LEAN/SIX SIGMA TOOLS DEFINITIONS




5 Whys:  A simple yet effective tool that can be used to identify the root cause of an issue by asking the question “Why?” five times.  The use of this tool will help to develop increased understanding of an issue and provide a forum to challenge current procedures.  

5S Work Place Organization (5S):  A 5 step process for achieve and maintain a clean and well organized workplace.  The 5 S’s are: Sort, Store, Shine, Standardize and Sustain.  

7 Types of Waste:  Production of defects, over processing due to rework or poor process design, overproduction by producing before there is demand, unnecessary movement of employees, unnecessary transportation of materials, holding more inventory than the absolute minimum and waiting or delays which interrupt flow.  

Affinity Diagram:  A tool used to generate and organize ideas or issues into common themes.  Once appropriately grouped, a single statement can summarize the idea or issue for each group.  These statements can then be ranked in order of priority for further analysis or resolution.  

Andon Lights:   A set of signal lights associated with a piece of equipment or a work cell that indicate its operational status.  An andon light can alert workers and management to a variety of conditions including:  need for part replenishment, quality issues or operation at less than expected rate.  The term and on is also associated with visual displays which provide workers with process performance information such as defect rate or throughput. 

Attribute Data:  Data that can be divided into various groups or categories on the basis of some non-numerical characteristics.  This term is often used to describe data reflecting conformance or non-conformance to specifications, including observation of the presence or absence of some quality characteristic.  

Average:  A measure of the central tendency of a sample or the population.  Synonymous with mean.

Bell Curve:  A bell shaped curve that graphically describes the probability distribution for “normal” data.

Binomial Distribution:  The probability distribution which describes a process or experiment with only two possible outcomes (pass/fail, defective/not defective, presence/absence, etc.).  

Bottleneck:  The rate limiting step of a process when demand exceeds capacity.  Continually improving the performance of the bottleneck should improve the output of the whole operation until it has been elevated to the point where it is no longer the rate limiting step.  

Brainstorming:  A group technique used to generate many ideas about a specific topic, issue or solution.  It is particularly good at inspiring creativity and synergy between the participants.  

Capability (of a process):  A measure of the quality of a process (and its output) that is derived through a comparison of the process variation to the process specification limits.  Capability measures include sigma capability, Cp , Cpk and defects per million (dpm).   

Cause and Effect Diagram (CE):  A pictorial tool that is used to categorize, display and examine potential causes or contributing factors (inputs) related to a specific observed condition (output).  This tool is also known as a fishbone diagram or and Ishikawa diagram.  

Central Tendency:  A measure of the point about which a group of values is clustered.  Examples include:  mean (average), median and mode.  

Changeover:  All of the activities associated with switching the materials, operating settings or tooling on a piece of equipment so that it can produce a different part or perform a different task.  Changeover time is usually defined as the time elapsed between the production of the last good part of one batch until the first good part is produced of the next batch.  

Common Cause Variation:  The sources of variability in a process that are truly random and occur naturally as an inherent part of the process.  

Continuous Variable:  A variable that is derived from measurement data (time, weight, pressure, linear measure etc.) and is considered to have an infinite number of possible values.  

Control (of a process):  A process is considered to be in a state of statistical control with respect to a specific quality measure, when it exhibits only random (common cause) variation.  A controlled process is consistent and predictable, but the state of control does not imply anything about the quality of the process or its output.  

Control Limits:  A pair of boundaries (upper and lower) that demarcate the region of natural variation (common cause variation) on a process control chart.  The position of these boundaries is set according to the amount of variation in the process, typically ±3 standard deviations from the center line.  

Correlation Coefficient:  A measure of the linear relationship between two random variables.  

Cost of Poor Quality (COPQ):  The cost of the “waste” associated with all of the processes (administrative and production oriented) within a business.  This includes costs related to internal and external failures, appraisal, prevention and lost opportunities.  Often associated with any activity related to not doing the right thing right the first time.  

Cp:  A capability index which describes the best potential capability of a process, i.e. when it is centered between the specification limits.  

Cpk:  A capability index which compares the natural tolerance of a process to its specification limits.

Cycle Time:  The elapsed time for a product to progress completely through its process, i.e. from the start of the first step until the end of the last step.  The term cycle time can also be applied to an operator or a piece of equipment.  Operator cycle time is the time required for an operator to complete the full sequence of his or her assigned tasks and return to the start position.  Similarly, equipment cycle time is the time required for a machine to complete the full sequence of operations associated with its production task.  

Defects per Million (dpm):  A capability measure which describes the number of defective units per million units produced.  

Design for Six Sigma (DFSS):  A series of Six Sigma tools which are applied during the early phases of the product development cycle with the intention of reducing variability and “building the quality in” rather than improving the process or product later.  

Design of Experiments (DOE):  An organized method of collecting data regarding a process by purposefully changing the process inputs in order to observe and measure the corresponding changes in the process output.  DOE provides a method to develop powerful empirical models which approximate the true relationships between various process inputs and the output.  Understanding these relationships allows us to improve the process performance characteristics.  

Discrete Variable:  A variable that is based upon “count” data (the number of defects, the number of births, the number of units produced etc.).  Although in some cases the numbers may be very large, they are finite and considered to be “countable”.  

Error Proofing (Poka-yoke):  A series of techniques which are used to reduce the opportunity for errors to occur.   

Failure Modes and Effects Analysis (FMEA):  A tool which is used to systematically identify, analyze, prioritize (on basis of a risk assessment) and reduce or eliminate potential process or product failure modes.  

Flow:  The continuous forward movement of a product or service through each of its process steps.
 
Histogram:  A vertical bar graph which illustrates the distribution of numerical data according to the frequency of occurrence within defined value ranges (classes).  

Input Process Output Diagram (IPO):  A pictorial tool which is used to describe all of a process’ inputs (e.g. people, policies, methods, equipment, materials or environment) and its outputs (e.g. producing a product, providing a service or performing a task).  

Just In Time (JIT):  A manufacturing strategy which strives to rapidly respond to the pull of the customer and to reduce inventories by producing the exact number of units required, at the right time with no waste and continuous flow.  

Kaikaku:  Radical improvement or transformation of an activity to reduce or eliminate waste.  

Kaizen:  Continuous small (incremental) improvements in an activity over time which reduce waste.
 
Kanban:  A Japanese term which means “card signal”.  However, this term is often used more generally to describe any visual signal used to indicate the need for material replenishment by an upstream activity.  

Knowledge Based Management (KBM):  A management philosophy predicated on good decision making using knowledge and data rather than opinion and perception.  It is composed of three elements:  the Questions that Managers Need to Answer, Questions Managers Need to Ask and the Lean Sigma tools and methodologies.  

Lean Production:  The principle philosophy of lean production is the elimination of waste, particularly as defined by the 7 types of waste.  However, in a broader context, lean production refers to many of the techniques and methods described in this guide which contribute to the company’s ability to respond quickly and efficiently to customer demand by delivering high quality products when and where needed.

Lean Sigma:  A business philosophy which combines the strategies of lean production (reduction of waste and queue times) and six sigma (reduction of variation).  These concepts must be applied to all facets of the business in order to achieve a truly “lean enterprise”.  

Load Chart:  A vertical bar chart which compares the workload, in units of time, assigned to each operator or piece of equipment in a work cell.  The cycle times within the cell should be balanced so that they are approximately equal thereby minimizing idle time and delays.  The overall process cycle time should be equal to the takt time.  

Mean:  A measure of central tendency of a sample or the population.  It is calculated by adding all of the values within the data set together and dividing by the number of values in the set.  The mean of a sample is usually represented by a variable such as  (read “x bar”) and the Greek symbol m is used to represent the average of the population.  Synonymous with average.  

Measurement System Analysis (MSA):  A statistical technique that quantifies the amount of the variability that originates from the measurement system itself.  In this context, the measurement system includes the operator and equipment used to perform the measurement.   Muda:  Any activity which is wasteful, i.e. it uses time or resources, but does not add any value from the perspective of the customer.  

Multiple Regression:  A mathematical model where several independent variables are used to predict the value of one dependent variable.  

Multi-Voting:  A group technique used to prioritize long lists of activities or issues.  

Muther’s Grid:  A tool used to systematically organize the interrelationships amongst a group of activities to determine the optimal physical layout of the work area.  Also known as a relationship diagram.  

Nominal Group Technique:  A group technique for prioritizing relatively short (usually no more than 10) activities, ideas or issues.  

Non-Value Added (NVA):  Activities which consume time or resources but do not directly contribute towards meeting the customer’s requirements.  

Normal Distribution:  A probability distribution, also known as the Gaussian distribution, which is graphically represented by a smooth bell shaped curve.  

Outlier:  A data point which is very unusual when compared to the rest of the data set.   Overall Equipment Effectiveness (OEE):  A measure of equipment performance in terms of downtime, throughput relative to design specifications, and the quality of its output.  

Pareto Charts:  A vertical bar chart that relates various attribute (non-numerical) categories to their respective cost or frequency of occurrence.  

PCOR:  The four strategies of process improvement:  Prioritize, Characterize, Optimize and Realize.  

PF/CE/CNX/SOP:  A methodology that is used to reduce extraneous variation.  It uses process flow diagrams (PF) and cause and effect diagrams (CE) to identify and sort the causes of variation.  The causes are further categorized as controllable (can be held constant “C”), noise (too difficult or expensive to control “N”) or experimental (factors which must be investigated to understand their impact “X”).  Standard operating procedures (SOP) are used to establish methods for holding the “C” factors constant.  

Physical Process Map:  A layout diagram of the work area that illustrates the path followed by materials or parts through the facility.  It can be used to highlight excessive amounts of material transport or employee movement while following the steps described in a process flow chart.  Synonymous with spaghetti diagram.  

Poisson Distribution:  A probability distribution often used to describe the number of occurrences per unit interval of time or space.  

Poka-Yoke:  Any technique(s) which is used to reduce the likelihood of and error (e.g. color coding, unique design characteristics which prevent improper assembly).  These techniques are best applied in the design phase of a process to prevent errors but can also be implemented as a corrective action to prevent recurrence of an error.  

Population:  The set of all objects or individuals to be studied, or the data which represents some characteristic of those objects or individuals.  

Probability Distribution:  A table, graph or formula that describes the probabilities of all of the possible outcomes of an experiment.  Many naturally occurring phenomena and processes can be described by a probability distribution.  Knowing the characteristics of the distribution that is associated with a specific experiment, process, or phenomena, allows us to make accurate predictions regarding the likelihood of certain events related to that experiment, process or phenomena.  

Process Capability:  A measure of the quality of a process (and its output) that is derived through a comparison of the process variation to the process specification limits.  Capability measures include sigma capability, Cp , Cpk and defects per million (dpm).   

Process Control Chart:  A fundamental tool of statistical process control (SPC) which is used to determine if a process is statistically under control i.e. stable and predictable within the bounds of natural variation.  

Process Flow Chart:  A diagram that describes in sequence all of the major steps and decision points of a process.  It improves understanding of the process and can facilitate the identification of problem areas.   

Process Observation Chart:  A table used to record specific attributes of each step while carefully watching the process.  The attributes of interest usually include the type of process step (e.g. operation, transport, inspection, waiting, storage or decision), quantity, distance and elapsed time.  

Process:  A series of activities, administrative or production oriented, which are usually performed in a particular sequence to accomplish a specific objective.  

Pull Production:  A production strategy which requires downstream process steps to signal a need for replenishment, before activity in each of the previous (upstream) process steps is initiated.  In the ideal pull system, no production begins until the customer purchases the product, and even then, it is only produced in the quantity consumed.  The customer demand “pulls” the product through the process. 

Push Production:  A batch and queue production strategy which produces product by controlling the input i.e. the “front end” of the process.  When orders are placed on the system, often based on forecast rather than true customer demand, batch production is initiated at the first step to meet and usually exceed the demand.  

Quality Function Deployment (QFD):  A systematic process used to translate and integrate the “voice of the customer” (performance requirements or design attributes) into the development of products or services that will meet those specifications.  

Random Sample:  A sample drawn from the population in such a way that every element of the population had an equally likely chance of being selected.  

Random Variable:  A definition of the possible outcomes of interest from a given experiment.  It converts the experimental outcome to a numerical variable that can be analyzed with statistical techniques.  

Range:  A measure of the variability or “spread” of a data set.  It is the difference between the maximum and minimum values in the data set.   

Reality Tree:  A tool used to analyze cause and effect relationships in order to identify root causes.

Regression Analysis:  A statistical technique for determining the mathematical relationship between a measured quantity and the variables it depends upon.  

Relationship Chart:  A tool used to systematically organize the interrelationships amongst a group of activities to determine the optimal physical layout of the work area.  Also known as a Muther’s grid. 

Rule of Thumb (ROT):  A simplified, practical procedure or guideline which can be used in place of a formal statistical test and will produce approximately the same result.  

Run Chart:  A graphical tool that sequentially charts a process quality measure over time.  

Sample:  A set of objects, individuals or values selected from the population.  

Scatter Diagram:  A graphical illustration representing the relationship between two variables, by charting ordered pairs data (x, y data).  

Sigma Level:  A commonly used measure of process capability that represents the number of standard deviations between the process center (the mean) and the closest specification limit.  

Sigma:  A Greek letter (s) used to represent the standard deviation of the population.  

Simple Linear Regression:  A mathematical model where one independent variable is used to predict the value of one dependent variable.  

Single Minute Exchange of Dies (SMED):  A series of techniques used to facilitate the rapid changeover of any process or equipment.  

Single Piece Flow:  A production strategy in which units progress individually (as a batch of one unit) through each of its associated process steps with continuous flow in the forward direction (i.e. without rework).  It can be contrasted with the batch and queue strategy in which batches of multiple units proceed as a group through each of the production steps.  This procedure requires each unit to “stop and wait” for each of the other units in the batch to be completed before the group proceeds to the next process step.  

Six Sigma:  A quality improvement and business strategy that began in the 1980’s at Motorola.  Emphasis is on reducing defects, reducing cycle time with aggressive goals, and reducing costs to dramatically impact the bottom line.  

Spaghetti Chart:  A layout diagram of the work area that illustrates the path followed by materials or parts through the facility.  It can be used to highlight excessive amounts of material transport or employee movement while following the steps described in a process flow chart.  Synonymous with Physical Process Map.  

Special Cause Variation:  Non-random causes of variation that can be detected by the use of control charts and good process documentation.  A process is said to be in a state of statistical control when all sources of special cause variation have been eliminated.  

Standard Deviation:  A commonly used measure of variability or “spread” in a data set or population.  It is equal to the square root of the variance.  

Standard Operating Procedure (SOP):  An up to date written procedure that clearly and concisely describes the exact method to be followed in order to complete a specific task.   

Standard Work:  A method of improving work efficiency and reducing variability by ensuring that everyone follows exactly the same procedure to complete a specific task.  The procedure may be improved as new ideas or technology become available.  However, the modified procedure must first be demonstrated to be superior and if adopted as the best practice, everyone who is performing that task must follow it.  

Statistical Process Control:  The use of graphical and statistical methods to analyze and improve a process by reducing its variation and increasing its capability.  

Subgroup:  A sample of units all drawn from a process at, or near, the same time.  Data collected from measurements of the units are used to plot control charts.  

Takt Time:  The average time to produce a unit in order to meet the current rate of customer demand.   t-distribution:  A symmetrical, bell shaped, distribution that is similar to the normal distribution except that it has a larger portion of its area in the tails, i.e. it has greater variability.  It is typically associated with small samples (less than 30 units) and is also referred to as the Student t-distribution.  

Time Value Map:  A time scaled graph that illustrates the amount of active and inactive time spent during one complete cycle of a process.  

Total Productive Maintenance (TPM):  A comprehensive and coordinated maintenance program designed to maximize equipment effectiveness by minimizing downtime and optimizing output in terms of speed and quality.  This approach is founded upon the ability of well trained equipment operators to proactively identify and correct small maintenance issues before they become significant and lead to breakdowns.  

Value Stream Mapping:  A diagram which describes all of the activities in a product value stream.

Value Stream:  All of the activities which a required to progress a product from concept development all the way through to delivery to the customer.  

Value:  Some feature, condition, service or product that the customer considers desirable and which is delivered to them when and where they want it.  

Variables Data:  A measurement whose value is only limited by the sensitivity and resolution of the measuring system.  Opposite of attribute data.  

Variables:  values which are subject to change or variability.  

Variance:  A measure of the variability or “spread” in a data set or population.  It is equal to the square of the standard deviation.  

Visual Management:  A series of visual techniques which are used to communicate the status of a system in such a way that it can be understood at a glance by everyone concerned.  The term “system” is used broadly to describe everything from simple conditions such as the presence or absence of the tools required to perform a specific task (e.g. a changeover system) or more complex operations such as graphs reporting key performance indicators for a business.  

Voice of the Customer:  A term which describes any method of collecting information about what is important to the customer.  It includes but is not limited to:  customer specifications, customer design requirements, customer surveys and listening to customer feedback.  

Waste Analysis:  An examination of a process that is conducted to identify and separate value added activities from non-value added activities (waste).   

Waste:  Any activity or process that does not add value to the product from the perspective of the customer.  

Work Cell Design:  A technique used to develop the optimal physical layout of the work area to improve the worker’s ability to maximize product throughput and quality.  

Work Cell:  A work area, often including machines, which has been designed to facilitate production activities by one or more operators.  The physical layout and activities in a cell are very organized and carefully sequenced to optimize flow of materials with minimal effort.  

Work In Process (WIP):  Material which has been partially processed, but has not yet reached the state of approved finished product.  

Yield:  The proportion of “good” units produced (output) relative to the theoretical maximum number of units which could be produced on the basis of material input quantities.  Yield is usually expressed as a percentage of theoretical maximum, but is also sometimes described in terms of absolute quantities (e.g. kilograms of product).  

Z-Distribution:  A standardized normal distribution or random variable having a mean of zero and a standard deviation of one.    Z-value:  A standardized value calculated by subtracting the mean and then dividing this difference by the standard deviation.  It represents the number of standard deviations between a specified value and the mean.