Driverless cars conjure up the “Johnny Cabs” in Total Recall powered by those creepy androids, or Minority Report—ironically, both Philip K. Dick visions.
But life is finding a way to mimic a sci-fi head-trip (the U.S. will have autonomous cars within a decade). There’s a lot of news on the self-driving front: below, some questions answered on the how, who and when.
Who’s making the driverless cars? Although Google’s “Prototype” is a celebrity, other major auto engineers are crafting their own iterations, such as Ford, GM, Honda, Mercedes Benz, Audi and Tesla.
There’s also a supermarket of other tech firms auto manufacturers are aligning with. For instance, Apple’s “Titan” has tapped figures at Tesla, Volkswagen, Nvidia and A123 Systems; BMW has partnered with Baidu, China’s Google, in creating a venture; and GM has combined their resources with Chinese auto manufacturer SAIC.
Automakers have also heavily invested in Delphi, an automotive technology outfit that develops high-end software and sensors to be retrofitted into currently operating models, such as Audi’s SQ5.
How do they operate? Self-driving vehicles are kind of like people: they need a body, eyes and a soul to interact with the environment. Using Google’s “Prototype” as a lens, the body is embedded with sensitive sensors and radars while other obsolete interior parts are removed and replaced with outsourced artisan engineering.
As for the eyes, Velodyne ‘s LiDAR is a company that creates remote sensing technology utilizing lasers to map out surrounding milieu usually accessed in agriculture, geology and military applications. Mobileye, also a craftsman camera developer, offers high-end video resolution and panoramic visibility. Most auto engineers have partnered with them as a more cost effective alternative.
The backend needs to process all the incoming data. One such company is Nvidia, which produces the hardware. Several manufacturers have sourced the hyper processing power, which is based on an AI phenomenon, known as “deep learning,” which interprets volumes of visual data.
When will we see them on public roads? Cultivating a thoroughly autonomous car is a process. As such, The US Department of Transportation has already drafted a policy sheet defining the stages of development, ranging from level 0 to level 4—level 0, where the human driver has complete control, and level 4 where all safety functions and inputs are controlled by the vehicle.
Analysts predict close to 2025, which will be dominated by the sharing economy in major metropolises, such as Uber. By 2035, 12% of global sales will come from robo-taxis, while sales of fully autonomous cars will reach $39 billion, 10% of the total market for light vehicles. Thus, fully self-piloted cars won’t flood the industry until around 2050.
How much will it cost? If you want to buy one, an estimated $7000-$10,000 extra will be added to sticker prices in 2025.
What are the current regulations? Like previous tech leaps, there’s a wake of confusion and caution. The federal government controls how cars are made and only mandates manufacturing standards, i.e., airbags and ABS braking regulations; however, states control how vehicles operate on their roads. As a result, certain states have penciled laws that are vague.
However, following President Obama’s most recent state of the union address, the Department of Transportation has invested $4 billion over 10 years towards R&D pilot programs throughout the nation. Also, the administration is spearheading an effort to work with automakers and other influencers to develop a basic state-based policy, which would eventually lead to a national model.
Liability is a necessary concern. For instance, Consumer Watchdog has cited safety hazards throughout the R&D stages including defective vehicles, recalls, the Volkswagen scandal and Google’s baby falling short of expectations. Not to mention incorporating programmed robots that will follow traffic laws to their destruction, like Terminator into society with chronic road rage.
How will self-driving cars affect the economy? Morgan Stanley recently conducted an analysis and found that driverless cars would save the US economy $1.3 trillion a year.
A snapshot reveals $158 billion/year could be saved in fuel economy with an eventual average of 54.5 miles per gallon in 2025 resulting from a smoother ride and cruise control without manual braking and throttling and less congestion.
Accident savings would amount to $563 billion, currently costing $7.9 million per death with 32,885 deaths and 2.24 million injuries, 90% of which are caused by human error. And there would be productivity gains allowing people to work without having to focus on driving. Morgan Stanley projects that $422 billion extra could be generated.