What Is the Google Car, and Why Is It Important?
Google is a technology leader in autonomous transportation, and any Google Car that makes it to market is expected to have the capability of being fully autonomous. They currently have 34 podlike electric prototypes, and 23 Lexus RX450h crossovers equipped for autonomous driving are being tested on public and private roads.
Google touts the autonomous technology as a way to provide safe mobility to people who otherwise could not be mobile, including the visually impaired, elderly, or otherwise physically challenged. Google argues that autonomous vehicles could be safer than those piloted by humans, as 94 percent of accidents in the U.S. involve human error, according to the company.
Nobody outside of Google or its partners knows what the product they bring to market will look like or whether it will even be badged as a Google product. The company recently inked a deal with Chrysler to convert about 100 2017 Pacifica minivans with autonomous technology.
Don’t confuse the Google Autonomous car with the Google Street View fleet that is constantly traversing the world’s roads. The Google Street View fleet uses roof-mounted 360-degree cameras to gather images for Google’s street-level mapping product, but those cars are standard vehicles modified with cameras and computer systems to gather the stream of data.
How Does the Google Car Work?
The Google prototype cars use data from a number of sensors, including laser, radar, and cameras to create constantly updated images of what’s going on around the car. Computers then interpret that data to classify the objects they see and how those objects might interact with the car.
The computers can differentiate between pedestrians, cyclists, and other vehicles and then predict what their intentions are in relation to the Google Car. The systems then determine the best course of action for the Google Car to take to get to its destination while obeying all the rules of the road and without colliding with anything in its environment.
Some of the cars can operate at highway speeds and not be impediments to the normal flow of traffic while others are limited to 25 miles per hour.
What Makes It Different?
The Google Car will be quite different from a traditional car. It will essentially be a technology with a car wrapped around it, rather than a car adapted to carry autonomous driving technology. Google is even working with regulators so they’ll be able to sell a vehicle with no steering wheel.
Unlike many of the semi-autonomous vehicles on the road today that rely on pavement markings to stay in their lanes, the Google test fleet uses extremely accurate three-dimensional maps that include details such as how high the curbs are and how tall the traffic signals are. The technology allows the Google Car to know its position within 10 centimeters without relying on GPS data.
Many of the autonomous car concepts that we’ve seen from others have familiar interiors, with a dashboard and steering wheel like a traditional car, and appear to be designed to go in and out of autonomous mode, kind of like an airplane’s autopilot. In fact, that’s what Tesla calls their current semi-autonomous driving mode – Autopilot.
Google appears to want to take the technology a step further, eliminating those controls completely from the vehicle or delivering manual driving functions in an all new way. Given the current prototypes’ reliance on mapping data, it would seem as though some manual driving mode has to be included for unmapped areas.
What Have We Seen So Far?
The Google Autonomous driving program began in 2009, though many of the people working on it had years of prior experience with the self-driving challenges of the Defense Advanced Research Projects Agency in the mid-2000s. The first test vehicles that Google deployed were modified Toyota Prius hybrids and an Audi TT Coupe.
Google currently operates a fleet of autonomous test vehicles including 23 modified Lexus RX 450h hybrids and 34 electric vehicles produced for them by Roush Enterprises. Testing locations include Kirkland, Washington; Mountain View, California; Phoenix, Arizona; and Austin, Texas.
The vehicles produced by Roush for Google are limited to a maximum speed of 25 miles per hour, making them suitable for urban and campus testing only.
Google’s autonomous cars have been involved in a number of minor accidents, but Google claims that all but one were the responsibility of the driver of the other car or occurred while the Google vehicle was under human control. In the one accident that they claim some responsibility for, the Google Car collided with a transit bus after the Google Car and the bus driver incorrectly made assumptions about what the other was about to do.
When Will We See a Google Car Available to the Public?
In an interview last year with International Business Times, Google Car’s project director, Chris Urmson, indicated that the company will have a self-driving car ready by 2020, though in another interview with Automotive News he said only that the car will come to market “when it’s safe and ready.”
One of the company’s founders stated in 2012 that a Google Car would be available to the public by 2017. Sergey Brin’s initial timeline now seems to have been overly optimistic.
The 2020 timeline puts Google in the ballpark of when Tesla’s Elon Musk has indicated his company will be able to bring a fully autonomous car to the market. Part of Google’s timing will depend on whether they build the cars themselves or bring in an established manufacturer as a partner to build the vehicles.
Complicating the timing will be dealing with government regulations, both in the testing of prototypes under varying conditions and for end users. It’s an area of the law where the rules haven’t caught up with the available technology. Right now there’s a patchwork of states that allow testing and those that don’t, and there will need to be national standards for both testing and operation before widespread distribution seems feasible.
Partnerships
Google recently made headlines with the announcement of a partnership with Fiat Chrysler Automobiles to convert about 100 2017 Chrysler Pacifica Minivans to operate autonomously using systems that Google has developed for their current prototypes. The first modified minivans are expected to hit the streets by the end of 2016.
Google has already been working with major automotive suppliers including Continental AG, Robert Bosch, Roush Enterprises, ZF, and LG Electronics. LG supplies the battery packs for their current podlike prototypes that are assembled by Roush Enterprises. Some of the prototype’s microprocessors come from Nvidia, a supplier of graphics processing and display technology.
Industry Experience
Google’s no stranger to the automotive industry and its product life cycles, with many vehicles on the road already employing the company’s Android Auto and Google Maps technology. Android Auto allows the infotainment system to channel information from the driver’s smartphone, creating a familiar user interface.
Many vehicle navigation systems use Google Maps data, and automaker Audi extends the technology one step forward, augmenting the map data with Google Earth images of the car’s surroundings.
Who They’ve Hired
Last fall Google made a splash by hiring John Krafcik as the CEO of its self-driving car project. Krafcik brings a wealth of product and industry experience to the company following stints with Ford, Hyundai, and TrueCar. He served as Hyundai’s president and CEO for five years.
The industry took Krafcik’s hiring as a signal that Google is very serious about bringing a car to the market in the coming years. Chris Urmson, who led the autonomous car project from its beginnings, is continuing to lead the company’s technical efforts.
Google has continued to hire staff for the project in marketing, engineering, and manufacturing. In April they hired Apple’s former iPhone global supply manager Daniel Munoz. In December Google snagged Robert Rose, one of the top engineers in the Tesla’s Autopilot program.
What Could Slow Them Down?
There are some hurdles that must be overcome before an autonomous car comes to market. As advanced as Google’s cars are, they’re limited to routes that have been precisely mapped. That mapping must become significantly more widespread before Google’s project can gain wide acceptance.
While the Google Car has been trained to respond to many driving situations, there are still some that it needs work on. Imagine an accident scene with a police officer directing traffic into what is normally the oncoming lane. Such exceptions to normal driving conditions will be a challenge to any autonomous product.
It will take some time for the public to gain enough trust in the systems for them to gain widespread acceptance. Google is working hard to raise people’s comfort level with the technology, but they face an uphill battle in the eyes of consumers. A recent University of Michigan survey showed that 4 out of 5 respondents have concerns over the safety of autonomous vehicles.
It will also take some time for regulation and legal opinion to catch up with the technology. One common question involves liability for accidents. Is it the car’s fault, the occupant of the vehicle, the manufacturer, or maybe even the vehicle’s data provider who is ultimately held responsible?
What’s Next?
There’s going to be a lot happening in the next few years regarding autonomous vehicles, and Google is likely to end up as just one of the players in a crowded market. Other manufacturers already have nearly autonomous cars on the market, and in some ways Google is playing catch-up.
But that weakness could also prove to be Google’s strength. Unburdened by a history that tells them something is impossible, Google’s clean sheet approach may find a way that is possible. They will also have the ability to create partnerships with appropriate teammates, and they have the resources to invest in overcoming any hurdles placed in their way.
Watch for them to continue to educate regulators and the general public on the importance of autonomous vehicles and their potential to save lives.
Is it reasonable to expect an autonomous car to be available by 2020? Maybe not, but it’s probably not going to be long after that. Will it be a Google Car? It’s too early to tell, but the company does have an amazing amount of resources and depth of knowledge to invest in the project.
Recommended Articles
-
2019 Toyota Tundra vs. 2019 Chevrolet Silverado: Head to Head
-
2019 Honda Pilot vs. 2019 Acura MDX: Worth the Upgrade?
-
2019 Chevy Equinox vs. 2019 Toyota RAV4: Head to Head
-
2019 Nissan Rogue vs. 2019 Nissan Murano: Worth the Upgrade?
-
2019 Nissan Rogue vs. 2019 Toyota RAV4: Head to Head
-
12 Cheapest Lease Deals Under $300 in October
-
Best Minivan Deals in October
-
2019 Best Cars for the Money
-
2019 Best Cars for Families
The Google Car
The auto industry as we know it is about to be joined by a couple of new players, though you already know their names and products well. Apple and Google are rapidly pursuing the automotive market both as technology providers and, perhaps, as carmakers themselves.
Of the two, Google is much further down the road, with several generations of autonomous cars already in testing. The company’s vehicles have charted more than 1.5 million miles of computer-controlled driving. They’re currently logging between 10,000 and 15,000 miles of autonomous vehicle testing per week on public streets.
Might you have a Google Car in your driveway in the coming years? Check out the slide show to see how Google’s working toward that goal.