Google's (NASDAQ:GOOG) (NASDAQ:GOOGL) head of driverless cars, Chris Urmson, recently predicted that its driverless cars could hit public roads within two to five years. That bold declaration complements the tech giant's previous goal of putting at least 100 driverless vehicles on public roads throughout 2015.
Yet Google doesn't have any car manufacturing aspirations. Instead, it will possibly sell its self-driving system to manufacturers, which would tether more users to its Google Maps ecosystem and ferry passengers across a Google-controlled smart grid.
Google's prototype vehicle, which it unveiled last year, replaces the steering wheel, accelerator, and brakes with simple "stop" and "go" buttons. The fully automated system could reduce the number of car accidents, which claimed between 32,000 to 43,000 American lives annually over the past decade, and get seniors and people with disabilities back on the road.
Four states -- California, Nevada, Florida, and Michigan -- have already legalized driverless vehicles, but I believe that Google's driverless cars could eventually hit a dead end due to three factors -- its proprietary mapping process, high costs, and inevitable privacy concerns.
1. A proprietary mapping process
Google's driverless cars can't drive down any road previously mapped out on Google Maps. Instead, the roads need to be exhaustively scanned again in "high definition." Last year, Google mapped around 2,000 miles of road in this manner, but it still has a long way to go -- California alone has 170,000 miles of public roads.
Google must scan roads in high detail because the rotating laser on top of its vehicle, which "reads" its surroundings in 3D, can't process all of that information in real time. Google's prototype, which has a top speed of 25 mph, also can't recognize humans, potholes, and temporary traffic signals, according to MIT Technology Review.
Meanwhile, the auto industry won't simply wait for Google to sluggishly upgrade its systems and map out roads. Around 90% of leading auto manufacturers have already signed deals with MobilEye (NASDAQOTH:MBBYF), a maker of "crash avoidance" systems which calibrate cameras and radars with a vehicle's brakes. Mobileye's system automatically applies the brakes if drivers get too close to surrounding objects, other vehicles, and pedestrians. It can also react to traffic lights and signs.
MobilEye has also developed a "semi-autonomous" system which lets vehicles cruise at highway speeds in the correct lane by simply scanning lane markers. The system is comparable to a self-steering version of cruise control, although lane changes must still be done manually. In my opinion, simpler systems like MobilEye's are more practical than Google's proprietary one.
2. An outrageous price tag
Another big problem for Google is the current cost of its driverless car, which is reportedly outfitted with a whopping $250,000 in equipment. Granted, that price will come down as Google improves the technology, but it makes the company's goal of getting its vehicles out to the public within five years seem utterly unrealistic.
By comparison, MobilEye's ADAS (Advanced Driver Assistance System) only costs around $1,000 to install, making it a much more reasonable choice for automakers and aftermarket consumers.
That's probably why Google hasn't announced any partners for its autonomous vehicles yet, even though The Wall Street Journal reports that it held talks with several Detroit-based automakers in the past.
3. Privacy and security issues abound
Google already tracks users via GPS, wireless networks, and Google Maps, but controlling the transportation grid could open the floodgates to all new privacy and security concerns.
For example, a court could subpoena your autonomous driving records like phone records. Insurers could track their customers' driving history to deny them coverage based on "risky" lifestyle choices. Governments might start zoning areas to prevent driverless vehicles from entering. Hackers, who already hacked Google's "smart home" Nest thermostat last year, could possibly hijack vehicles.
All these scenarios indicate that cheaper semi-autonomous vehicles, which can react to surroundings without leaving a trail of digital breadcrumbs, could be preferable to handing the keys over to Google.
Shooting down another moon shot
Google is known for some great ideas, but it's also known for "moon shots" that make sense in Silicon Valley but aren't practical elsewhere.
For example, Project Loon -- its effort to deliver Wi-Fi to low income countries with weather balloons -- made little sense because many people in those countries couldn't afford computers. Google Glass seemed like a revolutionary device, but it flopped due to the creepiness of wearing a $1,500 camera on your face.
Google's driverless vehicles similarly leapfrog over human habits and market scalability with cutting edge technology. As such, Google's "go it alone" strategy with driverless cars remains oddly disconnected with the ongoing "semi-autonomous" evolution of the auto industry, and could eventually lead its costly project into a dead end.