If you asked people what skills are needed for an engineering career, most would probably respond with “math and science.” That’s largely true, but which math and science skills? Are calculus and physics enough to carry you through your whole career?
After a couple decades as an engineer, I’ve realized that there are some critical skills that engineers need that aren’t so obvious, and often not covered in our basic schooling. Let’s take a look at some of these and why you need them.
In college my buddy Joe was a management major who wanted to work in baseball. He had to take a lot of math classes, especially statistics, and it’s served him well. In the Moneyball era, baseball has become dominated by Sabermetrics – the application of statistical analysis and methods to baseball records – and it has revolutionized scouting and player selection.
As a mechanical engineering major, I wasn’t required to take a statistics class, but we covered statistical process control (SPC) as part of manufacturing quality.
However, I’ve found that engineering decisions are often business decisions. People at Amazon are fond of saying “Amazon is a data-driven company,” and it’s true. If you propose a solution for a problem, you will often get challenged to show the data that supports the choice and the trade study that compares your solution to alternatives. I’ve used statistics more than I expected as part of data analysis, trade studies, and generation of key performance indicators (KPIs) to measure the health of processes.
Joe, by the way, is now the Assistant General Manager of a major league baseball team and the reason I haven’t had to buy a baseball ticket in over 20 years.
As a mechanical engineer I never expected to write algorithms and programs; that’s always felt more like computer science. Surprisingly, I end up writing quite a few programs to parse the data I’ve collected to identify trends and form conclusions.
Often when you’ve collected data, it’s not enough to calculate the mean, standard deviation, confidence interval, variance, and so on. In addition to looping through large sums of data, you want to analyze the relationship between multiple factors. This involves evaluating conditionals while processing the data or parsing a large dataset into additional vectors and matrices that you will evaluate separately.
The term algorithm can sound daunting initially, but it simply means defining a logical process for solving a problem. Product development is at its core an algorithm.
I started out my career as a structural analyst where I performed a lot of finite element analysis (FEA). One of the most common outputs from FEA are color contour plots that look like heat maps, going from blue to red, with red indicating peak values.
One of my biggest mistakes was presenting results where I didn’t normalize the values – in other words, scale them to the material yield strength or design limits. I would show the actual values for displacement, stress, temperature, and so on. The audience would see red and immediately think there was a failure.
Is red always a failure?
That made me realize that it doesn’t matter how good your results are if you can’t convince your customer of your intended conclusion. Sometimes good “viewgraph engineering” is the difference between making the right or wrong engineering or business decision.
Here’s how you can use your tools to become better at communication and presentation:
Much of mechanical engineering education and training is based around design, simulation, and manufacturing, but the business side of product development involves skills based around decision making and management. If you want to develop your proficiency in these areas, the capabilities exist in your CAD tools and engineering math software like PTC Mathcad – explore them!
