Different Paths Toward Democratization
June 1, 2018
The lore of my family includes the story of my father taking apart a car’s engine with a hammer, a screwdriver and a pair of pliers to replace a gasket with one he had cut from a cardboard box.
The car was an early Ford, really early. My dad was born in 1922, so the beater of his youth had a hand crank. Automotive technology was different then. Still, he kept up with it, working as a mechanic for awhile and then maintaining the family cars. Oil, filters, tires, brakes, radiators, batteries, exhaust systems and more were repaired or replaced in our garage, no mechanic required.
From Cars to ComputersNow, I look under the hood when buying a car only as a force of habit. If anything more serious than an oil and filter change or battery replacement comes up, I know I’d be dropping it off at the mechanic. The complexity of automotive technology today, combined with a fairly low incidence of catastrophic failure and convenient access to specialists has won out over the auto repair democratization ideals my father prescribed to. As my dad would eventually say when he ran across an auto repair he couldn’t do: “Too many computers in cars today.”
I saw that as a good thing then. The car of his youth was the personal computer of mine: a revolutionary technology that was changing the world. While playing the role of mechanic’s assistant as a tween, I would have rather been tinkering with my computer. I could upgrade RAM, hard drives, and power supplies, overclock processors, install graphics cards and replace motherboards. I knew every jumper setting.
The last computer upgrade I undertook was a simple solid-state drive installation in my laptop. My daughter assisted, glancing up from her phone occasionally as I pointed out what the various components were and what they did. She would have rather been Snapchatting.
When it comes to cars and personal computers, their widespread use didn’t come in the form of “simpler” technology, but via shifts in reliability, affordability and ease of use—backed up by convenient access to specialists—which continue to take place over generations. I don’t need to know how to change a tie rod to steer my car. My daughter doesn’t need to know the difference between integrated and discrete graphics to watch a video on a computer. If something catastrophic goes wrong, we have options when it comes to hiring specialists to make it right.
From Computers to SoftwareYou can take your pick of revolutionary technologies changing the world today. In the field of design engineering, technologies like simulation, cloud computing, artificial intelligence, augmented reality, additive manufacturing and sensor-laden, connected products are just a few with society-altering promise.
I would argue that most of those will eventually follow the usual democratization rules. Proving themselves useful, they “just” need to be easy to interact with their underlying complexity, decrease in cost and increase in reliability. And with the current rate of technological innovation, I don’t think democratization will take generations to happen.
But simulation is different. The fact that simulation software (and the AI that it will increasingly employ) is used by engineering teams to help create and validate complex technological products puts it on a different democratization bell curve. We want someone on the engineering team to be able to “look under the hood” and understand what the software is doing and why. For many teams, that mechanic equivalent—the simulation analyst—is not available in-house or is only available from the software vendor’s network.
You could also argue that simulation is actually on multiple paths because different products in different industries with different physics require their own cycles of simplicity, reliability and affordability to become democratized. And can a specialized use case for one type of product in one industry ever truly be considered to be “democratized”?
When we talk about the democratization of simulation, we can’t expect it to meet the same definition of the ubiquitous technologies in PCs or cars. We can expect easy-to-use, modern interfaces that hide the complexity when it’s not needed and help explain what’s happening when it is, affordable options for different engineering team sizes and use frequencies, and increasing reliability when applied to more and more types of product design and development scenarios. As you’ll see in this issue and our upcoming coverage of the Conference on Advancing Analysis & Simulation in Engineering (CAASE, June 5-7), software vendors, academia and end users have made great strides in the democratization of simulation. What more needs to be done? Tell us here via a very brief survey.