Chapter 9 - Key Terms

Design of Experiments:

Family of statistical tools designed to build quality into the product and process designs so the need for inspection is reduced. DOE achieves this by optimizing product and process designs and by making the product and process designs robust against manufacturing variability. Experimental designs are used to identify or screen important factors affecting a process, and to develop empirical models of processes. DOE techniques enable teams to learn about process behavior by running a series of experiments. The goal is to obtain the maximum amount of useful information in the minimal number of runs.

RACI Matrix:

Identifies the roles and responsibilities for all stakeholders involved in any process; the activity of creating a before and after RACI matrix is called responsibility charting or RACI analysis. Benefits include fewer misunderstandings, less time wasted in meetings, increased productivity and capacity, fewer disputes about responsibilities and authorities, and better ownership of roles and responsibilities. Responsible, Authority, Consulted, Informed.

Process Capability and Performance:

Lean sigma methodology that measures the ability of a process to consistently meet quality specifications.

DMAIC:

Lean sigma problem solving approach: 1. Define the problem, scope, metrics, team and sponsor 2. Measure: Collect data and then measure and map the "as-is" state 3. Analyze: Identify the system of causes and develop and test hypotheses (solutions, countermeasures). Select the best set of solutions, including just do it, implement now, and implement later solutions. 4. Improve: Implement the best solutions 5. Control: "Sustain the gains" by creating new roles and responsibilities, standard operating procedures, job descriptions, metrics, and reviews. Also, share learning and identify potential projects.

Acceptance Sampling:

Methods used to make accept/reject decisions for each lot (batch) based on inspecting a limited number of units. a. Attribute Sample Plans: accept/reject decisions based on a count of the number of units in the sample that are defective or the number of defects/unit. b. Variable Sample Plans: accept/reject decisions are based on measurements.

Sigma level:

Metric that measures defect rate for a process in terms of the standard normal distribution with an assumed shift in the mean of 1.5 standard deviations. Sigma level metric is a measure of effectiveness for a process.

Failure Mode and Effects Analysis:

Process that identifies the possible causes of failures (failure modes), scores them to create a risk priority number, and then mitigates risk starting with tehe most important failure mode. Failure modes must be scored on three dimensions: severity, occurrence, and detection. Scored on 1 to 10 scale where 1 is low and 10 is high. 3 scores are then multiplied to produce Risk Priority Number (RPN). Intended to be proactive tool.

Supplier Qualification and Certification:

Programs designed by purchasing organizations to test if suppliers can meet certain standards.

R-Chart:

Quality control chart that monitors the range (variability) of a process. A sample of n parts is collected from the process every so many parts or time periods. The range of the sample is plotted on the control chart. If sample range < specification limits, the process is said to be under control.

Attribute:

Quality management term used to describe a zero-one (binary) property of a product by which its quality will be judged by some stakeholder. Inspection can be performed by attributes or by variables. Inspection by attributes is usually for lot control (acceptance sampling) and is performed with a p-chart (to control the percent defective) or a c-chart (to control the number of defects). Inspection by variables is usually done for process control and is performed with an x-bar chart (to control the mean) or an r-chart (to control the range or variance).

PDCA Cycle:

4-step approach for process improvement. 1. Plan: Recognize an opportunity and plan the change. Plan to improve operations by first finding out what things are going wrong and then generate ideas for solving these problems. 2. Do: Test the change. Make changes designed to solve the problems on a small or experimental scale first. 3. Check: Review the test, analyze the results, and identify learning. Use data to determine if the change was effective in reducing variation. 4. Act: Take action based on what you learned in the check step. If change was successful, implement the change and identify opportunities to transfer the learning to other opportunities for improvement.

Statistical Quality Control (SQC):

A set of statistical tools for measuring, controlling, and improving quality.

Common Cause Variation:

A statistical process control term for natural or random variation that is inherent in a process over time and affects the process at all times. Common cause variation includes the normal, everyday influences on a process. If a process is in control, it only has common cause variation. This type of variation is hard to reduce because it requires change to the fundamental process. Referred to "chronic pain." Alternative to common cause is special cause.

Six Sigma (Lean Sigma):

A formal process improvement program that combines six sigma and lean thinking principles. Five views of lean sigma: 1. Metric View: Lean is about maximizing the sigma level (minimizing the defect rate or some other process capabilities metrics) often with the target of six sigma 2. Tool View: Applying managerial tools (brainstorming) and statistical tools (design of experiments) to problem solving 3. Project View: Defining and executing lean sigma projects with black/green belt project leaders using the DMAIC 5-step problem solving methodology. 4. Program View: Program management office that finds and prioritizes problems that need to be addressed and then charters and resources projects to address those problems. 5. Philosophy View: About building a sustainable culture of leaders who are relentless about continuously improving processes to eliminate waste and defect.

Control Chart:

A graphical tool used to plot the statistics from samples of a process over time and keep the system in control. If all points are within the upper and lower statistical control limits, variation may be ascribed to "common causes" and the process is said to be in control. If points fall outside the limits, it is an indication that "special causes of variation are occurring and the process is said to be out of control. Eliminating special causes and reducing common causes can improve quality. X-bar is an example of a control chart.

Pareto Chart:

A historical bar chart that helps identify and prioritize the most common sources of errors or defects. A pareto chart highlights the important few by displaying the frequencies for the causes of the problem, sorted from highest to lowest.

X-bar Chart:

A quality control chart that monitors the mean performance of a process. A sample of n parts is collected from a process at regular intervals (either time intervals or production quantity intervals). The mean of the sample is plotted on the control chart, and the process is evaluated to see if it's under control.

p-chart:

A quality control chart used to monitor the proportion of units produced in a process that are defective; a unit is considered defective if any attribute of the unit does not conform to the standard.

SIPOC Diagram:

An acronym for suppliers, inputs, process, outputs, and customers which is a tool used to identify all relevant elements of a process for the purposes of process improvement. Cause and Effect Diagram (causal map): Graphical tool often used for identifying the root causes of a problem. Shows C&E relationships in a system. Can add value to organizations in many ways: 1. Process improvement and problem solving 2. Supporting risk mitigation efforts 3. Gaining consensus 4. Training and teaching 5. Identifying the critical metrics

Pugh Matrix:

Decision tool that facilitates a disciplined, team-based process for concept generation, evaluation, and selection. The Pugh Matrix is a scoring matrix that defines the important criteria for a decision, defines the weights for each criterion, defines the alternatives, and then scores each alternative. The selection is made based on the consolidated scores. Allows organization to compare different concepts, create strong alternative concepts from weaker concepts, and arrive at the best concept, which may be a variant of other concepts.

Assignable Cause (Special Variation):

Deviations from common values in a process that have an identifiable source and can eventually be eliminated; also known as assignable cause. Causes of variation that are not inherent in the process itself but originate from out of the ordinary circumstances. Special causes are often indicated by points that fall outside the limits of a control chart.

Statistical Process Control (SPC):

Set of statistical tools that can be used to monitor the performance of a process. Implemented using graphical methods called control charts. Control limits are based on the process standard deviation (sigma) and are often set at plus and minus three sigma. Unlike control limits, specification limits are not dependent on the process in any way. Specification limits are the boundary points that define the acceptable values for an output variable of a particular product characteristic. Specification limits are determined by customers, product designers, and management. Specification limits may be 2-sided, with upper and lower limits, or 1-sided with either an upper or a lower limit.

Total Productive Maintenance:

Systematic approach to ensure uninterrupted and efficient use of equipment; also called total productive manufacturing. Focuses on preventative and predictive maintenance rather than only emergency maintenance. Indications that TPM may be needed include frequent emergency maintenance events, long downtimes, high repair costs, reduced machine speeds, high defects and rework, long changeovers, high startup losses, high Mean Time to Repair, and low Mean Time Between failure. Benefits include reduced cycle time, improved operation efficiency, improved OEE, improved quality, and reduce maintenance cost.

Acceptable Quality Level:

The maximum percentage defective that can be considered satisfactory as a process average. When deciding to accept a batch, a sample of n parts is taken from the batch and a decision is made to accept the batch if the percentage of defects is less than the AQL.

5 Whys:

The practice of asking "why" many times to get beyond the symptoms and uncover the root cause of a problem. Problems rarely have only one cause and assuming a problem has only a single root cause can prevent investigators from finding the best solution. Root cause analysis should focus on finding and fixing the system of causes for the problem rather than finding someone to blame.

Inspection:

The process of checking units to make sure they are free from defects. Can be done for batch control (accept/reject batch) or process control (check if process is in control). Ideally inspection is performed at source so the process has immediate feedback and a sense of ownership. Should also be done just before bottleneck so valuable bottleneck time is not wasted on defective parts. a. Source Inspection: operator checks his or her own work b. Successive Check: next person in the process c. Final Inspection: final product is inspected

Root Cause Analysis (RCA):

Tool used to identify the contributors to an adverse event after the fact. Purpose is to identify what caused the event and then improve the system so the problem does not reoccur. 5 whys and pareto analysis are useful starting points.