Centre for Learning

Design of Experiments for Product and Process Control

This course shows how to make practical use of Design of Experiments by combining inputs, like temperatures, times and dimensions, to give maximum information from the least number of trials.

DOE identifies important factors, shows how much affect they have, and allows you to predict process changes and to define optimal design settings.

Explanations are practical and visual. Graphs clearly show major inputs and their effect, and spreadsheets generate charts and calculations.

Features

  • Simple, practical approach
  • Rapid, visual, interpretation of results
  • Taguchi in a usable form

You will learn how to…

  • Identify key design parameters
  • Discover important process inputs
  • Untangle interactions
  • Pin down elusive causes.
  • Multi factor process solution

This course gives you tools to:

  • Specify critical outcome parameters
  • Search for possible input factors
  • Select the key drivers
  • Estimate the effects of multiple input factors – individually and in combination, and
  • Optimise the system.

PROGRAM CONTENT

Specify an outcome to measure

  • Find parameters which matter to the customer, and ‘feel the pulse’ of the process.
  • Deal with intermittent faults, and subjective assessments.

Search for possible inputs

  • Practical search techniques.
  • Two kinds of input factor: control variables and noise variables.

Select key factors

Use a screening experiment to test up to seven factors in only eight trials.

Learn how to:

  • set the input levels
  • check for interference and impossible combinations
  • plan for maximum information gain
  • use visual methods to select key factors.

Estimate Effects

  • Fine tune the input factors, discover how much effect they have and how they work in combination.
  • Anticipate interactions
  • Estimate effects with a second small experiment.

Visual interpretation of results

  • Line plot: discovers outliers and hidden trends
  • Effect plot: decides reproducibility
  • Interaction plot: shows how factors work in combination
  • Regression plot: relationships between input and outcome.

Optimise the system

Use the results to find settings which:

  • move the outcome closer to target, and
    reduce variation.

WORKSHOPS

The presentation is combined with a sequence of ‘hands on’ workshops which form an integral part of the course and ensure genuine learning.

Workshop #1: Process Improvement

  • Improve the outcome of a model process.
  • Apply the five steps: Specify outcomes, Search, Select, Estimate, and Optimise, to achieve a result which is closer to target and has less variation.

Workshop #2: Spread sheet familiarity

  • Use Excel spread sheets, provided as part of course material, to confirm your interpretation of workshop #1.

Workshop #3: Design Optimisation

  • Optimise the design parameters for a model product:
  • select the key factors to investigate in detail
  • discover relationships and interactions
  • find optimum parameter values
  • use spreadsheets to speed the task.

Workshop #4: Plan your own multi factor solution

  • Specify your own critical outcomes, begin to search for possible inputs and plan to select key factors.

Presenter
Alan Long is a Chartered Statistician and is an acknowledged expert in the application of statistical methods to a wide range of business and manufacturing environments. He has been applying Design of Experiments and related techniques such as Statistical Process Control to Australian businesses since 1982. Alan enjoys an excellent reputation as an instructor and wrote the STATMAN series of training books.

Fee

$795 excluding GST. Fee includes training Manual, computer spreadsheet files for data analysis, lunch & refreshments. Please refer to our standard terms and conditions for further registration details.

What to Bring?

Information on current trial work, problems or development issues.
If possible, a laptop with a 3 1/2inch floppy drive.

Who Should Attend?

All problem solvers faced with multiple inputs, or a range of influencing parameters.

© Copyright Centre for Learning 2018