Tag Archives: autonomous vehicles

fatal automated vehicle crash in Arizona

A Google automated vehicle has killed a pedestrian in Arizona, which has some of the loosest regulations on testing them. I learned about this from the Bicycle Coalition of Philadelphia.

This is certainly a tragedy. The Bicycle Coalition seems to condemn Google for testing the car, then goes on to make the point that human drivers kill people every day, every one is a tragedy, and there isn’t much public outcry about it. I agree with this, and yet I find it interesting that logic and our gut feeling about the morality of the situation seem to be so different. Imagine that changing all cars to self-driving ones would cut the number of people killed by 50% (and I have seen estimates of much larger reductions than that). It would seem immoral not to make that change. But at the same time, it would seem immoral to unleash a fleet of robot cars, knowing that a certain fraction of them are going to kill people, and by killing a few people learn how to not kill as many people in the future. I don’t know the answer to this, except that the technology will gradually get better, and insurance companies may eventually decide the human drivers are not worth the risk.

2018 transportation trends

U.S. News has predictions for transportation technology in 2018. In a nutshell, sales of electric vehicles will take off, but people will also keep buying inefficient gas vehicles because is relatively cheap right now.

According to this article, the commercial rollout of autonomous vehicles has suffered some setbacks, and 2018 might not be the big widespread adoption year.

Automated vehicle developers will push back their timelines for deployment, for good reason. Automated vehicles are coming, but the real question is when. Just over a year ago, tech magnate Elon Musk said he “felt pretty good” about a Tesla driving completely autonomously from Los Angeles to New York without any human interaction by the end of 2017. But at a recent conference, he pushed that date back another two years. Chevy delayed the debut of Super Cruise, and driverless shuttles have yet to move beyond pilot phases.

Society will greatly benefit from automated vehicle safety improvements. However, it’s a good thing that automated technologies are delayed. Automakers are finding it more difficult to design the system than they originally expected. Instead of putting a product on the road that is unsafe, they are responsibly taking the time they need to make sure the system is ready for the public. Patience will pay off in the long run.

I wonder if it is really a setback in technology, or a matter of a few high profile accidents getting a lot of media attention. Markets and regulatory agencies are going to respond to perception, no matter how clearly the statistics show that imperfect computer-controlled vehicles are a huge advance over human controlled vehicles. Insurance companies are somewhat immune to emotion and responsive to hard numbers though, so at some point when there are safer options available they may just jack rates up on people who don’t take those options. It probably won’t pay to be a late adopter.

RethinkX

this private think tank report on the future of transportation claims to have used a system dynamics approach and to have reached some radical conclusions, like the collapse of private car ownership, the oil industry, and major decreases in the cost of getting around within a decade or so. The buzz phrase is “transportation as a service”.

https://static1.squarespace.com/static/585c3439be65942f022bbf9b/t/591a2e4be6f2e1c13df930c5/1494888038959/RethinkX+Report_051517.pdf

Google may release self-driving taxis in 2017

Google may be about to release a self-driving taxi service in months, not years.

Real driverless cars could come to the Phoenix area this year, according to a Monday report from The Information’s Amir Efrati. Two anonymous sources have told Efrati that Google’s self-driving car unit, Waymo, is preparing to launch “a commercial ride-sharing service powered by self-driving vehicles with no human ‘safety’ drivers as soon as this fall.”

Obviously, there’s no guarantee that Waymo will hit this ambitious target. But it’s a sign that Waymo believes its technology is very close to being ready for commercial use. And it suggests that Waymo is likely to introduce a fully driverless car network in 2018 if it doesn’t do so in the remaining months of 2017…

According to Efrati, Waymo’s service is likely to launch first in Chandler, a Phoenix suburb where Waymo has done extensive testing. Waymo chose the Phoenix area for its favorable weather, its wide, well-maintained streets, and the relative lack of pedestrians. Another important factor was the legal climate. Arizona has some of the nation’s most permissive laws regarding self-driving vehicles.

autonomous vehicles and safety

This article from the University of Illinois says that not only are autonomous vehicles safer than human drivers, but having just a few of them mixed in with the humans is actually safer for everyone.

The presence of just a few autonomous vehicles can eliminate the stop-and-go driving of the human drivers in traffic, along with the accident risk and fuel inefficiency it causes, according to new research. The finding indicates that self-driving cars and related technology may be even closer to revolutionizing traffic control than previously thought.

“Our experiments show that with as few as 5 percent of vehicles being automated and carefully controlled, we can eliminate stop-and-go waves caused by human driving behavior,” said Daniel B. Work, assistant professor at the University of Illinois at Urbana-Champaign, a lead researcher in the study…

The team conducted field experiments in Tucson, Arizona, in which a single autonomous vehicle circled a track continuously with at least 20 other human-driven cars. Under normal circumstances, human drivers naturally create stop-and-go traffic, even in the absence of bottlenecks, lane changes, merges or other disruptions, Work said. This phenomenon is called the “phantom traffic jam.” Researchers found that by controlling the pace of the autonomous car in the study, they were able to smooth out the traffic flow for all the cars. For the first time, researchers demonstrated experimentally that even a small percentage of such vehicles can have a significant impact on the road, eliminating waves and reducing the total fuel consumption by up to 40 percent.

transitioning a highway to driverless cars

This white paper proposes some ideas for transitioning a highway to driverless cars in several steps.

The first step would be allowing autonomous vehicles into HOV lanes, the next step would be dedicating a lane to autonomous vehicles, and the final step would be expanding until all lanes are autonomous only. Regulators may want to phase these steps in with vehicles reaching various levels of the NHTSA/SAE autonomous driving framework (e.g., Level 3 automatic braking and lane control or 4 automatic control of all aspects of driving without the need for human intervention). We believe the first step of allowing autonomous vehicles in the HOV lanes could begin immediately with Level 3 and higher vehicles. A dedicated lane (each way) for autonomous vehicles could occur as soon as we have a significant number of such vehicles on our roads which could be 2025 or earlier if a tipping point has occurred. We can imagine that limiting non-autonomous vehicles to one lane on I-5 could occur beginning in 2030, which would mean at least three exclusive lanes for autonomous vehicles from Seattle to Everett, two exclusive lanes from Everett to Marysville and one north of Marysville to Vancouver’s downtown area.

autonomous trucks

This article from Wired brings up a couple points about autonomous trucks I hadn’t thought of before.

The startup Peloton is working on “platooning” trucks, or groups of vehicles that communicate via a wireless connection that helps them time their movements. At some point soon, the system might let a lead driver take over the steering for a bit, while those at the wheels in the vehicles behind could snooze, catch up on paperwork, meditate, whatever. California-based Embark would like to see driver-monitored trucks pilot themselves on interstates but be manually driven into warehouses by nimbler humans. Starsky Robotics has a similar vision, but says that the trickier driving maneuvers could be done by a human in a remote location, Predator drone-style. It’s unclear if any of these companies want—or will be able—to ever take the human out of the picture entirely…

So you could have one driver or no drives for a whole line of trucks on the highway. Then, when they get to town, a driver or even a remote driver could bring them in one at a time, if the computers are not yet up to the job.

One thing I think about is that this sounds a lot like trains. It’s hard for me to believe that platoons of trucks will be cost-effective with trains even after you take most drivers out of the equation. But remember that the highway system is automatically funded by all highway users through the gas tax, while rail companies are required to build and maintain their own infrastructure. Politicians in rural areas, which are greatly overrepresented in our electoral system at the state and federal levels, like this system because it makes taxpayers in the economically productive areas (aka cities) fund inefficient road networks in the mostly empty rural areas they represent. So this is not a level playing field where victory goes to the most efficient technology.

Logan and autonomous trucks

Not being a big superhero fan, I haven’t yet seen the movie Logan. Apparently, one thing in the movie is autonomous trucks.

About midway through the second act of the new X-Men movie Logan, there’s a scene that brings the superhero film’s vision of the future to life. Our heroes, on the run from the villainous Reavers, happen upon a car accident in the midwest. An autonomous truck vehicle has hit a horse trailer—and with nobody at the wheel, it’s hard to know why it happened.

I won’t go more into detail about the scene, spoilers, etc. But Logan’s writer and director James Mangold’s inclusion of the self-driving trucking machines makes it clear that the filmmaker understands the writing on the wall about the future of shipping. It’s a future without truck drivers.

Maybe as humans we find this idea creepy, but today’s human-driven trucks are incredibly dangerous and the autonomous trucks of the near future will almost certainly save human lives.

Tesla trucks

Tesla is trying out automated trucks that can drive in platoons.

Reuters has learned that Tesla wants to experiment with “platooning,” where several autonomous trucks, some without drivers, follow a lead truck on a highway using autonomous driving tech. This technique is viewed as allowing trucks to travel in small groups, creating some aerodynamic benefits for the following trucks and permitting the drivers in the group to switch to a “follow” mode that could not be accomplished with human drivers tailgating each other. In this platoon mode, all the trucks communicate to stay close, taking directions from a lead truck and reacting accordingly.

Read more: http://autoweek.com/article/autonomous-cars/tesla-wants-test-autonomous-trucks-without-drivers-them-report-says#ixzz4sc2PMY93

It was only a matter of time until attention turned to trucks and buses rather than just cars. One thing I wonder is at what price point would this technology be competitive with rail, given that trucks can deliver door to door? Especially if we continue to subsidize our roads with public money and our rails not at all.