What has four wheels, no steering wheel, no accelerator pedal and no brake pedal?
If your first guess was “Radio Flyer,” think more high-tech. Think more “Googley.” Think more…cartoonish.
“It looks like something out of a Pokemon game, with its inoffensive size and surprised little face,” declared blogger Vijay Pattni of the British motoring enthusiast show, “Top Gear.”
That was no mistake, according to Director Chris Urmson of the Self-Driving Car Project.
“In our case we wanted to find something that’s very Googley,” Urmson was quoted as saying in various media reports. “It’s friendly, it’s kind of cute. We hope it fits into neighborhoods.”
Indeed, media buzz ranneth over with characterizations like “cute,” “clownish” and “wheeled bubble” May 27 after Google unveiled its working prototype of an entirely autonomous vehicle – a significant public update for the Self-Driving Car Project that began in 2009.
Historically speaking, Google’s “handless carriage” is more or less the 21st century’s horseless carriage – one Urmson says will revolutionize the DNA of transportation as we know it.
No hands? No feet? No attention span? No problem.
The zippy automotive “feels like a theme park ride,” quipped Recode Senior Reporter Liz Gannes, who took a test drive in the newest model last month. Other folks were also treated to a joy ride, and gave their feedback to Google:
Unlike earlier prototypes in Google’s fleet of self-driving cars, which consist of modified Lexus SUV hybrids, the latest and cute-as-a-button generation is sans steering wheel, accelerator and brake pedals. All it has (for now) is a screen displaying the time, weather, mph and route; two seats with seatbelts; a space for passenger’s belongings; and two buttons to start and stop.
So how exactly does the robo car get from point A to point B without bumping into fire hydrants, bowling over bushes or careening into cyclists? Google had to map 2,000 miles of road, then create a virtual world out of those streets and pre-load that data into the car’s memory.
The fully electric car uses software and sensors, including a Velodyne LIDAR – or light radar system – with a spinning laser beam mounted on the roof. Using a combination of this technology to measure the environment in 360-degrees around the car, plus high-resolution maps of the real world, the car produces different types of data models that allow it to drive itself.
In a country where 33,561 people died in car crashes in 2012, according to the U.S. Department of Health and Human Services, Urmson makes a point to reiterate the bottom lines: improving road safety, transforming mobility for millions of people, and reducing road congestion/fuel consumption.
“Ever since we started the Google self-driving car project, we’ve been working toward the goal of vehicles that can shoulder the entire burden of driving,” he wrote in a recent blog post. “Just imagine: You can take a trip downtown at lunchtime without a 20-minute buffer to find parking. Seniors can keep their freedom even if they can’t keep their car keys. And drunk and distracted driving? History.”
And then there’s Urmson’s other bottom line.
“People hate driving,” he opined at a recent press event at the Computer History Museum. After you get to work in the morning, “it takes 30 minutes to decompress from that [jerk] who cut you off.”
The guys over at Top Gear, the Duke cousins, NASCAR and everyone else who doesn’t hate driving would probably grimace at that first remark [See: Jeremy Clarkson gingerly test drives a self-driving BMW 330i, and screams like a little girl].
What Urmson probably meant to say was, “people hate traffic.”
Can we get an “amen,” Bay Area commuters?
Google has announced its plans to build about 100 prototype vehicles, and safety drivers will be testing early versions with manual controls later this summer in Mountain View. The company will eventually roll out a small pilot program in the next couple of years in California, and hopes to have the cars ready for public use between 2017 and 2020.
The latest mile marker in the world of self-driving technology heralds a futuristic trend zooming towards us at the exhilarating speed of, well, 25 mph, in the case of Google’s clown car.
“Self-driving cars are coming. That’s essentially a given: the technology already mostly works, and nearly all automakers believe autonomous vehicles are both a good and feasible idea,” writes the Verge’s David Pierce. “The biggest remaining challenges appear to be regulatory rather than technological, as governments start to answer questions like who’s responsible when a self-driving car gets in an accident.”
Hitting the road with intelligent imaging
We’re big nerds for intelligent imaging here at ArcSoft, so we’re always dreaming about how our software – which enables visual ‘thinking’ capabilities in your favorite electronic devices and appliances – can help make smart devices even smarter.
In the case of Google’s self-driving car, some limitations to consider are the fact that it can’t understand hand signals from traffic cops, or detect small animals that might dart out into the road [something Conan O'Brien is gravely worried about. See: "Google's cutting-edge new driver-less vehicle prototype has no steering wheel... or mercy."]
It doesn’t fare so well in blizzards or torrential downpours, either.
“Snow on the ground and water kicked up by other cars messes with the spinning laser that sits on the roof, while fog limits how far the radar can see,” observed Wired author Alex Davies.
These are just a few scenarios where imaging processing technologies such as gesture recognition; face detection/recognition; and object/scene detection software could possibly refine and improve the overall functionality of the self-driving car.
Another factor? Cost reduction. As Extreme Tech’s Sebastian Anthony points out:
“Disappointingly, Google’s new car still has a ton of expensive hardware — radar, lidar, 360-degree cameras — sitting on a tripod on the roof. This is to ensure good sightlines around the vehicle, but it’s a shame that Google hasn’t yet worked out how to build the hardware into the car itself, like other car makers that are toying with self-driving-like functionality.”
Lower-cost 3-D imaging sensors could potentially replace the expensive LIDAR while achieving comparable functionality. This might significantly reduce the manufacturing cost of the car, and may help to expedite the availability to consumers.
Which would be great, because it’s about time someone other than Batman got to scoot around town in a self-driving car.