How Fast Is Autonomous vehicle Coming?

I recently read a new book named “Three Revolutions: Steering Automated, Shared, and ElectricVehicles to a Better Future” by Daniel Sperling.

In the book, the author expresses very restrained and rational views on when an autonomous vehicle is used for large-scale commercial use. I also agree with him very much.

AV technology is not new

The author reviews the development history of autonomous technology in the United States, dating back to the 1960s, with several key milestones as follows:

（1）A team headed by an electrical engineering professor at Ohio State University built the first AV in 1962.

（2）The University of California, Berkeley, developed an AV in the late 1980s that was guided by data from magnets in the roadway.

（3）In 1995, a team at Carnegie Mellon University built an automated car that completed a drive (with two researchers in the car) from Pittsburgh to San Diego in a project dubbed No Hands Across America.

（4）In 2004, DARPA issued the first DARPA Grand Challenge, requiring vehicles to leave Barstow, California, and autonomously navigate a 142-mile course across the desert to Primm, Nevada.（from 2004-2007，DARPA has held three Grand Challenge for autonomous car）

（5）It was soon after the DARPA Urban Challenge that announcements from companies like Google and others began to pop up.

Hype outweighs technical realization

The author thinks the hype about AVs has indeed stimulated innovation, bringing new investment and raising expectations. But it is hype！

Gartner, a firm that monitors technology hype for its clients, posits that autonomous vehicles were just past the “peak of inflated expectations” in 2018 and wouldn’t reach mainstream adoption for more than ten years.

According to a recent test by the Insurance Institute for Highway Safety, today’s autonomous vehicle makers Level 2 driver assistance are not capable of driving safely on their own. 

The production a vehicle that can safely drive itself anywhere, anytime isn’t available at the local car dealer and won’t be for quite some time.

The Uncertain Timeline Ahead

Assuming the technical challenges can be met, how soon will driverless vehicles serve the majority of travelers? Probably not for a long time—several decades at least, even in ambitious scenarios. He cited some facts to prove his concern.

（1）     Unlike typical Silicon Valley technologies, which pose minimal safety risks, automated cars will receive intense scrutiny from those tasked with protecting consumer safety. The result is that commercialization happens at a much slower and more deliberate pace than Silicon Valley technology start-ups is used to.

（2）     Car companies have mixed views about whether humans really want to hand over full control of their cars, it looks to be a long leap to a passenger-centric car that excludes a steering wheel and brakes.

（3）     Moving upward to level-4 self-driving and level-5 driverless cars will eventually generate huge safety benefits, but those benefits might not be realized for a long time because of the challenges of mixing driver-operated and driverless vehicles on the same roads.

（4）     Who is responsible for an AV crash? Car insurance and the law is expected to change radically in the era of self-driving cars.

（5）     How does one program an AV to make ethical choices? The right way to program automated cars is not at all clear. “Should people inside or outside the car be protected in an emergency ? ”， “Could you imagine if we had one algorithm and Toyota had another and General Motors had another?”

（6）     Even more troubling is the risk of hacking. hacking. As vehicles become more like computers and robots, they become more susceptible to hackers taking control.
（7） There are nearly three hundred million vehicles in the United States, with a turnover of around seventeen million annually. If every single vehicle sold from 2018 forward were driverless, it would take until 2034 for all the cars on the road to be driverless.
.............................................................