Search SciPol

Brought to you by
December 8, 2016

Driverless cars for travelers: More questions than answers

  • Agency
  • Robotics/AI

USA Today – In a series of initiatives that took place so quickly many Americans didn’t notice it happening, autonomous vehicles have suddenly lurched from science fiction to reality. And even at such a nascent stage, self-driving cars are poised to soon provide more choices to travelers opting for renting a car or hailing a taxi.

Ordinarily, this column strives to provide answers to travel issues and problems. But in the case of self-driving cars, we’re all still developing the questions, and it’s a list that seems to continuously expand.

For some, automated highways are the fulfillment of a dream first articulated at the New York World’s Fair in 1939-40. But the automotive industry has long lagged behind the aviation industry in this regard. The first autopilot system for airplanes was developed by aeronautical pioneer Lawrence Sperry in 1912, more than a century ago. In an infamous event that took place exactly 100 years ago — in November 1916 — Sperry employed that device while pioneering another aviation first, the Mile High Club. But regardless of his intentions, there’s little doubt of the success and ubiquity of the aircraft autopilot systems that have evolved since then.

Among most airlines worldwide, airplanes that carry more than 20 passengers are equipped with autopilots, so they are an intrinsic part of training for crewmembers and mechanics. But if many of the prognosticators are correct, automobile self-driving systems also could become commonplace within the next decade.

Even so, comparisons between autopiloting systems on airplanes and automobiles don’t always hold up. In fact, when it comes to autonomous cars, even the nomenclature itself has come under fire by critics. As Chris Woodyard pointed out here in October, transportation regulators in Germany and Holland have called on Tesla to stop using the name “Autopilot” for its partial self-driving system because it’s misleading. The concern is that drivers can place too much reliance on a system that still requires them to “stay in charge.”

In one way, the timing for a “partial” driving system is not good, since distracted driving has reached a crisis stage. According to, the U.S. Government’s official website on this topic, 431,000 people were injured and 3,179 were killed in 2014 due to this phenomenon. Now we need to imagine a scenario where partial distraction is actually permitted.

How will such training be implemented for a much broader audience of non-professionals? Who will oversee it? Will there be recurrent training requirements as there are for commercial and military pilots and mechanics? What methodologies and/or technological tools will be employed to ensure drivers don’t become too complacent and/or distracted?

Questions about safety

Federal authorities are still investigating a fatal accident in Florida last May involving a Tesla Model S equipped with Autopilot; the driver died when the system was in use but failed to detect a tractor trailer that turned in front of the car. A second crash of a Tesla Model S with Autopilot engaged occurred in China in August but did not result in any fatalities. Other crashes alleged to have been caused by Autopilot are under dispute by Tesla.

Some experts claim the first self-driving accidents were inevitable, and of course these incidents pale beside the overall toll of about 96 vehicle-related deaths per day in the United States. But coincidentally or not, the government’s response was recently put forth.

In September, the U.S. Department of Transportation revealed its Federal Automated Vehicles Policy, which is built on a 15-point checklist for new technologies. The “proactive safety approach” to Highly Automated Vehicles (HAVs) will be overseen by the National Highway Traffic Safety Administration, a subsidiary of the DOT, and will include these key components:

  • Development of a safety assessment for safe design, development, testing and deployment
  • Distinctions for federal and state regulators
  • Interpretation of current regulatory tools while identifying new regulatory tools

As reported here, even President Obama weighed in on the issue by noting in an op-ed that 94% of the 35,200 deaths on American roads last year were the result of human error or choice. It’s this very point that the developers of self-driving cars hammer home: The overwhelming cause of automotive deaths is human drivers, and they are seeking to largely eliminate that factor.

Obama also asserted the HAV policies that will be inherited by a new administration have been “designed to evolve.” And he stated: “Regulation can go too far. Government sometimes gets it wrong when it comes to rapidly changing technologies.”

How specifically can regulation go too far when it comes to safety? And how will the incoming Trump administration adopt and/or modify these new guidelines? How will the specific burdens of overseeing safety be split among federal and state regulators? How will increasingly complex transportation systems safely integrate traditional vehicles, driverless vehicles, bicycles and pedestrians, particularly in congested urban areas?

Questions about new initiatives

In September, the co-founder of Lyft issued a lengthy online manifesto entitled “The Third Transportation Revolution: Lyft’s Vision for the Next Ten Years and Beyond.” John Zimmer’s bold prognostication: “Autonomous vehicle fleets will quickly become widespread and will account for the majority of Lyft rides within five years.”

For those who haven’t followed such developments, it may be surprising to know that when it comes to developing new autonomous technologies, the floodgates have opened, and those in the forefront include a mix of familiar corporate names as well as new entrants. What’s striking is that unlike the development of mass-produced automobiles early in the 20th century, many of these entrepreneurs clearly are looking at the self-driven vehicle market not just for individual drivers but as a fundamental component of the travel industry.

This applies both to the sharing economy taxi segment, as well as to the car rental sector. Earlier this year, Avis Budget CEO Larry De Shon stated: “The combination of connected cars and self-service rentals and ultimately autonomous vehicles opens up tangible opportunities to enhance customer service, lower operating costs, improve fleet utilization and more.”

Consider just some of these burgeoning self-driving initiatives, many of which are coupled with vehicles that do not run on gasoline:

  • Ford. In August, the automaker announced a massive investment in this technology, doubling the size of its Silicon Valley team and tripling its fleet of self-driving Fusion Hybrids, which are being tested on roads in Arizona, California and Michigan. The goal? To deliver high-volume, fully autonomous vehicles for ride-sharing in 2021.
  • General Motors. In March, General Motors acquired Cruise Automation for a reported $1 billion in an effort to “accelerate” development of its driverless technology. In addition to establishing a separate corporate unit for autonomous vehicles, GM also entered into a ride-sharing alliance with Lyft and formed Maven, a “personal mobility brand” for car-sharing fleets.
  • Google. The company first began testing a self-driven Toyota Prius on California freeways in 2009. Today Google has logged over 2 million driverless miles by expanding onto roads in Arizona, Texas and Washington.
  • NuTonomy. Just before Thanksgiving, this Cambridge, Mass.-based company announced a memorandum of understanding to test self-driving Renault Zoe electric vehicles on designated streets in Boston. NuTonomy previously began public road tests and public trials in Singapore, with the goal of launching a self-driving mobility-on-command service there.
  • Uber. In September, Uber launched a pilot program of self-driving cars in Pittsburgh (along with “a safety driver up front to make sure the ride goes smoothly”). Uber also has partnered with Volvo, and predicts that self-driving Ubers will soon be on the road 24 hours a day.
  • Other endeavors. You may not know or, but some investors are betting you will soon; the first dubs itself “ghostriding for the masses” and the second proclaims “artificial intelligence for self-driving cars.” Both state they’re currently hiring.

Which of these projects will succeed and which will not? How will older technologies adopt — or not?

More questions than answers…

Of course, from a consumer perspective there are many more questions that need addressing.

  • What effects will such vehicles have on insurance? And who will bear responsibility? The “driver” who may not be driving? The vehicle’s owner? The ride-sharing company or rental company? The manufacturer?
  • If many of these new vehicles are used as part of the sharing economy, how will taxation be assessed? Will it be sufficient to offset required investments on the part of federal, state and local authorities?
  • How will the travel industry evolve to accommodate such wholesale changes to its existing business models? How will travelers book such products? What concerns will arise over pricing transparency and consumer protections?

With such a vast array of unanswered questions, it’s worth noting that only one issue seems all but resolved. Autonomous vehicles will be impacting more travelers’ lives sooner rather than later.