Self-driving cars have had a storied history in film and television, stretching as far back as the early 1980s. From Knight Rider and Timecop to Minority Report and I, Robot, the autonomous vehicle has long established itself as a common trope in the science fiction genre, and continues to capture our imagination today. However, in the past decade, this particular fantasy is rapidly becoming a tangible reality.
In 2011, Nevada became the first of several states to pass laws on the development and operation of Automated Vehicles (AV). Nearly all major car manufacturers have announced a vested interest in the burgeoning research and development of AV technology, and anticipate releasing their own AV within the next 10 to 15 years. Tech giant Google has already begun test-driving its prototype in the San Francisco Bay Area, and has successfully clocked in nearly 2 million miles of autonomous driving. Even though the California DMV has ruled that practice to be illegal, this is seen as more of an attempt to get regulatory control rather than an effort to quash a new enterprise.
In partnership with Volvo, ride-share company Uber recently began testing their fleet of human-supervised, self-driving cars to a select number of participants in Pittsburgh, Pennsylvania. In August 2016, Singapore became the first country to offer self-driving taxis. It seems that the future we have collectively envisioned has already arrived, and with it comes a whole host of questions about safety, function, and liability.
To start, it is worth noting that the car you drive today may have some of the seeds of the technology upon which these vehicles are built. Many vehicles on the road today have technology which grants extra layers of safety to human operators by providing automated or machine-actuated assistance. This includes things as common as power steering and electronic stability control, to newer technologies such as adaptive cruise control, automatic collision avoidance systems, and lane departure warning systems. Those technologies are a far cry from truly self-guided automobiles, but they lay the foundation for that next evolution. Robotic cars will rely on some of those same cameras, radar systems, and computing models to assist in the independent operation of the AV.
Given that vehicles available today such as the Tesla Model S seem to differ from the test-driven prototypes that have yet to be released, there has been some confusion as to where this technology is headed. A call for clarity arose following news of the first self-driving fatality in June 2016, prompting many to question the presumed safety of autonomous driving and whether the accident could have been averted if the Tesla vehicle had been equipped with the same radar technology as the seemingly safer Google prototype.
In September of 2016, in its Federal Automated Vehicles Policy, the Department of Transportation’s National Highway Traffic Safety Administration adopted the following rubric of determining the functionality of automation technology:
Level 0: The human operator performs all driving functions.
Level 1: The vehicle’s automated system can occasionally assist the operator with some driving functions. An example of this would be adaptive cruise control or parking assistance.
Level 2: The vehicle’s automated system can conduct some driving functions as the operator monitors the environment and performs the rest of the driving task. An example of this would be highway driving assistance.
Level 3: The vehicle’s automated system can conduct some driving functions as well as monitor the environment in some instances under very specific conditions. Some of the more recent advancements in automation are Level 3, and include lane-changing, acceleration, and deceleration.
Level 4: The vehicle’s automated system can perform both driving functions as well as monitor the driving environment without oversight or control from the operator. However, it can only operate in specific environments and conditions. Currently, there are no Level 4 vehicles available to the public.
Level 5: The automated system can perform all driving tasks under all conditions that a human operator can. This level removes the agency of the human driver entirely.
Using this classification, Tesla’s Model S, a vehicle with an Autopilot function, would be categorized as a Level 3 vehicle due to the unique situations in which self-driving can actually take place. Conversely, Google’s prototype constitutes a Level 4 vehicle, in which the human driver serves as more of a passenger than an active operator. Level 4 and 5 are more aligned with the idea of self-driving cars that we’ve seen in our movies and TV shows, and these are still relatively far in the future.
However, several consulting firms and organizations have already started polling the public as to whether or not they would ride in a self-driving car. The Insurance Information Institute conducted a survey in May of 2016 and found that over 55% of respondents would not ride in a self-driving vehicle. Other industries that would be affected have voiced their concerns regarding the mass-production and proliferation of AV tech, such as the trucking industry, a widely popular profession in many states.
Generally speaking, the public is often skeptical, even fearful of new technology until it has been fully vetted by the government, the media, and the experience of first adopters. Such is the case with air bags, which was originally patented in the 1950’s, but wasn’t widely implemented until 1990. The emergence of that particular technology came with mixed expectations and reactions.
Several industry experts predicted that air bags would ultimately replace seat belts as the primary safety measure in vehicles; others, including car manufacturers, fervently opposed adopting air bags, unsure as to whether consumers would even opt for a safety feature that has proven fatal to women, children, and the elderly. No one at the time suspected that air bags would serve as a supplement to the seat belt. If viewed as an analog to AV technology, perhaps it would be best if the public scales back its expectations to allay some of its fears.
But, how will AVs affect driving laws and auto liability? Many presume that automated driving will drastically reduce the number of car-related accidents by eliminating the main element responsible: human error. This is not entirely without merit, in spite of the Tesla self-driving accident. The Insurance Institute of Highway Safety believes the wide implementation of electronic stability control, a form of AV tech required in all consumer vehicles since 2012, is a major contributor to the significant reduction in car-related deaths over the years.
If this speculation proves true, in the event of an accident, will the blame shift from the operator to the car manufacturer? Thus far, it appears that the liability will follow the operator—that is the person sitting in the vehicle. But what if there is an accident with a car with no occupant that has been summoned to pick someone up, or to make a delivery?
Some have speculated that the advent of AVs will also give rise to “real-time insurance premiums” through additional technological advances such as biometrics, in which a vehicle can discern the operator’s blood alcohol content level upon activation. In this hypothetical scenario, the vehicle will warn the driver and require verbal confirmation to proceed, in which case the driver would be considered at-fault should an accident occur. Others believe that auto-coverage for AVs will be grouped into homeowner’s insurance in the same manner as bicycle crashes. Another speculation that has been gaining recent traction is the use of telematic devices, or “black boxes,” in order to monitor the driver. The National Association of Insurance Commissioners has predicted a significant growth in telematic usage for underwriting, which several usage-based insurance programs have already begun incorporating.
More than a few have begun to question how cyber liability will come into play given the potential hazards of hacking. Recently, a team of Chinese security researchers hacked the Tesla Model S, effectively remote-controlling the car from 12 miles away via its controller area network. The team posted its findings on YouTube to underscore the vehicle’s vulnerabilities to the public, and advised all Tesla S-owners to be diligent in updating their firmware. This type of automobile maintenance is already becoming increasingly common as the majority of automotive recalls now include the installation of new software, not hardware, and given the complexity of these new systems, many foresee these services becoming quite costly to the consumer.
The liability implications of AVs are as far-reaching as its application, and while there is no real way of telling how much will dynamically change with the emergence of this technology, there is one thing we can count on. Historically speaking, insurance companies have often served as intermediaries in the investigation, recognition, and adoption of technologies that reduce risk and increase safety. When AVs become available to the rest of us, you and I will be the first to know if it’s a good idea to sit back and let our cars drive us to work.
The Application of Innovation
We racked our brains to think of how AVs might be useful to special districts. We had a lot of great ideas, and some of them strayed out of AV and into robots in general, like robot lawn mowers, pesticide sprayers, and watering trucks.
But here are a few that qualified in the category of AV. If you think of any, we would love to hear your ideas! You can contact us at info@csdpool.org.
- Cars that drive automatically on the most efficient route possible, even avoiding traffic and road work to ferry meter readers, park attendants or others
- Automated carpooling vehicles offered as a benefit to employees, especially during inclement weather
- Cars that ferry workers to off site meetings or trainings, then picks them up at a predetermined time and location
- Automated golf carts
- Fire engines, pumps, or tankers
- Buses or other personnel carriers
- On-call vehicles that deliver additional equipment to field crews as needed for fire fighting, water line maintenance, etc.
- Self-driving ambulances that free up EMTs to attend to critical patients
- Automated library vehicles that pick up inter-library loan materials or redistribute assets between multiple locations
Sources:
- http://www.nytimes.com/2016/12/14/technology/uber-self-driving-car-san-francisco.html?_r=0
- http://triblive.com/news/allegheny/11383487-74/driving-self-uber
- http://www.wsj.com/articles/worlds-first-self-driving-taxis-hit-the-road-in-singapore-1472102747
- http://www.nytimes.com/interactive/2016/07/01/business/inside-tesla-accident.html?_r=0
- http://www.nhtsa.gov/nhtsa/av/pdf/Federal_Automated_Vehicles_Policy.pdf
- http://www.iii.org/issue-update/self-driving-cars-and-insurance
- http://www.latimes.com/projects/la-fi-automated-trucks-labor-20160924/
- http://www.rand.org/pubs/research_reports/RR443-2.html
- http://www.iihs.org/iihs/topics/t/automation-and-crash-avoidance/topicoverview
- http://www.forbes.com/sites/patricklin/2016/04/25/self-driving-cars-wont-kill-insurance-industry/#10437f936afa
- http://www.fordhamiplj.org/2016/09/16/self-driving-cars-liability-implications/
- http://www.naic.org/cipr_topics/topic_usage_based_insurance.htm
- https://www.theguardian.com/technology/2016/sep/20/tesla-model-s-chinese-hack-remote-control-brakes