CJSteele Probabilistic Services




Theory Update




predicting and mitigating the effects of uncertainty and random variability on technical and commercial systems




Hi, as you might know now, my name is Clint Steele and I assume that you would like to know more about who I am before you look too much more at my website.


I think the best way for me to explain why this website is here and why should listen to what I have to say about quality is to give you some history on myself and how I got here.


Fist off I am an engineer. I decided to become an engineer after I read a hotrod magazine when I was 13.


It was not so much the aggressive styling of the cars, the mountains of chrome sticking out of the bonnets or the thoughts of the acceleration pushing me into the seat, which got me so enthralled. However, they did help. What really got me, was the technical section of the magazine and the way they demonstrated that through having a good understanding of the science of a system one could optimise that system for greater performance.


So I went and studied engineering at Swinburne university, which served me incredibly well, to study engineering and work in the automotive industry.


However, two things happened while I studied.

  • The first was that I worked in the auto industry for six months as part of me course, and learnt that the automotive industry didn’t quite offer the room for comprehensive product development and optimisation of new ideas as I had hoped.

  • The second was that I came across a subject that focused purely on design for quality. This subject was taken by Dr John Browne who was ex Xerox from the days of Fuji Xerox. For those of you who don’t know, there was a significant flow of knowledge on design for quality from Japan to the US though Fuji Xerox. Also, there were the beginnings of an American quality system developing independently at Xerox. This was a time and place where a lot of work in the are of design for quality was being done and a lot of the lessons being learnt were then, lucky for me, were later delivered to me in my undergrad course. If you want to find out more about this time then look for anything written by Don Clausing, who was a major part of what happened then. Anyway, this subject was so intriguing to me; at first I thought basic optimisation was intriguing, but the improvement of quality, reliability and robustness through engineering design optimisation really captured my mind, that I decided that I had to do a PhD on this topic.


So instead of applying for the graduate positions at one of the automotive companies I started my PhD.


During this time I first focused on my thesis topic, understanding the variability in manufacturing and how to account for that during design, and design theory in general.


However as I stayed on I ended up teaching the subject that had convinced me to do the PhD. From this perspective I learnt that most people who learn about design for quality (or probabilistic design) had no trouble seeing the value in it. However, they all found that mathematics gave them severe headaches. I was so enthralled by the topic that I was able to push through those difficulties, and be able to get the full benefit of what was being taught.


Nevertheless, two things came from my observation of the difficulties that the students had.

  • I developed a piece of software that took all of the theory of optimisation for quality, or robustification as I prefer to call it, which I called Robustica

  • I refocused the education so that even if the mathematics was hard to grasp, the students would gain an ability to intuitively spot a more robust and high quality design


As my PhD came closer to an end, things slowed down a bit, and I had more time to do other things while I got ready to start working in industry.


Because I had this free time I decided to do the Master of Entrepreneurship and Innovation (MEI) at the Australian Graduate School of Entrepreneurship (AGSE). I always felt that the commercial side of ‘real’ engineering (the ingenious type) was entrepreneurship, and I wanted to get a handle on that too.


A major part of one of the subjects in the MEI was a research project on the application of new methods to the entrepreneurial process.


Naturally, I wanted to apply my probabilistic skill somehow. Therefore, I took a case that we had done, Clarion Optical, and looked at it from a probabilistic perspective.


I don’t want to say too much about it because the case is still used for educational purposes.


However, I found that it was remarkably easy to start quantifying and ranking the risks that initially were only identified as issues of concern.


Then I found that with some modification to the way the case was modelled, it became possible to optimise the investment strategy and the planned operation of the business to minimise the risk to everyone involved in the business in the case.


Once again, I was faced with the situation where everyone watching could easily see the value in the approach, but just couldn’t get passed the mathematics.


I really felt that I needed to do something to help everyone better understand how they can use probabilistic techniques to minimise risk and increase quality and robustness.


Anyway, the MEI ended and my PhD was passed. I was actually examined by Don Clausing himself; that was pretty good. Not to mention Ken Swift, of Hull in the UK, another man big in the area so I was very happy at that time.


It was now time to work in industry in product development. That was the reason why I became an engineer, to contribute to society by developing new products, and I also wanted more chances to apply what I had learned.


So what happened? Well I got to do the following:

  • Work in a couple of different companies

  • Work in a couple of different countries

  • See the real quality issues that companies faced when developing and manufacturing products

  • Understand the process of managing the commercial risks of introducing new products

  • Confirm what I had learned about the importance of concurrent engineering

  • See my own designs go out into the market

  • Fully appreciate that only some people can or want to really get their heads around quality and that having that gives one a real edge

  • Realise that a lot of people out there need to learn a lot more about probabilistic design if they want to see higher quality


Throughout all of this, the prediction and management of the effects of random variability upon the function of a system (either technical or commercial) has always been my passion.


Because this is my passion, I have been frequently disappointed at the nature of training that is offered in this area. Too often I have seen people with little real understanding speaking about quality and risk. Not only that, I have readily seen people put forward ideas that are just plan wrong. Chances are if you have done a course in quality control, then you've been told something about control charts or capability that is not correct.


And this is all because quality and risk management has become dominated by people who are removed from the actual fundamental process related to the systems that they are meant to be analysing and who have no grasp of basic probability.


That brings me to here and now.


That's why I created this website: to provide you with the probabilistic knowledge and skills that you really need. Not the watered down, simplistic process based methods that you have probably been exposed to.


I set up this website to share knowledge about probabilistic methods so that people can better manage risk and quality.


Please take a look around the site to see if there is anything that interests you.


I strongly recommend you download Robustica and sign up to the article review series.


Also, don’t forget about the free consultation service; I like to hear about the issues you have. If you’re not sure if probabilistic methods can help you, then make contact; we can have a preliminary email discussion first.


Either way, I hope I can help you and I would like to hear from you.




Clint Steele