Pareto :-1. Prioritizing Problems.2. Analyzing Symptoms.3. Proving Effectiveness of remedy.

    Misapplication :-Objective data rather then opinion.

  • BOX and WHISKER PLOTBox plots like histogram, provide a graphic summary of the patterns of variation in a set of data.The box plot is especially useful when working with small sets of data or comparing many different distribution.

  • BOX and WHISKER PLOTHIGHEST VALUETHIRD QUARTILEMEDIANFIRST QUARTILELOWEST VALUE

  • BOX and WHISKER PLOT

    Sheet1

    ItemNumber of ErrorsPercent of TotalCumulative Percent of Total

    G442929

    J382555

    M312175

    Q161186

    B8591

    D5395

    C3297

    A1197

    O1198

    R1199

    N1199

    L11100

    I00100

    E00100

    H00100

    K00100

    F00100

    P00100

    Total150

    Sheet1

    29.333333333329.3333333333

    25.333333333354.6666666667

    20.666666666775.3333333333

    10.666666666786

    5.333333333391.3333333333

    3.333333333394.6666666667

    296.6666666667

    0.666666666797.3333333333

    0.666666666798

    0.666666666798.6666666667

    0.666666666799.3333333333

    0.6666666667100

    0100

    0100

    0100

    0100

    0100

    0100

    &A

    Page &P

    Percent of Total

    Cumulative Percent of Total

    Sheet2

    ContractorResponse time in minutes

    12345678910

    A45183207226243251268272290345

    B102125145156188192215228251279

    Sheet3

    Sheet1

    ItemNumber of ErrorsPercent of TotalCumulative Percent of Total

    G442929

    J382555

    M312175

    Q161186

    B8591

    D5395

    C3297

    A1197

    O1198

    R1199

    N1199

    L11100

    I00100

    E00100

    H00100

    K00100

    F00100

    P00100

    Total150

    Sheet1

    29.333333333329.3333333333

    25.333333333354.6666666667

    20.666666666775.3333333333

    10.666666666786

    5.333333333391.3333333333

    3.333333333394.6666666667

    296.6666666667

    0.666666666797.3333333333

    0.666666666798

    0.666666666798.6666666667

    0.666666666799.3333333333

    0.6666666667100

    0100

    0100

    0100

    0100

    0100

    0100

    &A

    Page &P

    Percent of Total

    Cumulative Percent of Total

    Sheet2

    ContractorResponse time in minutes

    12345678910

    A45183207226243251268272290345

    B102125145156188192215228251279

    Contractor

    AB

    Minmin45102

    First Quartile((1)n+2) / 4207145

    Median(n+1)/2247190

    Third Quartile((3)n+2) / 4272228

    Maxmax345279

    Sheet3

  • Median response time of B is better than A.Variability of A is high than BEven though the A had shortest response time only about 25% of As response time were less than 207, while 75% of Bs were less than 228min.Nearly 25% of As response times were longer than Bs longest response time.

  • Comparing Patterns of variationsDifferences in the location of the MEDIAN.Differences in the amount of the variation.The presence or absence of outliers.Symmetry or asymmetry in the data.

  • BOX PLOT Vs HISTOGRAMBox plot indicates about the width of variation and the presence of various forms of asymmetry. Whereas, Histogram provides information about distribution of data along with width of variation and central tendency.

  • BOX PLOT Vs HISTOGRAM

  • Potential Pitfalls and problems in interpretationCurrent condition true data representation .More than 40 points - good enough.Just help you / lead you to establish a THEORY, which needs further confirmation.