Google released some footage this week of people in the company’s self-driving car in a cutesy video reminiscent of how a toy company would show newborns engaging with mobiles, colorful blocks and stuffed animals. The reason this video merits our attention is that Google is not an automotive company. Many companies that make cars are experimenting with self-driving cars, and we think that they can find success given that they know a lot about cars in general.
Google, on the other hand, has the blessing and curse of not knowing how to make a car. The parallel between this video and a toy company is not dissimilar in the sense that many people would approach this type of vehicle with more curiosity and trepidation than they’ve encountered anything since their very first years. It’s also a parallel that makes Google the newborn in a way. Though they possibly employ and/or consult with former automotive engineers, there is a sense of naiveté in the company’s efforts to not only make a car, but make one that has a wholly different experience for consumers.
The car has no steering wheel. There are no pedals. It has two seats, safety belts, and a start/stop button. Sensors and software speed or slow the car as it whisks you to your destination, just like a taxi. It’s just a prototype, and thus probably runs only a preprogrammed route, but the principle seems to be that you would input your destination in Google Maps, press start and let the car do the rest. One lady in the video commented about how it slowed down before the turn and accelerated out of it, just like she was taught in high school and how she’s always telling her husband to do. It’s like an automated taxi that completely does all of the work for you and is faster at seeing trouble and avoiding it than a human could ever possibly be.
This is made possible by software, which Google is very good at making. Google has been working with cars for years thanks to their Street View and Google Maps products. Combining that knowledge with the mechanics of collision avoidance and user input is the challenge at hand. By the way, it’s worth noting that the amount of programming code in the average car is massive. Like, a lot. An infographic prepared by independent sources shows that the tens of millions of lines of code in regular cars is more than the operating system that your computer or phone is running right now to let you read this article. A modern car needs more programming code to run than Microsoft’s Windows, Apple’s OS X, Facebook, the Hadron Collider, an F-35 Fighter Jet, Google Chrome or the Mars Curiosity Rover. Think about that: cars, which have been around for more than a century, need more computer code to run than computers.
Clearly the automotive industry is ripe for change in the software department. With programming code, more does not necessarily equal better. It could be that Google can do code for cars better than most automotive makers, and would make their car quite viable. That in turn could spur automakers to streamline their code, which would make everyone’s products better. Google’s self-driving car is thus a good thing, we think, because it raises competition and introduces innovation. It’s goofy looking, but we don’t mind. The new kid is often made fun of for looking funny. We love how the video points out the advances in mobility a self-driving car could enable: blind people, the elderly, those without a license, those who don’t know how to drive, etc. People now can use their driving time to do other things, which presents a tremendous potential boost in leisure and productivity.
Google will be testing about 100 of their self-driving cars in real-life conditions for the next 12 months or so. They will use real people as testers and no doubt learn a lot. We wish them well in their enterprise, and look forward to the great things that will emerge from their work.