Do Autonomous Vehicles Get In Crashes or Accidents?

Brad Templeton has a detailed discussion of the issues that arise from collisions involving self-driving cars. The piece is long and worth reading, but why does he call the incidents "accidents"? There is sustained effort to replace the commonly used accident with crash (see here) to more accurately reflect that crashes are preventable, though some more so than others. With autonomous vehicles major sources of crashes such as drunk driving will be eliminated (unless Bender takes the wheel), but there will still be some crashes. These should not be called accidents.

In The Future No One Will Know How To Drive

The above video is of a New Scientist reporter experiencing a "road train." From the story:

The most alarming thing about taking your hands off the steering wheel when hurtling along the road at 90 kilometres an hour is just how quickly you get used it. There is a brief moment of initial uncertainty, but then you quickly stop worrying about who is control and just sit back, relax and enjoy the ride.
Welcome to the brave new world of semi-autonomous cars. I say "semi" because this car is not entirely driving itself. Volvo calls it a platoon: a convoy of moving cars that are wirelessly coupled together, one behind the other, into a road train, all under the control of a single professional lead driver.

I do expect that people will get used to not driving pretty quickly, though the road train technology doesn't seem as likely to take off as fully autonomous vehicles. Here is another story about autonomous cars that explains that cars communicate around corners and substantially reduce collisions. And here is a CNN/Wired piece that suggests no one will need a drivers license by 2040.

Protect Florida from Driverless Cars and Toll Roads, Says the Committee to Protect Florida

Jeff Brandes is a Republican running for state senator in Florida. He also wants to use driverless cars to kill old people using walkers and supports road tolls to pay for infrastructure. I learned this from the attack ad below that highlights some of Brandes' work as a state Representative.

Road Trains: Like Autonomous Cars, but Different

Live Science writs about the public debut of SARTRE cars in Europe. Unlike Google's autonomous cars, the SARTRE cars communicate with each other and form 'road trains' that allow for semi-autonomous driving in certain situations. From the story:

Smart cars obediently followed the leader in the first public test for a futuristic "road train" concept. The idea of having networked cars follow a truck could pave the way for tomorrow's self-driving cars to drive together.
The recent test featured a Volvo XC60 compact crossover SUV, a Volvo V60 sports wagon, a Volvo S60 compact executive car and a truck allconnected wirelessly to a lead truck. Such networked "talking" allowed the cars to mimic the accelerating, braking and turning of the human driver in the truck — all while traveling at 53 mph (85 kilometers per hour) on a public highway in Spain.

I suspect that these technologies will be introduced prior to lots of fully autonomous cars entering the auto fleet. Highway driving is boring, and people may be happy to give it up. However, the proposed SARTRE road trains do require a human driver in the lead vehicle, which seems like a disincentive to adopt the technology by other drivers. After all, everyone thinks they are above average drivers. So who will lead the train?

Autonomous Car Reviews and Legislation

California's Senate passed a bill that allows autonomous cars to drive on California's roads:

A bill that allows for the use of self-driving cars on California’s roads passed the California State Senate.
SB1298 by State Sen. Alex Padilla (D-Pacoima) would establish guidelines for such "autonomous vehicles" to be tested and operated in California. The bill now goes to the Assembly for consideration next month.
Tech giant Google Inc., Caltech and other organizations have been working to develop such vehicles, which use radar, video cameras and lasers to navigate roads and stay safe in traffic without human assistance. Google has said that computer-controlled cars should eventually drive more safely than humans, who, after all, get sleepy and distracted and can't see in every direction at once.
Padilla said his bill passed without objection, a demonstration of the bipartisan support the technology has engendered.

 Google's autonomous car is reviewed. Apparently the car can't use turn signals yet. You'd think it was a BMW or something:

Since the Google car only just got its learner's permit, it drives accordingly. During our test loop, it stopped a few times for phantom threats, like a parked truck that was just a little wider than the cars around it. Then there was the jerking halt on a side street caused by a car that stopped a little abruptly almost two car lengths ahead.
When it wasn't sure what to do, the car would hand control back to the driver, announcing it was doing so in a friendly female voice. (The driver can always take control at any time by just by moving the steering wheel or touching the pedals, even slightly.) "Self driving" was resumed by pushing a big green button on the Prius's center console near the even bigger red "kill switch."
Surprisingly, one thing the car can't do all on its own is use the turn signals. The driver still has to do that.
"That's been on our to-do list for a long time now," said the engineer riding shotgun.

I Don't Want to Believe This is Real

Presenting the U.S. House of Representatives Unmanned Systems Caucus! It's real. Not a joke. Here are the mission statement and goals:

Mission & Main Goals

The mission of the U.S. House Unmanned Systems Caucus is to educate members of Congress and the public on the strategic, tactical, and scientific value of unmanned systems; actively support further development and acquisition of more systems, and to more effectively engage the civilian aviation community on unmanned system use and safety.

As members of this Caucus, we:

1. Acknowledge the overwhelming value of these systems to the defense, intelligence, homeland security, law enforcement, and the scientific communities;
    
2. Recognize the urgent need to rapidly develop and deploy more Unmanned Systems in support of ongoing civil, military, and law enforcement operations;
    
3. Work with the military, industry, the Department of Homeland Security, NASA, the Federal Aviation Administration, and other stakeholders to seek fair and equitable solutions to challenges created by UAV operations in the U.S. National Air Space (NAS);
    
4. Support our world-class industrial base that engineers, develops, manufactures, and tests unmanned systems creating thousands of American jobs;
    
5. Support policies and budgets that promote a larger, more robust national security unmanned system capability.

And the message from the Chairmen (remember, these are elected Representatives of actual people, not drones):

Chairmen's Message

We want to thank you for interest in this industry. This is an exciting and existing technology that will continue to grow, and improve our lives as public acceptance progresses. The Congressional Unmanned Systems Caucus’ goal is to educate members of Congress on every facet of this industry. We are this industry’s voice on Capitol Hill, and will work closely with industry to ensure we continue to expand this sector through efficient government regulation and oversight. Thank you for visiting our page and please don't hesitate to contact us for further information. 

Sincerely,

H/T to James Fallows, who has more thoughts on this. It's nice of these elected representatives of actual people to serve as the industry's voice on Capitol Hill, especially since the Chair on the Unmanned Systems Committee is also the Chair of the Armed Forces Committee.

If the House were getting behind autonomous cars and driverless trains I may be supportive of this...

This is another example where public choice theory is applicable. I near as I can tell the whole plan here is to maximize the spending on and size of the program regardless of benefits.

Steady Progress Toward Autonomous Cars

New Scientist magazine has a nice story about the slow but steady adoption of autonomous cars. There isn't any new news, but the story highlights the incremental adoption of autonomous driving technologies that has already occurred:

Promising improved safety, better fuel-efficiency and freedom from the boredom of long drives, autonomy has been coming piecemeal to our cars for some time - and it has always had its critics. In 1994, on a UK motorway, Jaguar and Lucas Industries demonstrated the safety of adaptive cruise control and automatic lane keeping; both technologies are now commonplace on our roads. The media were not impressed, describing the idea of cars that drive themselves as "madness".

The incremental nature of technological change is an important factor in transport policy, yet too often transport planning chases "revolutions" where brand new (or redeveloped old) technologies will be widely deployed and adopted. As nice as a revolution may sound, that's just not how change actually occurs.

Rather than one day having a brand new model of fully autonomous vehicles, it is most likely that parts of the driving experience will be computerized (for instance fully autonomous freeway driving, just as fully autonomous parking is now commonplace) before the entire experience will be computerized. By encouraging and supporting incremental advances the technological, political and legal issues will have a chance to develop together, and adoption of new and safer technologies will happen faster. Consider what is happening with electric cars as a similar type of issue with deployment of new technologies. Electric cars are just not very popular as few people are adopting the revolutionary powertrains. The Chevy Volt is on hiatus, and Chevy just announced that it will be on hiatus an additional week. George W. Bush's "Freedom Car" and Arnold Schwarzenegger's "Hydrogen Highway" were failures, as was the entire "hydrogen revolution" because the infrastructure and technology isn't there to support widespread adoption, and still aren't. Maybe someday in the future hydrogen will be a standard fuel, but now people tend to really like their hybrids (Toyota is setting sales records with it's Prius models) which are an incremental improvement to what people already use.

All of this is to say don't hold your breath that autonomous vehicle technologies will first appear in a fully, full-time autonomous car, but also don't expect that driverless cars will not become commonplace. Drivers are already ceding some driving to their cars, and over time they will cede more and more until the driver is simply a passenger. Then someday no one will remember that is was ever any different.

UPDATE: Chevy announced that the Volt will go back into production on the original schedule due to strong March sales. Let's hope the strong sales continue!

UPDATE, Part Duex: Adam Ozimek at Modeled Behavior offers an alternative path to autonomous cars where demand for driverless cars comes from particular groups such as the elderly or disabled. This seems more along the lines of incremental shifts in the demographics of demand rather than incremental shifts in technology, as I argued.

Getting Closer to Autonomous Cars

Some points of interest about autonomous cars:
From USA Today:

 The National Highway Traffic Safety Administration will start studying aspects of autonomous driving in August with a one-year pilot project in Ann Arbor, Mich., to test 3,000 cars with equipment to communicate with one another to prevent accidents. Officials have expressed support for technology that addresses distracted driving and prevents accidents.

Issues still to be resolved include who is liable in a crash and whether drivers of autonomous cars are legally exempt from bans on texting.

And Google's car took a blind guy to Taco Bell:

Much has been written about how self-driving cars may ease traffic and solve the texting-while-behind-the-wheel problem. But there’s another benefit: Such autos will allow people previously unable to drive to become more independent. In the video below, Steve Mahan, who is 95 percent blind, takes a Google self-driving car for a spin.

The promise of increased mobility for visually impaired  travelers is great. Using a "south of the border" run to demonstrate how cool the technology is probably won't gain any converts from people who think Americans are fat because of cars, however. (I'm not one of those people.)

Autonomous Cars and Transport Finance

A few links of interest:
CNN has a video with a UT researcher working on autonomous cars.


The LA Times reports on car to car communication to improve safety.

And two from the NY Times on transport finance:

The federal tax credit for alternative powertrains will increase to $10,000. Will this lead to a corresponding increase in prices? We shall see!

And a post about whether or not electric vehicle drivers should be taxed for road use. As gasoline and diesel decline in importance as a transport fuel for personal travel the conventional way to pay for transportation will also decline. In my opinion, the sooner we shift towards more user fees to finance transport the better.

Two bonus links:

The unions finally have some good news.

Lots of new money for NASA's space taxi. There are lots of private companies also developing space taxis.  The good news is space travel will be cheaper. The bad news is that there will be more space travel, which will accelerate climate change emissions entering the atmosphere.  These effects will likely more than offset any reductions in emissions from ground transport. More bad news is that orbital space junk will make private space travel impossible. Or so says Policymic. Unless this works.