Pictures of the Future    Spring 2006

Simulation – Scenario 2020

Hidden Identity

May 2020. A tour of a future virtual automobile production center reveals the extraordinary scope of simulation and optimization—as well as a case of hidden identity.

I'm what you call a car guy. Show me anything with four wheels and a motor, and I'll start dreaming about how fast it can lap the north circuit of the Nürburgring, the Indy super speedway or the tri-oval Daytona—some of the toughest tracks in cyberspace. Sure, it's all virtual now—everything from materials and parts design to engine, tire and vehicle testing, not to mention the races themselves. But there are still plenty of real customers out there and they still need real cars to get around in.

Take this baby here. She's going to be the world's first hybrid-electric multi-fuel-powered car equipped with a micro HCCI—that's a homogeneous charge compression ignition engine—a very small yet impressive power plant that combines near-zero emissions with outstanding fuel efficiency. Once development is complete—and by the way, there's nothing like simulation and optimization for accelerating that process—and the assembly plants are humming, it'll be bye-bye Oil Age.

This car will be able to run on anything from switch grass juice to Grand Marnier, not to mention hydrogen. What's more, according to my own calculations, it will be 27.32 % more fuel efficient than any other internal combustion engine tested to date.

Sorry, I tend to get a bit carried away when it comes to numbers. But numbers are what make this business buzz. Just look at this place, it's nothing but numbers. I mean it's all simulated—the engines, the cars, the factory, the sounds, the interactions… It's even amazing for Prof. Bunsen, our head of worldwide R&D.

In fact, Bunsen was here just a few minutes ago to see how things were going. We toured the entire plant—a virtual fly-through you might say. We slowed down the mobile production lines from their high-speed real time values long enough to witness how a simple change in a new engine component would alter the movements and coordination of the robots, the performance of software, and even the associated logistics. It was a great example of what we call "parallel engineering."

Every now and then one or two engineers from remote sites literally popped in to run software on a virtual prototype, simulate a remote maintenance routine, or conduct a virtual meeting. Surprisingly for me, Bunsen did not seem to notice any of them. But when I pointed this out, he just gave me one of his knowing looks and said "Mr. Sonn, it's all in your head." What did he mean by that?

When we got to the control center and saw existing software being tested on banks of virtual monitors, Bunsen looked impressed. But then he asked, "Are we really going to need on-site presence?"

"At the beginning, yes," I said. "But the system will get better and better each day. Eventually, remote presence will be more than adequate. You see, sir, the plant will be equipped with hundreds of thousands of sensors that will respond not only to local processes, but to what other sensors are reporting.

At first, they'll use networked gaming to compete against each other to see which parts of the plant can achieve the highest efficiencies. But eventually—since they are a learning system—they will achieve an optimized equilibrium. If errors are detected, they'll be instantly corrected and the resulting information will be used to improve processes. Yes, sir, I expect the resulting quality levels to be exceptionally high. Our customers will surely love it!"

We moved on, and after a while Bunsen stopped in front of one of the new HCCI cars, its dirt- and scratch-resistant chameleon coating glowing like a desert sunset in a perfect simulation of its future appearance. "And this is the car that's going to change the world," Bunsen said, giving me one of those knowing smiles again from over the rims of his narrow glasses. I was not sure what he meant. "Sir," I said with pride, "would you like to meet the engine?"

As I pronounced the last three words, the hood grew as transparent as cellophane and the engine, which we could now see in cross-section, came to life with the soothing purr of a stroked cat.

I slowed down the simulation with a voice command so that we could watch the individual cylinders moving in perfectly choreographed rhythm, barely visible mists of fuel spraying into the space left by each receding head, then igniting at exactly the right moment and with exactly the right heat distribution pattern as each piston neared its zenith.

We were able to zoom in on these patterns and "walk through" the mist of fuel droplets, experiencing how the patterns changed in response to simulated alterations in the car's speed, altitude and load.

By reaching into the simulation I was able to separate and enlarge individual sections of it, and demonstrate the interactions of flow dynamics and acoustics as algorithms neatly optimized the outlines of each color-coded parameter.

"Thrilling," said Bunsen, finally stepping back from the vehicle as if the hyper reality of the demonstrations had left him breathless. Then, abruptly changing gears, he looked at me and asked. "What's it like being, well, virtual?"

"I don't understand, sir. What do you mean?" I said.

"Mr. Sonn, don't you know? Hasn't anyone told you? Why your very name… it means Self-Optimizing Neural Network."
                                                                                                            Arthur F. Pease