Key TQM Tools -Check Sheet An organized method of recording data -Scatter Diagram-A graph of the value of one variable vs. another variable-Cause-and-Effect Diagram-Pareto Chart-- Identifies and plots problems in descending order of frequency-Statistical Process Control-Uses statistics and control charts to learn systematically (rather than haphazardly) about process output-Since variability is inherent in every process, seeks to identify and differentiate between common-cause (random) and special-cause (non-random) process variation à when to take corrective action or not-Systematic process improvement through elimination of special causes and gradual reduction of common causes -Constructed from historical data, the purpose of control charts is to help distinguish between common-cause and special-cause variation-There are many types of distributions, including the normal (bell-shaped) distribution, but distributions differ in terms of central tendency (mean), standard deviation or variance, and shape-If only common causes of variation are present, the output of a process forms a distribution that is stable over time and is predictable-If special causes are present, the process output is not stable over time and is not predictable-Which of the following is a TQM tool? Ishikawa diagram-Lean?? Andon cord, kanban card, heijunka plot, shitsuke cheetscompute LCL & UCL for eachWhen we know σ, the population variation, either from historical data orbased on a given targetL C Lx=x-zxU C Lx=x+zx x is the mean of the sample means or a target value set for the processz is the number of normal standard deviations sx is the standard deviation of the sample means = n is the sample sizeWhen we do not know σ, the population variation: L C Lx=x-A2RU C Lx=x+A2RSetting limits for the R (variation) chart – regardless of whether σ is known or not!: L C LR=D3R U C LR=D4R , whereå==niiRnR1/1 Level Scheduling (Heijunka)Goal: achieve a smooth, stable production flow. Why would this be desirable?Idea: process frequent small batches rather than a few large batches à a.k.a. mixed-model (“jelly-bean”) scheduleJIT Layout - Group technologyGroup similar products into familiesGroup processes in work cellsX-bar & R charts, when population variance is known or unknownControl Charts-A variety of charts can be constructed, depending on the metric that is measured-Continuous variables (length, weight, thickness, etc.) – X-bar and R chartsc charts – if dealing with counts of defects, p charts – if interested in the proportion of defects u charts – if interested in the number of defects per unit of output Multivariate charts – if interested in tracking multiple dimensions simultaneously X-bar charts – track changes in central tendency (i.e., the mean) of a processR charts – track changes in process variability-Variability in the weight of cereal boxes is tracked with? R charts Must be used together, as both the mean and the variability are critical metrics that need to be tracked to assess process performance xbar chart has no points outside l/ucl, is process in control? No, abnormal patterns might be present, must check rchart alsoTQMRefers to a consistent and comprehensive focus on quality that encompasses the entire organizationQuality is defined by identifying and meeting customer expectationsManagement is directly responsible for quality improvementRequires a continuous effort to analyze and improve organizational processesMajor elements-Continuous improvement Never-ending process of continual improvement à the goal is zero defects Covers people, equipment, materials, procedures Based on Shewhart/Deming PDCA model-Some additional principles-Employee empowerment, Internal and external benchmarking6sigma A comprehensive program focused on total customer satisfaction, including aDMAIC model: Define à Measure à Analyze à Improve à ControlSet of tools, e.g., Statistical Process Control (SPC), Pareto charts & histograms, flowcharts, cause-and-effect diagrams, etc. Define – project’s purpose, scope, and outputs subject to the customer’s definition of qualityMeasure – the target process and collect data Analyze – the data ensuring repeatability and reproducibilityImprove – by modifying or redesigning existingprocesses and proceduresControl – the new process to makesure performance levels are maintainedThere is absolutely no reason for having errors or defects in any product or service, as quality is free according to Philip Crosbycost of poor quality is typically underestimated à should include all the consequences of not building the right product or delivering the right service the first timea focus on quality is required regardless of the organizational strategy Determining the Number of Kanban cards (containers) neededeach container represents the minimum production lot size and the number of such containers determines directly the amount of WIP in systemK = (Expected demand during lead time + Safety Stock) / Size ofthe container = (DL+S)/CD = average demand per period L = lead time to replenish an order S = safety stock expressed in days (hours, etc.) of demand C = Container size Little’s Law=WIP = THt x THr, whereWIP – represents amount of work in processTHt – is the throughput time (i.e., time in system)THr – is the throughput rate Defines the output rate that a process is expected to produce over a period of time (units/time)Represents a measure of process capacityEquals the inverse of the process time of a stationLean Production- Respect for people-Cross-functional teamsPreventive maintenance-Periodic inspection and repair Standard Work-Use clear, standardized, consistent task specs Value-chain mapping-Map processes and identify value-added and non-value-added activities Continuous improvement (Kaizen)Quality at the source – andon cordUpgrade housekeeping (5S)-Seiri (Sorting)Seiton (Simplify)Seiso (Shine/sweep)Seiketsu (Standardize)Shitsuke (Sustain/self-discipline)What is a common cause of variation? Random sources (Cost of Quality)Prevention costs- employee trainingAppraisal costs- quality testingInternal failure costs-scrapExternal failure costs-returned goodsEliminating wasteReduce inventories of raw materials, WIP and finished products (overproduction)Reduce defects and rework due to poor qualityReduce waiting due to resource or part unavailabilityReduce unnecessary transportation and motionReduce overprocessing – work that