Everybody talks about innovation these days, but the word is used so lightly. Every new app, gadget or product feature is now “innovation”. A few decades ago, “innovation” implied a life-changing advance in technology: the transistor, the computer, space flight. Does it mean anything that we speak of innovation more casually today than we did in the past century? Maybe.
In October, 2000, the US Congress mandated this goal: “by 2015, one-third of the operational ground combat vehicles are unmanned.” We haven’t reached that goal. Yes, it’s a tough goal, and yes, Google, Daimler, Mobileye and others are making progress on driverless cars. But still, we didn’t make the goal. It seems a pretty modest goal compared to putting a man on the moon, and we did that in less than nine years, with resources that look mighty primitive by today’s standards.
Space travel called for developing a mix of new technology: materials that could withstand tremendous variation in environmental conditions, compact, lightweight food, and a myriad of analytic applications for flightpath computations, meteorology, monitoring astronaut life functions and other uses. That’s a lot of data analysis, especially considering that the most powerful computers in the era of early manned space missions had only around 100 kilobytes of memory.
Analytics plays an equally important role in today’s technology development. Cars already have engines, tires, all the parts except automation. The driverless car is all about gathering and analyzing data to control the vehicle in a safe and efficient manner. Speaking recently at a meeting for University of Chicago Booth Big Data and Analytics Roundtable, Matthew Walter, Assistant Professor at Toyota Technological Institute at Chicago, discussed his work developing an early driverless car, and explained that most of the considerable computing capability required for the car goes to analysis such as computing an appropriate path for the vehicle.
Because analytics is so important to automated driving, Google’s experience with significant advances in analytics applications such as web search and programmatic advertising may be key elements of its success with driverless vehicles. Still, other factors are in play.
Jon Gertner, in his book The Idea Factory: Bell Labs and the Great Age of American Innovation, outlined three requirements for innovation as it was viewed at Bell Laboratories decades ago: market, technology research and manufacturing. So, to be an innovation, a new development had to leverage the latest scientific research, be feasible to manufacture in large quantity, and have clients ready and willing to buy. “If you hadn’t sold anything, you hadn’t innovated.” That tough standard was at the core of the culture that invented the transistor, the laser and digital signal processing.
Today’s tech culture rarely sets such tough standards when speaking of innovation. Yet look at Google and its driverless car. There’s a case where all the elements come together. Google’s needs alone represent a significant market for self-driving vehicles. It builds on research funded by DARPA and begun at universities across the world. And it takes advantage of considerable prior manufacturing development, which has become less costly (and less unwieldy) with increasing volume.
The self-driving car is an analytics product, as will be much of the 21st century’s new technology. And data analysis will continue to have an important role in fundamental research and manufacturing improvements. Analytics will be key to the coming century of innovation, but not the only key. If we hope to accomplish great advances in technology in the decades to come, we’ve got to set high standards, and once again define innovation as much more than mere newness or novelty.
