|
BACK TO 100 examples in Business, Operations and
Engineering. |
|
Do it once, do it right, and do it now.
|
|
Back to Lawson Computing Homepage
|
Apply as needed,
when needed. |
|
Quality Control, Charts, Methods, People |
CHAP 1
14 POINTS FOR MANAGEMENT
1. CREATE CONTINUOUS PURPOSE TOWARD IMPROVEMENT AND SERVICE.
TO BECOME COMPETITIVE AND STAY IN BUSINESS.
2. ADOPT THE NEW PHILOSOPHY , THE NEW ECONOMIC AGE AND TAKE ON LEADERSHIP FOR CHANGE.
A. PEOPLE ARE BEST MOTIVATED BY SHEER JOY.
B. MANAGE THE PROCESS AND RESULTS.
C. MANAGE SO EVERYONE WINS NOT JUST SHAREHOLDERS.
D. COOPERATION WORKS BETTER THAN COMPETITION.
3. STOP DEPENDENCE ON INSPECTION TO ACHIEVE QUALITY.
A. THIS DOES NOTHING TO MAKE THEM BETTER.
4. END PRACTICE OF REWARDING BUSINESS BY PRICE ALONE.
A. MINIMIZE TOTAL COST BY BUILDING TRUST WITH ONE.
5. IMPROVE CONSTANTLY PRODUCTION AND SERVICE.
A. MGMT. KNOW BETWEEN SPECIAL / COMMON VARIATIONS.
B. MGMT.AGREE ON DEFINITIONS OF OPERATIONS.
C. SDSA=STANDARDIZE, DO, STUDY, ACT.
D. PDSA=PLAN, DO, STUDY, ACT. -DEMING-
E. EMPOWERMENT
6. TRAINING ON THE JOB.
A. EMPLOYEES ARE MOST IMPORTANT ASSET TO FIRM.
B. FORMAL AND THEORY TRAINING.
C. EXPERIENCE W/O THEORY WILL NOT GIVE A PERSON PREDICTIONS ON NEXT EVENTS.
7. INSTITUTE LEADERSHIP
A. MUST BE SOMEWHAT CAPABLE OF PREDICTING FUTURE.
8. DRIVE OUT FEAR.
~SECURITY, PHYSICAL HARM, UNKNOWN GOALS, SHORTCOMING IN HIRING/TRAINING, POOR SUPERVISION, FAIL TO MEET QUOTAS, BLAMED FOR PROBLEMS OF THE SYSTEM, FAULTY INSPECTION METHODS.
A. MANAGEMENT JOB PERIOD.
9. BREAK DOWN BARRIERS BETWEEN DEPARTMENTS.
10. ELIMINATE SLOGANS, TARGETS, ZERO DEFECTS.
A. THESE ARE NOT METHODS.
11.STOP WORK STANDARDS OR QUOTAS ON FACTORY FLOOR.
A. THESE ARE NEGATIVE.
B. HAVE NO BEARING ON THE PROCESS ITSELF.
11B. STOP MANAGEMENT BY OBJECTIVE. (NUMBERS AND GOALS)
SUBSTITUTE LEADERSHIP.
A. GOALS AND TARGETS ARE DYSFUNCTIONAL. DO NOT AID LEARNING.
STEALS PRIDE OF WORKMANSHIP / AIDS THE "EVIL OF MGMT" THOUGHTS.
KNOW THEORY’S ROLE IN DECISION MAKING.
12. REMOVE BARRIERS THAT ROB WORKER OF PRIDE OF WORKMANSHIP.
~ ~UNKNOWN MISSION, NOT USING SKILLS, BLAMING THEM, POOR EQUIPMENT.
NO APPRAISAL SYSTEM / THIS FOCUSES ON SHORT TERM.
13. ENCOURAGE EDUCATION AND SELF-IMPROVEMENT FOR ALL
14. TAKE ACTION TO ACCOMPLISH THE TRANSFORMATION.
IT WILL NOT OCCUR WITHOUT THE EXPENDITURE OF ENERGY.
LEADER MUST KNOW EACH WORKER’S REASON FOR OR AGAINST THE CHANGE , AND HOW EACH OF THOSE INTERACT WITH THE SYSTEM.
DEFINING/DOCUMENTING A PROCESS
EMPOWERMENT AGAIN
EMPLOYEE HAS POWER TO
1. DEFINE/DOCUMENT PROCESSES.
2. LEARN BY TRAINING AND DEVELOPMENT.
3. INNOVATE BEST METHODS AND MAKE UP PROCESSES.
4. USE THEIR OWN JUDGEMENT WITHIN THAT CONTEXT.
5. ENVIRONMENT OF TRUST WITHOUT NEGATIVE FEEDBACK.
SDSA
USING
STANDARDIZE: WITH METHODS SUCH AS FLOWCHARTS.
DO: EXPERIMENT PROPOSED BEST METHODS.
STUDY: RESEARCH THE EFFECTIVENESS OF THESE METHODS.
ACT: MANAGERS ESTABLISH BEST METHODS AND STARTS TRAINING.
PROCESS BASICS:
TRANSFORMATION OF INPUTS INTO OUTPUTS.
TIME VALUE: INPUTS ARE READY ON TIME.
PLACE VALUE: WHERE IT IS NEEDED.
FORM VALUE: AVAILABLE HOW IT IS NEEDED.
A. FEEDBACK LOOP= HAVE IT.
DEFINING AND DOCUMENTING THE PROCESS.
A. WHO OWNS IT/WHO IS RESPONSIBLE FOR IT.
B. WHAT ARE THE BOUNDARIES OF THE PROCESS.
C. WHAT IS THE FLOW OF THE PROCESS.
D. WHAT ARE THE PROCESSES OBJECTIVES,MEASURMENTS.
E. ARE THE PROCESS DATA VALID ?
TYPES OF FLOWCHARTS:
1. SYSTEMS FLOWCHARTS= SEQUENCE OF OPERATIONS
2. LAYOUT FLOWCHART= FLOOR PLAN OF AREA.
ADVANTAGES:
1. COMMUNICATIONS TOOL. EASIER TO CONVEY IDEAS.
2. FUNCTIONS AS PLANNING TOOL/VISUALIZE DATA.
3. PROVIDES OVERVIEW OF SYSTEM.
4. DEFINES ROLES. OF PERSONNEL,OPERATIONS,LOCATIONS.
5. DEMONSTRATES INTERRELATIONSHIPS/ RELATION TO EACH OTHER.
6. PROMOTES LOGICAL ACCURACY=SEE ERRORS EASIER.
7. FACILITATES TROUBLE-SHOOTING/DIAGNOSTIC TOOL
8. DOCUMENTS A SYSTEM
CHANGING SYSTEM WITH FLOWCHARTS.
FIND AREA WITH HIGH REJECTS, FIND AREA WITHIN OWNERS CONTROL,ISOLATE THESE ELEMENTS THAT EFFECT CUSTOMERS.
EQUALS = EXCELLENT TIME TO MODIFY SYSTEM.
PROCESSES OBJECTIVES:
CONTINUOUSLY MONITOR PURCHASING’S CUSTOMERS (MAINTENANCE,ADMINISTRATION) SUCH AS
1. NUMBER OF DAYS FROM REQUEST TO DELIVERY.
2. EASE OF FILLING OUT PAPERWORK.
3. SATISFACTION OF PURCHASED MATERIAL.
CONTINUOUSLY TRAIN AND DEVELOP PURCHASING PERSONNEL.
1. NUMBER OF ERRORS PER PURCHASE ORDER.
2. NUMBER OF MINUTES TO COMPLETE PURCHASE ORDER.
OPERATION DEFINITIONS.
ESTABLISH A LANGUAGE FOR PROCESS IMPROVEMENT.
EVERYONE MUST AGREE ON WHAT WHAT IS.
OPERATIONAL DEFINITION:
OBJECT OR GROUP ?
TEST OF IT.
DID IT MEET THE OBJECT OR GROUP CRITERION.
SUMMARY:
2 MOST IMPORTANT ASPECTS 1. THEY ALL HAVE CUSTOMERS AND SUPPLIERS. AND THEY ALL NEED FEEDBACK LOOPS FOR IMPROVMENT.
OPERATIONAL DEFINITIONS GIVE INCREASED COMMUNICATION AND WITHOUT THEM SERIOUS PROBLEMS CAN OCCUR.
Quality Control
test prep # 1
The deming cycle:
aid mgmt in pursuit of never ending process of improvement.
4 basic stages.
1. Plan
What, How, Who,
Here is where problems are identified and analysis done with one or more SPC tools. This and brainstorming together. See following writing.
Subject matter, purpose,location, sequence, people, method, cost.
What is being done? Can it be eliminated.
Why is it necessary? Clarify its purpose.
Where is it being done ? Does it have to be done there.
When ? Best time? Then ?
Who. Someone else ? Why them.
How best method, other way.
How much. Cost now? Cost later.
2. Do
Educate and train all.
Hold meetings not suggestions.
Make changes.
Collect/analyze data.
3. Check.
Continually monitor.
Expectations.
Monitor effects, look for backsliding.
Interpret results, conclusions inferred?
What do conclusions mean.
Did the "do" have an improvement.
Do results from statistical analysis in this stage, influence the act stage.
4. Act
Should we continue.
Collect more data.
Get more suggestions.
If success, standardize it.
Adopt it or abandon it.
How to change process now.
Run thru cycle now, possibly under different environmental conditions.
Continuous quality improvement.
(CQI) and PDCA cycle.
Step 1.
Select a theme. Background info, flow diagram, objectives.
2. Full grasp of situation. Control charts,check sheets, pareto charts, fishbone.
3. Analysis of present. Fishbone and Pareto.
4. Plan into motion. Trial basis.
5. Collect, study, analyze data concerning effectiveness of plan. Run and control charts.
6. Establish revised standard operating procedures.
7. Establish plans for future actions.
ISO 9000
An Unification plan.
Adopted in 1987. For standardization.
This verifies a set of guidelines for companies to demonstrate they have a working quality system.
89 countries have adopted ISO 9000 standards as their quality standards.
Over 25000 European countries are ISO 9000 certified.
Only 400 USA firms are.
Standards.
ISO 9000 is a standard which provides guidelines for selection and use of other for standards. (ISO 9001-9004).
ISO 9002 covers manufactures and installation only.
ISO 9003 covers final product inspection and testing only.
ISO 9004 provides direction for internal use by a manufacturer to develop a quality mgmt. System.
Benefits.
Competitive advantage.
International acceptance and recognition.
Improve supplier control.
Reduced customer audit.
Improved internal control.
Better quality.
Better trained employees.
Higher productivity.
Less waste, scrap, and inspection costs.
Higher return on investment.
Short coming.
These are minimum requirements for an effective quality system.
Only forces company to specify their assurance standards and procedures and they follow them.
Quality must be one part of quality improvement efforts.
ISO 9000 alone will not make European countries attain world class quality. JURAN.
ANALYZING PROBLEMS AND ELIMINATING NONCONFORMANCE.
1. Define situation. Clearly, plan solution.
2. Fix (temp) rework or repair, needless expense , not permanent solution.
3. Identify root causes. Cause/effect diagram, pareto chart. Fishbone chart.
4. Evaluate and follow up.
Pareto chart Is a bar chart in decreasing order of occurrence of problem causes.
Fishbone construction.
Idea generation. Amount of them.
Enforce rules of brainstorming. order, no criticize, ok to pass, build on ideas of others, list on board as going (to give ideas to others).
Continue until no more.
Clarification phase.
Team reviews list so all understand. Eliminate duplication of ideas.
Do not discuss at this time.
Evaluation phase.
Eliminate Irrelevancies or off limit issues.
Charting phase.
Team determine major cause grouping. Form, method, machine, material.
List causes on main branches. Interrelated on twigs and little twigs.
General approach to finding root causes. (Problem solving)
1. Understand process. (Flowchart)
2. List all possible causes fishbone.
3. Collect data.
4. Data analysis if necessary.
5. Identify relevant and major causes. Pareto chart.
6. Go back to step one.
Always evaluate and follow up.
TOTAL QUALITY CONTROL
Taguchi Methods.
1. Loss function. Quality losses are measured by system-wide problems, not just rework. Deviation from target, not conformance to specifications.
Target value=customer’s viewpoint of performance characteristics.
2. Quality of design. SPC will not solve poor design.
Robust design= combines engineering and Statistical methods to design products which are insensitive to different sources of variations.
Crosby
Quality mgmt is a systematic way of guaranteeing that organized activities happen in the way that is planned. Must start at the top, must be ingrained in company.
Quality is conformance to requirements, it is measurable (quality cost).
Workers perform like the attitude of managers.
Quality improvement though defect prevention ( 14 step program).
Act now for reward later. Zero defect is the attitude. Do job right first time.
85 % of problems can be solved at first level of supervision.
Crosby 14 step to quality improvement.
See computer.
JURAN
Japanese revolution is based on.
Massive quality training program.
Annual programs for quality improvement.
Upper mgmt leadership of quality function.
List see computer.
Costs of quality.
Prevention costs. Voluntary
Appraisal costs. Voluntary
internal failure. Not voluntary.
External failure cost. Not voluntary. Worst of all costs.
DATA ANALYSIS
TYPES OF DATA
Variable (continuous). Measurements which describe a quality characteristic.
Time, temperature, cost, weight.
Cost per patient meal, admission time, work order turn around.
Attribute (discrete)
Counting occurrences.
Number of defects, of good of bad, pass ,fail.
% of defective, # of billing errors, # of dented cans.
PARETO chart
Graph by reason of overall problem. Don’t just graph numbers, but by percentage of overall failure rate.
SAMPLING DISTRIBUTION
TYPES of control charts
see computer.
Total quality control section # 2 page 5,6,7.
Crosby 14 steps to quality improvement.
1. Clear that mgmt is committed to quality.
2. Team representatives from each dept.
3. Determine current and potential quality problems lie.
4. Evaluate "cost of quality" and explain use as a mgmt tool.
5. Raise quality awareness and personal concern of all employees.
6. Take action to correct problems found in previous steps.
7. Make committee for "zero defects program".
8. Train supervisors to actively carry out their part in the quality improvement program.
9. Hold "zero defects day" to let all know there has been a change.
10. Help and encourage employees to make self improvement goals for them and teams.
11. Help employees to tell mgmt their problems in getting the goal of quality.
12. Show respect and caring to all that participate.
13. Make "quality councils" to meet and discuss on a regular basis.
14. Do it all over again, it will never end. Quality, a never ending process.
J.M. JURAN
Fundamental changes for the west.
A structured annual improvement in quality. 98% of japan mgmt trained in quality.
JURAN 10 step to quality improvement.
1. Build awareness of need and opportunity for improvement.
2. Set goals for improvement.
3. Organize to reach goals, quality counsel,find problems,select projects,appoint teams,chose facilitators.
4. Provide training.
5. Carry out projects to solve problems.
6. Report progress or failure.
7. Give recognition.
8. Communicate results.
9. Keep score.
10. Maintain momentum by making annual improvement part of regular operations and processes.
External Failure Costs.
Labor on returned products, allowances, re-process engineering, travel costs to fix & inquire.
Warranty claims, liability claims, recalls, complaints, declining and lost sales, lower prices, lower future profits, lost reputation.
Internal Failure Costs.
Cost of scrap, cost of spoiled units, rework labor, materials, overhead, rescheduling of production, retesting, reinspection, excess handling, excess mfg cost due to poor design, reprocess of engineering and analysis of failure, design changes, lost contribution margin on unsaleable goods, disposal costs of defective product, maintenance of safety stocks.
APPRAISAL COSTS.
Testing and inspection of received goods, of in process goods, of finished goods, supplies used for testing and inspection, maintenance for testing-labor-supervision, units destroyed for testing, maintenance and depreciation of testing equipment, field testing and appraisal at customer’s site, supplier verification, and customer satisfaction survey.
PREVENTION COSTS.
Design engineering (materials choices and # of parts.),mfg process engineering, quality training, quality improvement projects, quality improvement motivation (empowerment, recognition) , continuous improvement methods, supplier evaluations (technical support), purchase of higher grade material, quality data gathering (analysis and reporting), audits of systems, SPC activities, equipment maintain.
Sampling distribution.
Types of control charts.
Control charts.
For variable data. For attribute data.
For center. For Variation. For defective. For defects.
X-bar chart Range-R-chart. Proportion-P-charts Number-C-charts.
Individual-x- moving range-MR-charts. Number-nP-charts. # per unit -U- chart
median charts.
Why control charts work.
Performance always vary, natural variation usually flows normal distribution, if not sample taking can create one (central limit theorem). Process or control limits are inherent natural variation of process due to chance causes. Upper and lower control limits can be determined using normal distribution characteristics as follows.
UCL=center + 3 standard deviation. / LCL= center-3 std.deviation.
This will guarantee 99.73 % accuracy.
This is to see if process variation is natural = the process is stable, stable process = state of statistical control. This means variations are predictable within limits. Out of limits causes will be assignable, evident ( backwards).
Understanding Variation.
Most be done so managers can act rationally, to variations in figures, performance, and the data with it.
Two types of variation, natural, and unnatural.
Natural variation is caused by randomness or chance. Many variables which cause minor shifts in variations day after day. Random causes are unexplainable and part of the system.
Unnatural variation is assignable or special causes. These are few and identifiable and cause variation which is above the "natural system". These come from people, machines, and materials. Process control finds these and corrects them. Control charts are effective here.
Basic characteristics of data.
Central Tendency Variability Shape
Mean. Range. Histogram.
Median (simplied of x&r) Std.deviation. Line,bar,scatter,pie,control,
mode (value 2=bimode) Comparison, pareto chart.
FLOWCHARTS
Graphical method of displaying a system’s operation and sequence.
Order of operations:
What is output of this step.
What materials are required.
What operations are required.
What is the consequence of no inspection here.
What quality control here is needed.
Do not show more detail than needed.
Methods:
1. Operations with major bearing in separate boxes.
2. Operations with little impact (or none) will be combined.
3. Sequential operations similar in nature into single box. 3 sequential welding jobs.
4. Use several simple charts instead or single complicated one. To show subassemblies.
4b. Subassemblies are best shown separately.
5. Show only detail necessary. To personnel.
6. Trace parts back to origin in plant, how was quality assured there.
7. Consider "rework on the fly" if defective are unavoidable.
8. Show external operations (customer source inspection or deliverable data as operations.
9. Continue the flow right though product delivery, including packaging and shipping.
STATISTICS.
Measures of central tendency.
Mean=typical value. Denoted bu "u" or "mew". The average
Sample mean. .x.w/bar called x bar.
Mean of sample subgroup. Average of averages= x bar bar (double bar on top).
Median= average of the cumulative frequency of fifty percent. High/freak values will skew the averages. Enumerative studies throw extreme points out (oddballs) , analytic studies do not. This means this must be used with caution.
Mean shows extreme values, median value does not.
The mode of distribution is the value which occurs most frequently. Not affected by extremes/freak points. One pt=unimodal. More=multi modal
Measures of variability.
All processes and populations have it. RANGE= largest data pt.--smallest data pt.
Standard deviation figures
data pts and distance from mean. Sigma O, sample called s.
|
BACK TO 100 examples in Business, Operations and
Engineering. |
|
Do it once, do it right, and do it now.
|
|
Back to Lawson Computing Homepage
|
Apply as needed,
when needed. |
