Quality management 

Chapter 1: Introduction

  • Variability:

    • Mostly proportional to variability in processes.

    • Variability is viewed negatively; the goal is to reduce it for predictability.

  • Specification Limits (LSL & USL):

    • LSL: Lower Specification Limit

    • USL: Upper Specification Limit

    • Specification limits defined by customer needs, R&D, or design teams.

    • Customers expect products within certain tolerances (e.g., shoe size 10 may vary to 10.0001 or 9.9999).

    • Difference with Control Limits:

    • Control limits are calculated by the manufacturer, not predetermined by customer specifications.

  • Quality Engineering Overview:

    • Quality is defined as reduced variability in products or processes.

    • Subjective definitions of quality exist, but the focus is on minimizing variability.

  • Critical Characteristics:

    • Metrics that can be physical, sensory, or time-oriented, critical for quality engineering.

    • Objective: Work with these metrics to reduce variability.

  • Engineering Types:

    • Over the Wall Engineering:

    • Traditional method where design is completed and handed off to manufacturing without collaboration.

    • May lead to issues during production if designs are not manufacturable.

    • Concurrent Engineering:

    • Design teams and manufacturing work together from the start.

    • Example from speaker’s past experience in manufacturing power cables illustrates benefits of collaboration by identifying issues in production early.

  • Data Types in Quality Engineering:

    • Variable Data:

    • Continuous, can have infinitesimally small increments (e.g., time measurements).

    • Attribute Data:

    • Discrete, consists of distinct categories (e.g., defects counted as whole numbers).

Chapter 2: Defects On Refrigerator

  • Nonconformity vs. Nonconforming Unit:

    • Nonconformity: A defect identified in a unit.

    • Example: If a refrigerator has 2 defects, those are nonconformities.

    • Nonconforming Unit: A unit that has one or more defects. If a refrigerator has at least one defect, it is considered nonconforming.

  • Statistical Distributions:

    • Binomial Distribution: Used when there are two outcomes (e.g., good vs. bad).

    • Example: Nonconforming vs. conforming refrigerators.

    • Poisson Distribution:

    • Counts of defects, which are discrete and not continuous like the normal distribution.

    • Characteristics:

      • Shape similar to normal distribution but based on discrete counts.

Chapter 3: Number Of Defects

  • Yield Calculations Introduced:

    • Throughput Yield: Calculation based on units coming out of a process.

    • More Yield and Defects Per Million Opportunities (DPMO): Additional yield calculations relevant to quality evaluation.

  • Throughput Yield Calculation Example:

    • Examines 4 refrigerators with varying defects:

    • 1st: 2 defects

    • 2nd: 1 defect

    • 3rd: 1 defect

    • 4th: 0 defects

    • Calculation of conforming vs. nonconforming units provides insight into yield rate.

  • DPU (Defects per Unit):

    • Formula for DPU is Total Defects divided by Total Units Inspected.

    • Example leads to average defects per refrigerator being calculated.

  • Throughput Yield Formula:

    • Uses exponential function related to the Poisson distribution:
      Y=e−NPUY = e^{-NPU}

    • NPU is the defects per unit calculated previously.

Chapter 4: Further Defects Assessment

  • Probabilities Using Poisson Distribution:

    • Understanding how to utilize the formula for defects to find probabilities is key.

    • Example provided illustrates how to find probability for no defects, and connects back to definitions.

Chapter 5: Different Process Steps

  • Rolled Yield:

    • Reflects the cumulative yield across multiple process steps.

    • Example:

    • Starting with 100 units, yields step-by-step through multiple steps (97%, 95%, etc.), resulting in final output yields.

  • High-complexity Processes:

    • Case of semiconductor manufacturing requires high yields and understanding of distributed processes.

Chapter 6: DPMO Calculation

  • DPMO:

    • Covering the calculation basis for defects per million opportunities.

    • Example with airline baggage handling:

    • 3 bags lost out of 8,000 passengers, each carrying 1.6 bags leads to effective DPMO calculation, identifying process effectiveness across complex operational environments.

Chapter 7: Statistical Process Control

  • Control Charts:

    • Used to monitor production processes and identify when they go out of control.

    • Distinction between control limits and specification limits emphasized; critical for operational efficiency.

  • Statistical Methods Overview:

    • Upcoming topics include design of experiments and quality assessments, stressing the importance of statistical foundations.

Chapter 8: Conclusion

  • Quality Cost Types:

    • Further classifications of quality costs into four quadrants: Prevention, Appraisal, Internal Failures, and External Failures.

    • Most serious losses arise from external costs, while investments in prevention are most valuable as a leverage effect in mitigating overall expenses.

  • Importance of Quality for Business Success:

    • By improving quality and reducing variability, organizations can enhance productivity and reduce costs over time.

    • The course will focus on methods to quantify and improve quality through various statistical tools and techniques.