Stratfor explains why self-driving cars won’t rule soon

Summary: Millions of jobs will be lost when autonomous vehicles dominate the roads. Here Stratfor explains that most difficult factor is not building the technology, it is our fear of new technology.

Autonomous Vehicles
Image by Volvo.

Driving Consumers Toward Automated Vehicles.

by Stratfor, 3 November 2017.

Highlights.

  • Recent technological developments will keep automated vehicles on track for a limited commercial launch in the 2020s.
  • Despite the advancements, however, policy initiatives to facilitate the use of automated vehicles — including regulations over the technology and over data privacy — will continue to lag behind.
  • Public acceptance of automated vehicles will be the biggest obstacle to their incorporation into the global fleet and will probably limit their use over the next 10-15 years.

Analysis.

The phrase “self-driving car” is a bit of a misnomer. Despite what the name suggests, automated vehicles do have a driver — just not the kind we’re used to. Developing a computer that is robust enough to operate an automated vehicle, small enough to fit in the car and efficient enough not to drain its power source is a difficult and costly endeavor.

But California-based computing company NVIDIA seems to have solved the riddle with its latest-generation processing platform, Pegasus. Roughly the size of a license plate and 13 times more powerful than previous iterations, the newly unveiled system will meet the requirements to run a fully automated vehicle and will be available starting in mid-2018. Pegasus is just one of the rapid-fire developments in computing power, data processing and artificial intelligence that will bring the automated vehicle industry closer to its goal of releasing the technology onto select markets by the early 2020s.

Once the cars are on the road, however, they will need passengers to revolutionize transportation as we know it. And so far, the technological advancements necessary to bring automated vehicles to the market are outpacing the changes in regulation and public perception that will push them into the mainstream.

Nvidia Drive PX Pegasus

Pegasus, Take the Wheel.

One of the driving forces behind the dizzying pace of developments in automated vehicles is the steep competition in the sector. NVIDIA, one of a handful of industry leaders that haven’t been bought up by a major automaker or tech firm, is on the cutting edge of automated vehicle technology and counts companies such as Toyota Motor Corp. and Tesla Inc. among its customers. But the firm has some formidable rivals in the field. Intel Corp., for example, plans to release a rival system in 2021, working closely with BMW and with Mobileye, an Israeli company that specializes in sensing technology using radar and lidar {story here}.

Beyond processing technology, moreover, several companies in the industry, including tech giant Alphabet Inc. and its subsidiaries, are developing their own platforms to operate automated vehicles.

Nevertheless, the number of firms racing to build better and faster processors and sensors doesn’t mean that autonomous vehicles will be flooding the market anytime soon. NVIDIA intends to roll out a limited run of its new platform at first, mainly in delivery vehicles and cars used for taxi or ride-hailing services.

Driving Levels for Automated Cars

Bringing the World’s Passengers on Board.

Over the next 10-15 years, automated vehicles will probably enter use primarily in these applications. Though lawmakers in several countries are starting to catch up to the existing technology, lingering concerns over safety will limit the use of even intermediate levels of automation in cars. Germany, for instance, stopped short of sanctioning fully automated vehicles of the sort that Pegasus would support when it recently legalized automation in cars, so long as a human driver is behind the wheel.

The U.S. Congress, meanwhile, is trying to refine legislation that would establish a standard federal policy on self-driving vehicle technology. As the technology continues to evolve, new regulatory issues will arise. The more automated vehicles enter the global fleet, after all, the more data they will collect, raising legal questions over privacy. The technology required to put self-driving cars on the road may be developing apace, but public policy is a different story.

And then there’s public opinion. MIT recently conducted a survey that revealed acceptance of automated vehicles, and especially fully automated vehicles, may be decreasing. Views of the technology vary among different age groups — younger drivers are more open to self-driving cars — but even so, worries over the safety and reliability of automated vehicles are prevalent. Furthermore, when fully automated vehicles hit the roads, they will have to interact with human drivers, and human error.

The potential benefits of automated vehicle technology on supply chain efficiency, urban congestion and safety may make its eventual incorporation into the global fleet inevitable. Without technological improvements like NVIDIA’s Pegasus, the transition would be impossible. But how quickly the cars enter use depends more on changes in policy and behavior than it does on changes in technology. Though the automated vehicle industry may be ready for self-driving cars, persuading the rest of the world to hop in and buckle up will take some time.

Driving Consumers Toward Automated Vehicles
is republished with permission of Stratfor.

From the MIT study.

Consumer Interest in Automation: Preliminary Observations Exploring a Year’s Change“.

By Hillary Abraham et al. at MIT Agelab, February 2017.

A third of people do not want their car to help with steering or determining the right speed. Almost-two thirds do not want the car to take control of the driving.

MIT study about autonomous vehicles - figure 5

The young are more willing to use automation — but not a lot more willing. And everybody’s enthusiasm for robot drivers dropped from 2016 to 2017, showing that these attitudes have shallow roots — and are easily influenced.

MIT study about autonomous vehicles - Table 3

——————————————

Stratfor-Worldview

About Stratfor

Founded in 1996, Stratfor provides strategic analysis and forecasting to individuals and organizations around the world. By placing global events in a geopolitical framework, they help customers anticipate opportunities and better understand international developments. They believe that transformative world events are not random and are, indeed, predictable. See their About Page for more information.

For More Information

The new industrial revolution has begun. New research shows more robots = fewer jobs. Also see the famous book by Wassily Leontief (Nobel laureate in economics), The Future Impact of Automation on Workers (1986). Also see the Frequently Asked Questions page at the website of the Machine Intelligence Research Institute.

If you liked this post, like us on Facebook and follow us on Twitter. See all posts about robots and automation, and especially these…

  1. A warning about the robot revolution from a great economist.
  2. How Robots & Algorithms Are Taking Over.
  3. Economists show the perils and potential of the coming robot revolution.
  4. Three visions of our future after the robot revolution.
  5. The coming Great Extinction – of jobs.
  6. Lessons for us about AI from the horse apocalypse.
  7. A Timeline for the Extinction of Jobs by Machines.

Books about the coming great wave of automation.

Rise of the Robots: Technology and the Threat of a Jobless Future by Martin Ford (2015).

The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies Erik Brynjolfsson and Andrew McAfee (2014).

Rise of the Robots
Available at Amazon.
The Second Machine Age
Available at Amazon.

 

8 thoughts on “Stratfor explains why self-driving cars won’t rule soon

  1. Strafor really falls for the self-driving hype here. They just ignore fundamental technical issues on the path to full automation that have not been solved. The director of the Carnegie-Mellon University Robotics Institute, Herman Herman, provides a more sober perspective: “The Head of CMU’s Robotics Lab Says Self-Driving Cars Are ‘Not Even Close’” in Medium, Sept 2016 — “A conversation on rebuilding after Uber gutted the top lab.”

    “With autonomous cars, you see these videos from Google and Uber showing a car driving around, but people have not taken it past 80 percent. It’s one of those problems where it’s easy to get to the first 80 percent, but it’s incredibly difficult to solve the last 20 percent. If you have a good GPS, nicely marked roads like in California, and nice weather without snow or rain, it’s actually not that hard. But guess what? To solve the real problem, for you or me to buy a car that can drive autonomously from point A to point B—it’s not even close. There are fundamental problems that need to be solved.”

    Others have cautioned not to expect full automation in our lifetimes: “The Hurdles Facing Autonomous Vehicles” by Preston Lerner, Automobile in 22 June 2015. Near, but so far away.

    Stratfor is basing their analysis (pretty credulously IMO) on a NVIDIA press-release, implying that hardware (essentially GPUs) is the major hurdle to full-automation. This is dubious – software (which essentially boils down to theories of computer vision) is as much a hurdle. Open questions in computer vision (questions critical to say the task of safely operating a vehicle) go back to the 60s at least. There’s a recurring theme in CV where problems predicted to be solvable in the next 5 years linger for decades without resolution.

    Fully-automated vehicles are more likely to be seen in aviation and agriculture than in commercial vehicles for reasons people like Herman Herman have discussed. Which begs the question, why are the Musks and Kalanicks and investors of companies like Uber so sold on the immanent arrival of fully automated self-driving cars?

    1. redkin,

      (1) “Strafor really falls for the self-driving hype here”

      Stratfor’s analysis is, like most professional forecasters, usually conservative. That’s why they have stayed in business so long. As it is here. You should re-read this report, and the links more carefully, since your characterization is quite false.

      (2) “The director of the Carnegie-Mellon University Robotics Institute”

      Beware what Andy Rivkin, the NY Times science reporter, calls the “single paper syndrome.” What one expert expert says is seldom definitive. I suggest paying attention instead to the consensus, which is why so many major corporations are investing large sums in driver automation. Also note that computer technology has a long history of experts underestimating what is possible.

      (3) “Others have cautioned not to expect full automation in our lifetimes”

      Since Stratfor says nothing remotely like that, your rebuttal seems quite misplaced. I suggest replying to direct quotes to avoid strawman rebuttals.

  2. Good stuff.

    Level 2 has been available for a couple years now as an option in most luxury cars. It’s nice. You can get elaborate foods at drive thru and eat guilt free without touching the steering wheel. You can nap peacefully while the driver next to you fumbles with his phone and laptop looking over notes on the way to a meeting…

    When you’re on a well maintained road, that its. Sometimes the lines on the road aren’t there. Bzzzz – wake up, driver. Sometimes it gets confused. Bzzzz. Level 3, not yet.

    US interstates are built with generous clearances and are usually decently maintained. Try it on the cramped pothole-ridden patchwork of NY roads, for instance. Or a windy county road… Road construction standards are going to be bumped up to accommodate unmanned vehicles. More land use, and a hidden cost. Once you can build a road without pedestrian crossings, it’ll work. You can always require all vehicles to have transponders.

    But my real beef with the hype surrounding unmanned vehicles: They don’t solve the problems we associate with transport. Resource use, pollution, congestion, lost time, cost/accessibility – all better addressed by increasing vehicle occupancy. Vs self-driving cars, which are looking like an increase in number of cars on the road…

    I think I might’ve said this here before, but … Mass transit. Buses, trains, park-and-rides – not new. Not even close to new. People-transport is not a technology problem.

    1. Pete,

      “they don’t solve the problems we associate with transport.”

      Or ingrown toenails or death or taxes. But 37 thousand people died in traffic accidents in 2016, and uncounted thousands more were injured. Self-driving cars have the potential to drastically reduced that. Plus they will make driver not just safer but easier. Many people find that having the car parallel park is a godsend.

      You ask too much from a single technology. There are no deus ex machina.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.