Archive for April, 2016

01
Apr
16

Coming to a galaxy not far, far away

0416MEM_Cover_no_boxLet’s face it, many of us would rather have root canal surgery than shop for a new car. Navigating the art of the deal on the showroom floor leaves even the best of us (maybe not you, Mr. Trump) ready for a cup of chamomile tea, if not an adult beverage.

Dealerships understand this and are undertaking efforts to make the process seem less contrived and the customer experience more pleasant. Sales people are being trained to make more eye-to-eye contact so they seem genuine. They smile more and pat little children on the head. Unfortunately, even the most savvy car makers and their dealers still give car buyers the impression they are being taken for a ride.

Wouldn’t it be a better experience to negotiate the purchase of that shiny red coupe you’ve always had your eye on with an automaton instead of a sales person?

Algorithms running on interconnected computers could reshape auto sales the way they have other industries. Look at where the brick-and-mortar travel agencies and video stores of yesterday are today: almost exclusively online.

In his revealing cover story this month, associate editor Alan S. Brown reminds us that automatic systems are depopulating professional offices as well as retailers. In some cases, Brown says, software has replaced loan officers, attorneys, and even writers and journalists; and engineers are relying on expert systems to evaluate designs and simulations. Even the investment community is adopting automated transactions. Robo financial advisers, which offer automated investment services and advice, often outperform human advisers who may be occasionally unscrupulous, on top of being unable to beat the market. (Nothing personal guys.)

What we’re talking about here is deep learning, and networks that think like brains. These are artificial intelligence (AI) systems that go beyond following hard and fast algorithmic rules like some robot on a factory floor. IBM, for example, is advancing its Watson to diagnose diseases and to read medical images. Companies such as Google, Facebook, and Microsoft have collectively spent billions to fund the development of neural networks that can understand human speech and recognize faces in photos. In the next decade, AI could well power thousands of machines and gadgets through cloud services.

Of course, robots come in all shapes and sizes, so the time is nearing—probably before the final chapter in the seemingly endless Star Wars saga is ever written—when droids like C-3PO become run-of-the-mill companions in this galaxy, not in the one that is far, far away. DARPA, the Defense Advanced Research Projects Agency, is investing millions of dollars on projects to develop a kind of smart robot that is stronger and braver than C-3PO. In the DARPA Robotics Challenge, an Olympiad for robots, teams from Boston Dynamics, NASA Jet Propulsion Laboratory, and other organizations competed to develop autonomous robots with dexterity matching our own human deftness, but that can survive and work in extremely hazardous conditions, perhaps even the automobile showroom floor.

Come to think of it, maybe Damari, the sales guy I met last week at the Volvo dealer, wasn’t so bad after all.

Advertisements



The Editor

John G. Falcioni is Editor-in-Chief of Mechanical Engineering magazine, the flagship publication of the American Society of Mechanical Engineers.

April 2016
M T W T F S S
« Mar   May »
 123
45678910
11121314151617
18192021222324
252627282930  

Twitter from John Falcioni

Twitter from Engineering for Change