Select Measures The Quality Improvement Model Define Process

Select Measures The Quality Improvement Model Define Process Select Measures Collect & Interpret Data Select Measures Purpose: • Develop understanding of customer needs. • Determine ‘possible’ process response control points. Is Process Stable ? No Investigate & Fix Special Causes Yes Improve Process Capability No Is Process Capable ? Yes Use SPC to Maintain Current Process 3 -1

Select Measures Importance of Measures Outputs Customers Process Products or Services Need Measures of Process Performance l To learn about the process. l To identify improvement opportunities. l To monitor performance of the process. 3 -2

Select Measures Use of Measures Products or Services Process Customers Measurement Process Dealing with the Future Dealing with the Past Process Control and Improvement Product Control and Disposition What Does This Information Tell Us About The Process? Make The Decision And Move On 3 -3

Select Measures Effective Measures 3 -4

Select Measures Example Measures Typical Measures of Process Performance Manufacturing Product characteristics (weight, dimensions, color, chemical composition, appearance, etc. ), % not meeting customer specifications, etc. Administrative Number of errors, time to complete activity, percent not meeting requirements, etc. Service time, delivery time, number of complaints, etc. Management Number of injuries, difference between amount budgeted & actual expenditures, etc. 3 -5

Select Measures Types of Measures l l Counting Measures § Count the number of items that do not conform to some standard. § Count the number of nonconformities in a given area, volume, or time period. Instrument Measures § Can count the good and bad. Can count the bad only! Quality characteristic is measured by some type of measuring instrument (weight, time, size, chemical analysis, dollars, . . . ) 3 -6

Select Measures Developing Measures l Understand the customer's needs, expectations, and wants. l Use process map to identify key points for measures. l Develop & document the measurement process. l Develop a process for collecting, recording, and analyzing data. l Try the measure. 3 -7

Select Measures Selecting Measures Outputs Customers Process Products or Services ? Gauge In-Process Final Product Gauge Customer Satisfaction l Where can we measure the process? l What are pros and cons of the above locations? l Where should we start? 3 -8

Select Measures Process Variables Suppliers Inputs (Raw Materials) Process Outputs (Product) Customers Input Properties Uncontrolled / Response / Output Product Controlled Variables Properties Process Variables -- All physical or analytical measurements related to a process. Measurements can be electronic, manual observations, or laboratory results. Measurements can be in-process or final product oriented. Process variables can be broken down in to several different types: • Input Properties - Inputs • Uncontrolled / Controlled Variables - Inputs • Response Variables – Outputs (Secondary) • Product Properties – Outputs (Primary – See on Big Block/SIPOC) 3 -9

Select Measures Input Properties Inputs (Raw Materials) Suppliers Process Outputs (Product) Customers Input Properties Uncontrolled / Response / Output Product Controlled Variables Properties Process Variables Input Properties (Inputs)-- Properties of inputs to the process that can be controlled through contact with supplier. Examples: • Manufacturing - Purity of raw materials, composition of mixes, dope solids, dope viscosity • Lab - Purity of reagents, brand/model of instrument • Maintenance - Condition of tools, size/composition of gaskets used in repair • Administrative - Number of mistakes in data 3 -10

Controlled/Uncontrolled Variables Suppliers Inputs (Raw Materials) Process Outputs (Product) Select Measures Customers Input Properties Uncontrolled / Response / Output Product Controlled Variables Properties Process Variables Controlled Variables (Inputs)-- Variables held at a set point or within a range (standard operating conditions), either manually or by an automatic algorithm (EPC). Set Point -- The desired level of a controlled variable. Uncontrolled Variables (Inputs)– Variables that do not have a control system. Can only be monitored. Examples: • Manufacturing - Feed rate of mix, reactor temperature settings (C), outside temperature (U), pump shaft speed (C), metering pump speed (C) • Lab - Hot plate temperature setting, time to cook a sample (C) • Maintenance - Alignment gauge settings, lathe settings (C) • Administrative - Procedures followed (U) 3 -11

Select Measures Response/Output Variables Suppliers Inputs (Raw Materials) Process Outputs (Product) Customers Input Properties Uncontrolled / Response / Output Product Variables Controlled Variables Properties Process Variables Response Variable (Outputs) – an analytical or measured variable that is used to control or monitor the process but is not a measurement on the final product. Time Lag -- A delay between expected effects of variables. Example: 20 minutes after increasing a temperature (controlled input variable), the reactivity (output variable) begins to increase. Examples: • Manufacturing - Composition before distillation, early stage viscosity • Lab - Visual check for dissolution during heating, final temperature of heated solution • Maintenance - Vibration of pump during operation, micrometer readings of machined part • Administrative - Budget variance during the project 3 -12

Product Properties / Output Variables Suppliers Inputs (Raw Materials) Process Outputs (Product) Select Measures Customers Input Properties Uncontrolled / Response / Output Product Controlled Variables Properties Process Variables Product Properties (Outputs) -- Measurements made on the final product that indicate its performance. Important to "Value". Probably will be on Big Block/SIPOC. Target -- The desired state of a response variable or product property. Examples: • Manufacturing - Lab test result on final product (IV, L*, a*, b*, lubricant level, denier, entanglement) • Lab - Lab test result on “known” reference material • Maintenance - Equipment performance in the field, length of time until failure • Administrative - Number of errors in Certificates of Analysis or purchase orders 3 -13

Select Measures External/Input Variables External Variables Inputs (Raw Materials) Suppliers Process Outputs (Product) Customers Input Properties Uncontrolled / Response / Output Product Variables Controlled Variables Properties Process Variables External Variables (Inputs) -- Variables that are not controlled but may have an effect on the process. Typically uncontrolled. Examples: • Manufacturing - Ambient temperature • Lab - Humidity • Maintenance - Vacation scheduling during deer hunting season • Administrative - Employee morale 3 -14

Customer Requirements / Output Variables Select Measures External Variables Suppliers Inputs (Raw Materials) Process Outputs (Product) Input Properties Uncontrolled / Response / Output Product Controlled Variables Properties Customer Requirements Process Variables Customer Requirements (Outputs) -- Fitness-for-use specifications given by the customer. Will be on Big Block/SIPOC or directly influenced by product property outputs. Specifications -- the range of measurements of a product characteristic within which the product is judged acceptable to meet customers’ requirements. Examples: • Manufacturing - Purity above 99. 7% • Lab - Sample Value reported within 2 hours • Maintenance - Repairs completed within 4 hours • Administrative - Errors below 0. 5% 3 -15

Select Measures Three Ways to Control Key Variables in a Process Control to a Standard Statistical Process Control Engineering Process Control 3 -16

Select Measures Statistical Process Control (SPC) • Involves use of control charts, special cause action plans and control strategies to identify and take action on special causes. • Primarily used on response variables and product properties (outputs). 3 -17

Select Measures Engineering Process Control (EPC) • EPC moves variation from where it bothers you to where it doesn’t bother you. (Ex. cruise control) • Automatically makes compensating adjustments at a specified frequency. • Must have reliable measurement, autocorrelated data, and process adjustments with clear cause & effect. • Applied to inputs and outputs 3 -18

Select Measures Control To A Standard • A standard way to perform a task (S. O. P. ). • A standard range that a key process variable should be within. • S. O. P. 's should be audited. • Primarily used in inputs. • Fundamental backbone of S. P. C. • MUST be done first! 3 -19