Learn and DEPLOY FastR&D

Innovation and R&D

Statistical Process Control for R&D

In my last article, I mentioned If you have a limited R&D budget or resources “Design of Experiments(DOE)” is one tool that can give you the biggest bang for the buck.

A similar powerful tool that can give the biggest bang for the buck for Manufacturing is Statistical Process Control. SPC for short. And since SPC is vital for Manufacturing, it needs to be equally important for R&D, and actually if you think hard, even more important for R&D!

Quite often we hear “Practice makes Perfect.” And occasionally we also hear “Perfect Practice makes Perfect.” So here is the question? How do we in R&D, help Manufacturing identify Perfect Practice? Trial and error at manufacturing scale? If not, is there a better way?

What is Perfect Practice for our new product or new process? How do we produce a quality product consistently day after day? What is quality?

What is Quality?

Quality is conformance to customer requirements. - Phillip Crosby [ref. 1]

Quality is fitness for use. - Joseph Juran

[Cost of] Quality is the loss to society. – Genichi Taguchi

Taguchi provides a very broad measure of quality through Cost of Quality as loss to society. If you see waste, you see loss to society! More importantly, if you see waste, you also see an opportunity for quality improvement!

And the concept of “loss” can be applied not only to chemical reactions, separation processes, product quality but also packaging, billing, cleaning, safety, timeliness and so on.

The real pioneer of Quality movement worldwide was Dr. Edward W. Deming. Dr. Deming was a US based Survey Statistician, assigned to Japan after World War II to help Japan run a national Census survey. During his stay in Japan, Deming offered to give some lectures on Statistical Process Control. It has been said Deming told the Japanese “use Statistical Process Control for product quality and you will conquer the world.” Over the years Japanese companies like Toyota, Honda, Cannon, Sony became household names because of their legendary QUALITY.

The good news is today, Indian companies are catching on. Over a dozen Indian companies have developed and demonstrated world class excellence in quality by winning the coveted Deming Prize. [ref. 2-3]

Deming does not provide a formal definition of quality but recommends you achieve quality/consistency by operating the process using Statistical Process Control and ‘Control to Target’ philosophy.

Deming also says that once a process is under statistical process control, if we tinker with the process, we only increase process variability/inconsistency. [ref. 4]

What is SPC?

Here are a few working definitions.

Statistics Science of Decision Making in presence of variation in data

Process A series of actions or operations aimed to produce a desired result

Control Skill in guiding or directing

How do we in R&D package and transfer our R&D know-how and know-why to Manufacturing?

SPC is one simple, efficient and effective vehicle to make that happen.

SPC can be started with seven simple tools, it can be learned and practiced easily by workers at all levels and can be and ought to be part of the common vocabulary/language for process tracking, control and improvement.

Kaoru Ishikawa, learned the basics of Statistical Process Control under Dr. Deming and started teaching and promoting Statistical Process Control. He was among the first to identify and use the seven basic SPC Tools. [ref. 5]

For SPC to be widely used

Management must be involved, commitment is not enough
Workers must be involved
SPC methods must be

simple
visible
fast
consistent

Basic SPC Tools

1. Flow Chart
2. Cause and Effect Diagram or Fishbone Diagram
3. Pareto Chart
4. Tally Sheet
5. Scatter Plot
6. Histogram
7. Run Chart or Control Chart ( run chart with control limits/rules)

All seven tools are simple, visible, fast and consistent.

By communicating though these simple tools, the Problem Solving evolves from opinion-based solutions to data based solutions. Discussions become focused.

What data do you have?” “Here is my Control Chart, Pareto chart of Defects and Cause and Effect Diagram related to Defects.

So, how do we in R&D package and transfer our R&D know-how and know-why to Manufacturing?

Use Statistical Process Control methodology:

Provide Process Flow chart
Provide Targets and Tolerances for each recipe variable
Provide Targets and Tolerances for each process variable
Provide Targets and Tolerances for each quality attribute
Provide Cause and Effect diagram for each quality attribute (and productivity)
Identify key variables for each quality attribute through a Pareto Chart
Provide Control Charts for various quality attributes

If R&D provides such data and documentation for each new product/process, Manufacturing will learn Perfect Practice faster. R&D better understands its key customer, Manufacturing’s requirements and R&D develops, uses and promotes better science. It is a win-win for everyone.

So crank up your Product Development engines… Let us speedup new product development and growth rates. And let the fun begin!

References:

1. Philip Crosby, Quality is Free
2. https://en.wikipedia.org/wiki/Deming_Prize
3. http://isqnet.org/demingprzwinners.html
4. Kaoru Ishikawa, Guide to Quality Control
5. Edwards Deming, Out of Crisis