How Big Data Knows My Son Will Never Learn To Drive

Driverless cars — technically known as autonomous cars — are advancing so fast, I predict my 2 year old son will never learn to drive.  He won’t have to.

As Bill Gates once said, “We always overestimate the change that will occur in the next 2 years and underestimate the change that will occur in the next 10.” In this case, we’re talking 14 years, when my son will be 16. As I explain below, I believe that, by the time he’s allowed to learn to drive, over 50% of cars sold won’t need a driver.  Heck — why wait until he’s 16?  I thinks it’s likely that he will have been “driving” — that is, travelling alone in an autonomous vehicle — since age 14 or 15.

Big Data On Wheels

Google is the most famous company working on autonomous cars.  And rightly so.  They were one of the first to identify autonomous cars for what they really are: Big Data projects, where the road is just another massive data source for computers to analyze and comprehend.  It’s a complex process that, as I explain in my Big Data presentations, involves 3 of the 5 Big Data data types: geographic data, sensor data, and media data. Google is a market leader in this space not because they know how to build cars, but because they are exceptional at understanding and manipulating data.

Even In 2 Years …

Even if we heed Bill Gates and lower near term expectations, we can’t deny what’s already happened:  Google has driven around 1,000,000 miles on public roads.  Delphi, the automotive parts company, just drove across the United States in an autonomous car.  Holland recently authorized the first fully autonomous vehicle on public roads (not even a driver in the vehicle!) to begin operation at the end of this year.

Even cars you can buy later this year are becoming autonomous: Volvo will offer autonomous functionality on its XC90 car at speeds below 35 mph.  Audi will offer a similar option and roughly 60% of customers appear ready to pay the extra $2,550.  And several weeks ago, Tesla announced a software update to existing Model S already on the road that will allow the car to operate autonomously on the freeway.  A software update!  Once a car has a few sensors and video cameras, what makes it autonomous is software.

What About 10 Years?

Most predictions say autonomous cars won’t be mainstream in the 2020s and instead reserve widespread adoption for the 2040s.  But I think Bill Gates is right and those predictions are wrong.  The benefits for autonomous cars are so compelling, and the economics of software so elastic, I predict 50% of the new cars sold by my son’s 14th birthday in 2027 will be fully autonomous.  Here’s why:

  • First, you can’t have widespread adoption unless there’s something to adopt.  And there will be, sooner than most people think.  Audi has announced plans to offer a fully autonomous car by 2018. Google, Ford, and Nissan all say they’ll offer autonomous cars by 2020.  Jaguar plans to provide them in 2024.  Honda won’t announce a date, but already has a working prototype on public roads.  By most predictions, 2020 is the year most auto makers will offer autonomous cars.  The market will still have another 7 years before my son turns 14.  Remember that number: 7.
  • Second, the value proposition is amazingly compelling.  This is actually three compelling value propositions, are three reasons, each one alone a game changer:
    • Autonomous cars are 78% safer.  Which isn’t surprising when you consider that human error accounts for over 90% of road accidents.  Indeed, autonomous cars may represent the single greatest improvement in automotive history since all safety features combined have “only” reduced motor vehicle deaths by 66% from a high in 1937 to a low in 2013.
    • Autonomous cars will greatly reduce traffic. That’s because they have such quick reflexes, autonomous cars can be 8 times closer to each other on the freeway and still hit the breaks fast enough.  So even if rush hour traffic is 8 times heavier than full speed traffic, well, you’ll still be going at full speed.  In fact, I bet the pressure to let autonomous cars go full speed will be so strong, I wouldn’t be surprised if HOV lanes were converted to Autonomous Car lanes, with 8 times the traffic, but still going full speed.  Indeed, cities will likely increase the number of autonomous car lanes until human drivers are relegated to a single lane… or banned from highways altogether.
    • Driving is no longer wasted time.  Autonomous cars will let you work, sleep, or watch a movie while in transit. How much is that worth to the average person who spends close to an hour a day (the US average) stuck in a car?  How often can you spend a few thousand and get an addition 5 hours a week of your life back? And that’s just commuting. It’s even more if you include all driving.
  • Third, the cost is minimal.  Audi is selling their current autonomous offerings for under $3,000.  But remember, the valuable part is just software.  And when everyone makes autonomous cars, market forces will push costs down to actual cost of production, which, for software sold by the millions, is amazingly cheap.

How Did I Ever Live Without This?

So, there you have it: amazingly valuable and amazingly cheap.  And what was the last technology that was such a leap forward everyone couldn’t believe how they lived without it. (Which is my personal test for the “killer app”).  I think the best example is the smartphone.  Apple introduced the iPhone in 2007, and by 2014, smartphones were 67% of the US mobile market.  In cities, which will benefit from the most from autonomous cars, smart phone adoption is almost 75%.  Just 7 years from introduction to becoming standard.  Remember — that’s the magic number when my son will be 14 years old.

To make the comparison more apples to apples, let’s relate monthly cost to total cost.  At $25 a month in average data cost over those 7 years, most consumers have paid over $2,000 in data charges for their phone.  For people who finance their cars, $25 a month is about the cost of the privilege of having an extra 5 hours a week (more if you include non-commute driving).  People will differ, but I believe that most people will put smartphones and driverless cars in the same “How did I ever live without this????” category.  Nonetheless, even though smartphones achieved 67% of the market in 7 years, I predict just 50% of the market will be autonomous in 7 years, by my son’s 14th birthday in 2027.

So if you’re like me and live in a city and not an early adopter, but just a mainstream adopter, I believe you are more likely than not to own an autonomous car by my son’s 14th birthday.