Module 3: Quality Assurance Tools Learning Objectives • Learn the theory and method of use of the seven tools of quality needed in process analysis and problem-solving techniques • Seven tools of quality • Pareto charts • Flow charts • Cause-and-effect diagram • Check sheet • Histograms • Control charts • Scatter diagrams Pareto Chart • Method of prioritizing problems or causes by frequency of occurrence or cost • Based in the 80-20 rule: • 80% of the problem is caused by 20% of the sources • “Vital few” and “trivial many” • Depicted by a vertical bar graph arranged from left to right descending order Advantages of a Pareto Chart • Focuses efforts on problems with greatest potential for improvement • Distinguishes the critical causes from the less significant causes • Helps prevent “shifting the problem” where the solution removes some causes but worsens others Advantages of a Pareto Chart (cont.) • Measure the impact of improvement projects when comparing charts “before” and “after” • The chart shows the relative importance of problems in a simple, quickly interpreted, visual format. • Progress is measured in a highly visible format that provides incentive to push on for more improvement. Steps to Build a Pareto Chart • Identify problem • Choose categories that will be monitored • Choose the most meaningful unit of measurement • Frequency • Cost • Determine time period • Long enough to represent situation • Scheduled time to collect data is typical of a workday Steps to Build a Pareto Chart (cont.) • Collect data • Compare the frequency of each category • Draw chart: • List the categories on the horizontal line • Descending order, from left to right • Frequencies on the vertical line Steps to Build a Pareto Chart (cont.) • Draw the cumulative percentage line showing categories contribution • Optional • Draw vertical line on the right side of the chart • Plot cumulative values from left to right • Interpret results • Tallest bar represents biggest contributor • Perform analysis of category that has the “most impact” Cum % Example of a Pareto Chart 100 % 97% 91% 79% 61% 50 % 30 35% 20 24 10 18 12 8 4 2 0 Late Wrong Missing Dam aged Wrong Not received delivery product parts container address Delivery complains Further analysis using Pareto charts • Major cause breakdown: • Tallest bar is broken down into sub-causes 30 20 24 10 18 12 8 4 2 0 Late delivery Wrong product Missing parts Dam aged Wrong Not received container address 8 6 4 8 6 4 4 2 2 0 Missing address Driver mistake Late Transportation Administrative documentation problems delays Late delivery Further analysis using Pareto charts •Before and after: • New Pareto bars are drawn side-by-side with the original Pareto showing effect of change 25 24 20 18 15 12 10 10 10 8 8 5 4 4 3 2 1 0 Late Wrong Missing Dam aged Wrong Not delivery product parts container address received Before After Further analysis using Pareto charts •Change measurement scale: • Same categories are used but measured differently. Typically cost and frequency. • This exercise defines the category of “most impact”. $3,000 30 $2,500 25 $2,000 20 $1,500 15 $2 ,50 0 24 $1,000 $1,8 50 10 18 $1,2 0 0 12 $500 5 8 $8 0 0 4 $2 2 0 $2 0 0 2 $0 0 Late Wrong Missing Dam aged Wrong Not received Wrong product Not received Missing parts Wrong Damaged Late delivery delivery product parts container address address container Delivery com plains Cost/month Flow Charts • Pictorial representation of the steps and decision points in a process. • Flow charts are used to identify the actual path of a product or service. • Flow charts show: • Sequential work activities • Inputs for each action • Outputs from each activity Advantages of a Flow Chart • Provide common understanding of how a process works. • Identifies problem areas, unexpected complexity, redundancies, and areas of potential improvement • Serves as a training aid • Provides basis for documentation. • Identifies location where additional data can be collected and investigated Symbols of a Flow Chart Shows an activity performed in the process Shows a point in the process where a Yes/No question is being asked or a decision is required Identifies a break in the flow chart and is continued A elsewhere on the same page or another page Shows the flow of the process Shows the beginning and end process Steps to Build a Flow Chart • List the basic steps of the process. • Determine the logical sequence of the steps. • Identify the inputs of each step. • Determine the outputs of each step. • Draw the chart using the symbols to show the actions and decisions. Example of a Flow Chart Team is given task to write procedure Review draft yes Complete Send to Develop Draft Final Release outline approved version Dept. No Release Complete Revise Draft procedure draft with changes Cause and Effect Analysis • Pictorial representation of all possible causes contributing to a problem. • Developed by Dr.Kaoru Ishikawa •Also known as “Fishbone” or “Ishikawa” diagrams Advantages of a Cause and Effect Diagram • Clearly illustrates the various causes affecting a process: • Relationship between these causes • Where are they occurring • Helps in finding the most basic cause of the problem • Motivates team members participation Steps to Build a Cause and Effect Diagram • Place problem statement in a box to the right-hand side of the paper. • Select the major cause or categories and place them to the left of the problem statement. • 4 M’s (Production process): • Materials • Manpower • Machinery/equipment • Method • 4 P’s (Service process): • Policies • Procedures • People • Plant / equipment Steps to Build a Cause and Effect Diagram (cont.) • Draw a box around each category and connect to a line pointing out towards the problem statement. • Using the brainstorming technique, generate ideas of causes, on the major categories. • Record these ideas on a line off the applicable major category line. Steps to Build a Cause and Effect Diagram (cont.) • For each cause listed on the diagram, ask “why does this happen?”. • For each response, ask the same question. • Each successive answer is another possible cause • Look for causes that repeat across major cause categories. • Ensure all team members agree on the problem and causes statements. Example of a Cause and Effect Diagram Machinery/equipment Manpower Drivers don’t Unreliable trucks Show up Drivers get lost Not enough trucks Wrong address Not capacity for on shipper peak periods Database Input error Late deliveries Poor dispatching Run out of products Poor handling of Poor planning large orders Lack of Lack of training training Method Material Check Sheets • Tool used to record and compile frequency of observations as they occur • Used for Pareto charts and histograms • Design varies depending on information needed Advantages of a Check Sheet • Patterns of information are clearly detected and shown • Easy-to-understand data from a simple process • Based on facts, not opinions • Forces agreement on the definition of each condition Steps to Build a Check Sheet • Agree on the definition of the events or conditions to observe. • Decide who will collect the data. • Determine the time period. • Design a check sheet form that is clear and easy to use. • Define Source of information. • Determine Content of information. • Collect data consistently and accurately. Example of a Check Sheet Project: Defects on fastener installation Performed by: John Doe Location: Cost Center 727 Reason: Defect reduction Time period: W/E 4/23/04 Type of defects 4/19 4/20 4/21 4/22 4/23 Total Gapped fasteners 15 15 12 10 14 66 Missing fasteners 3 0 0 1 1 5 Damaged fasteners 8 3 12 8 4 35 Defective fasteners 12 3 5 3 6 29 Total 38 21 29 22 25 135 Histograms • Graphical representations of the frequency distribution of data in bar form. • Summarizes data from a process that has been collected over a period of time • Provide a quick representation of the “spread” and “centering” of a process Advantages of a Histogram • Display large amounts of data that are difficult to interpret in a tabular form • Show the relative frequency of occurrences of the various data values • Reveal the variation, centering, and distribution shape of the data • Very useful when calculating capability of a process • Helps predict future performance of a process Steps to Build a Histogram • Collect data for analysis. • At least 50 to 100 data points • Use historical data to find patterns or to use as a baseline for past performance • Determine the range of the data set. • Smallest value subtracted from largest value • Determine quantity of categories. • Take the square root of total number of data points and round to nearest whole number Steps to Build a Histogram (cont.) • Determine each category’s data point. • Mid-point of each category • Plot data on a vertical bar-graph. • Frequency on Y-axis • Categories on X-axis • For each class interval, draw bar with the the height equal to frequency count Example of a Histogram Class Category Mid-point Frequency boundary 1 10.00 - 10.19 10.1 1 2 10.20 - 10.39 10.3 6 3 10.40 - 10.59 10.5 12 4 10.60 - 10.79 10.7 8 5 10.80 - 10.99 10.9 6 6 11.00- 11.19 11.1 3 14 12 10 8 6 12 4 8 6 6 2 3 1 0 10.00 - 10.20 - 10.40 - 10.60 - 10.80 - 11.00- 10.19 10.39 10.59 10.79 10.99 11.19 Control Charts • Line graph of measurements of a process overtime that has statistically based control limits placed on it • Process control charts monitor and display variations in a process output. • Control limits are based on process variation • Define expected variation range due to common causes • +/- three standard deviations from centerline • Centerline represents the average of all measurements used Types of Control Charts • Two primary types are: • Control charts for variables: • Most used: X –R • X is average values • R is range • Others: Run charts, moving range charts (MX –MR charts) • Control charts for attributes: • p chart: fraction defective • np chart: number of defectives • c chart: number of defects Advantages of Control Charts • Focuses attention on detecting and monitoring process variation over time • Distinguishes “special” from “common” causes • Helps predict performance of a process • Helps improve a process to perform consistently • Provides a common language to discuss process behavior Steps to Build a Control Chart • Select the process to be charted. • Determine type of chart. • Determine sampling method and plan. • Initiate data collection. • Calculate control limits and centerline. • Build the control chart. • Plot data. • Interpret results. Example of a Control Chart Point out-of-control (out of the control limits boundaries) Data plot Upper control limit 3 std.deviations Centerline 3 std.deviations Lower control limit Scatter Diagrams • Chart used to identify the possible relationship between two process characteristics • Advantages of the scatter diagram • Provides visual and statistical means to test the strength of a potential relationship • Provides a good follow-up to the cause and effect diagram to find out if there is a connection between the cause and the effect Steps to Build a Scatter Diagram • Collect 50 – 100 paired samples of data. • Construct a data sheet. • Draw the Y-axis and the X-axis of the diagram. • Plot the data on the diagram. Example of a Scatter Diagram Variable 1 Summary • The seven tools of quality discussed in this module are considered to make up the fundamental continuous improvement toolkit. • It is the intent of this course to touch basis on these tools and not to study them in depth. • Depth analysis of these theories is considered subject for another course.