SMT007 Magazine


Issue link:

Contents of this Issue


Page 23 of 93

24 SMT Magazine • February 2017 by Tom Borkes THE JEFFERSON PROJECT Where is our industry going? The beginning of a new year provides a good opportunity to look into the crystal ball and foretell what we think will be. Using general cause and effect ob- servations of the past, can we predict products and processes that will be developed in the near future? In science, we call it extrapolation: estab- lishing a relationship between variables that form data points (statistical inference) and us- ing the relationship or function to predict an occurrence outside the known observation range. In a sense, this is what Professor Marvel did in the Wizard of Oz when Dorothy asked him to tell her the fu- ture. He sees a young girl run- ning away from home with a picture of an older woman (Auntie Em) in her basket and combines this with the dan- ger of the forthcoming tornado and says, "I see a woman who is very worried about someone she loves very much." Professor Marvel infers a conclusion from a series of observations. We can create a predictive model with causes and effects based on independent and depen- dent variables. However, the model may have to be very complex and statistically based. Re- member the butterfly effect, where a tiny action like a butterfly's fluttering wings in Singapore is attributed to causing a hurricane in Florida. Starting with a set of initial conditions we apply the most likely changes to the indepen- dent variables (What we allegedly have control over, like climate change) and calculate the effect on the de- pendent variables—like the seas rising. Does the mod- el converge or diverge? The dirty little secret is that when we know the conclusions we want, we can rig the mod- el to predict the desirable re- sults. How? We keep chang- ing the model's relationships The Shape of Things to Come FEATURE The Wizard of Oz, 1939.

Articles in this issue

Archives of this issue

view archives of SMT007 Magazine - SMT-Feb2017