The innovation of autonomous vehicles could have a significant impact in terms of the number of annual automotive collisions in Canada and in the United States, which total 120,0003 and 10,000,000,4 respectively. In fact, according to Peter Sweatman, Director of the University of Michigan Transportation Research Institute5, this number could potentially be reduced by 90-percent, since nearly all motor vehicle collisions are caused by a form of human error, which would be minimized or wiped out with autonomous vehicles.6
The final manifestation of autonomous vehicles will largely depend on the manner in which regulators balance the issues that arise at the intersection of liability, freedom, and privacy. This paper will provide insights into the current state of the technology of autonomous vehicles and autonomous trucks before delving into a discussion about the shifting scope of liability and the potential consequences this may have on the calculus insurance companies use to apportion risk and determine the cost of premiums.
Current State of the Technology
I. Autonomous Vehicles
Although, in recent years, the advancements of autonomous vehicles has been exponential, the attempts to have self-driving cars connected to smart roads and smart highways has been in the works for more than half a century. At the 1939 New York World`s Fair, visitors were presented with the idea that automated cars and automated highways were indeed a possibility; however, the technological infrastructure, namely GPS positioning sensors connected to high-speed broadband, to spearhead such a project did not yet exist.7 In the mid 2000s, the Defense Advanced Research Agency (DARPA) organized the Grand Challenges where research teams gathered to compete to create the first set of self-driving vehicles.8
In 2009, Google gathered and recruited a team of engineers already working on creating self-driving vehicles to begin its own project at the company. This team has made developments in leaps and bounds, and are well on their way to reaching their goal of creating the first set of fully functioning autonomous vehicles by 2020.9 As Paul Godsmark, chief technology officer at Canadian Automated Vehicles Centre of Excellence points out: "...we`ll achieve as much in the next 8 years as we have in the last 40 ...although it sounds fantastical to say, this technology will be ready by 2023, [however] the technology exponential development curve suggests that this could be an underestimate".10
Google's autonomous vehicles have driven over a million miles on public roads... |
As of 2015, Google's autonomous vehicles have driven over a million miles on public roads in Mountain View, California, and in Austin, Texas.11 These autonomous vehicles comprise of a modified Lexus RX350 SUV, and a Google Car created from the ground up by the company.12 For all of its travels, a safety driver is on board of Google's autonomous vehicles for two reasons: (1) to commandeer the vehicle if the technology on board were to fail in any way, and (2) to interact with the public and elicit feedback on the idea of self-driving vehicles.13
The next logical extension of creating autonomous vehicles is to create an autonomous taxi service. Over the last year, this is precisely what Uber has committed itself to manufacturing.14 When the company isn't busy disrupting the taxi-industry by introducing software that can summon a taxi through a Smartphone, Uber is actively working with Carnegie Mellon University in order to make driverless taxis a reality.15
As Adam Jonas, an industry analysts for Morgan Stanley puts it: "...the journey to the end of humans driving cars may take a generation to happen and it will hit obstacles, but the process has been started...computers drive better than people, and once a fleet of connected cars is available, people's use of automobiles will change, ultimately obviating the need for the vast majority of individual vehicle ownership."16
The trucking industry has been eagerly anticipating and actively preparing for the introduction of autonomous transport vehicles. |
II. Autonomous Trucking
The trucking industry has been eagerly anticipating and actively preparing for the introduction of autonomous transport vehicles. In fact, this past May, Daimler acquired the world's first autonomous trucking license to drive on public roads, and routinely conducts tests on public roads in Nevada, United States.17 It currently has two Freightliner trucks that are equipped with a system meant to drive autonomously on highways, and expects these sorts of self-driving trucks to be commercially available as early as 2025.18 Not to be outdone, Suncor Energy Inc. is aiming to replace its fleet located in Alberta's oil sands with autonomous trucks "by the end of the decade".19
According to the Canadian Automated Vehicles Centre of Excellence, with the introduction of fully automated trucks, by 2030, we could be in a situation where there will no longer be a need for truck drivers. In 2013, Suncor began shifting its operation towards autonomous heavy-haul trucks, and recently, signed a five-year agreement with Japan's Komatsu Ltd. to purchase "autonomous-capable" trucks that will be operated autonomously or with the assistance of a single individual.20
A representative from Suncor indicated that the technology is being tested and will be gradually introduced into the daily operations of the corporations.21 Not all transport companies are taking this relatively gradual approach, Rio Tinto, which operates in Australia has at least 54 autonomous trucks currently operating handling various transportation-related tasks.22
From a business perspective, the advantages are irresistible, according to the Automated Vehicles Centre of Excellence, eliminating the need for a human driver could cut operating costs by 40 percent.23 Moreover, taking into account for maintenance, loading and unloading times, autonomous trucks could operate for 20 hours per day, whereas, regulations currently do not allow for a human driver to operate a truck for more than 14 hours per day.24
In addition to this, a collateral benefit redeemable by cargo companies and insurance companies is the elimination of theft. By necessity, an autonomous transport vehicle will have constant digital communication with its host, allowing for real time monitoring of location and speed, which results in sending out alerts explaining why the autonomous truck deviated from the original route, for any reason.
...Canada has yet to enact comprehensive regulations that reflect the current state of autonomous vehicles. |
Regulations
III. Placing the Cart Before the Horse
Unlike our neighbours to the south, Canada has yet to enact comprehensive regulations that reflect the current state of autonomous vehicles. In July 2015, the Ontario Good Roads Association published a white paper flagging issues that Canada must tackle before the introduction of autonomous vehicles. Amongst the recommendations was the need to update or amend the Ontario Traffic Manual, Highway Traffic Act,25 and Ontario Provincial Standards.26
Unlike Canada, to date, according to the Centre of Internet Society at Stanford Law School, the District of Columbia, and four states (Missouri, Nevada, California, and Florida) have enacted regulations accounting for autonomous vehicles, while another 18 have regulations in front of their legislative assemblies. The enacted regulations include but are not limited to: defining what constitutes as an autonomous vehicle,27 ensuring manufacturers disclose the information collected by the autonomous vehicle to the driver,28 and limiting manufacturer liability if the end-user modifies or augments the technological equipment found in the autonomous vehicle.29 The importance of enacting effective and comprehensive regulations and standards cannot be understated. In the initial stages of deployment of all autonomous vehicles, we will be faced with a situation in which conventional vehicles operated by individuals will be on the same public streets as autonomous vehicles, and we must have regulations and standards that take this to account.
IV. Shifting Liability - Who is Liable?
As one can imagine, the introduction of autonomous vehicles poses a serious question as to which party will be held liable when a collision occurs. Some have suggested that the list of potential parties that could be held liable could include: "the vehicle manufacturer, the manufacturer of a component used in the autonomous system, the software engineer who programmed the code for the autonomous operation of the vehicle, and the road designer in the case of an intelligent road system that helps control the vehicle".30
In addition to this list of potential parties that could be held liable, one could also add the municipality or township in which the accident occurred. A municipality could potentially be held liable if the road or street signs were not properly maintained, which in turn, could have enabled the autonomous vehicle to incorrectly misprocess the signage contributing to a risk of a collision. This is precisely the problem Tesla Motors encountered when attempting to test its own autonomous vehicles on California's 405 Highway. The company found that the concrete on the highway was light in color, the lane markings were faded, there were high-contrast black skid marks, and the width of the lane width varied significantly.31 Given the fact that these markers were what the autonomous vehicles used as guideposts to navigate roadways, it is conceivable to hold the municipality partly liable for any negligence arising out of poorly maintaining the roadway.
...the risk of a hacker gaining access to the autonomous vehicle's operating system and causing an accident from a remote location. |
An additional liability concern arising out of autonomous vehicles is the risk of a hacker gaining access to the autonomous vehicle's operating system and causing an accident from a remote location. Although this problem also exists with conventional vehicles with high-tech components such as Tesla's Model S,32 and the Jeep Grand Cherokee,33 the problem is vastly enhanced when the vehicle is completely autonomous and relies entirely on a computerized network to operate.
This raises significant questions as to the extent of duty the manufacturer of the vehicle, the manufacturer of the software, and the insurance company owe to the driver of the vehicle. Do the various manufacturers have a duty to continuously monitor and create software patches to every vehicle that they sell, no matter the geographic location of use? Does the insurance company have a duty to investigate and indemnify a driver if he or she claims that they have been the victim of hacking and were not engaged in reckless driving? Similar to regular maintenance of a motor vehicle, should drivers undergo regular software maintenance to ensure that they are protected as best as the technology at the time allows?
V. The Changing Nature of Insurance Premiums
As mentioned, with the creation of autonomous vehicles, regulators will need to balance the issues that arise at the intersection of liability, freedom, and privacy. The wealth of information that can be gathered at this intersection presents an opportunity for insurance companies to craft insurance premiums that better reflect the risk the company is taking in insuring an individual driver. In response to this, insurance companies may choose to create a spectrum of available policies. The amount of information an individual is willingly to provide their insurance will determine where they fall on the spectrum, and thus the amount and the particulars of their individual policy.
On the far end of the spectrum will be individuals who are not willing to provide analytic information to their insurance company. This could involve individuals who are not willing to purchase an autonomous vehicle, but instead, insist on driving a conventional vehicle. This could be for a number of reasons: this individual could be a traditional car enthusiast, or even an individual who values their freedom and privacy and detests the idea that the insurance company has detailed information in the manner in which they operate their vehicle. This leaves the insurance company in a situation where they are not able to access crucial analytic information via sensors on the vehicle that could immediately inform the insurance company of the accident and provide data regarding the speed, weather, and driving pattern of the driver in the critical seconds before the accident. The lack of access to the information will undoubtedly hamper the insurance company`s ability to apportion liability and defend itself from a civil action, thus justifying a higher premium.
On the other end of the spectrum will be individuals who choose to purchase an autonomous vehicle and provide analytic information to their insurance company. This individual has ceded a portion of their control both in terms of the manner in which the vehicle is driven, but also, in the manner in which the insurance company collects and retains information about their vehicle. If the claim that autonomous vehicles are significantly safer than conventional vehicles holds true, and individual drivers grant access to information that will allow their insurance company to apportion liability appropriately and better defend itself from a civil action, this driver will undoubtedly be a safer bet, thus justifying a lower premium.
In between these two extremes will be drivers that choose to purchase and operate autonomous or semi-autonomous vehicles where individuals retain control over the manner in which the vehicle is driven and emits information. For instance, an individual could choose to operate a vehicle that collects information such as distance driven, acceleration and braking, time of day of journey, and the circumstances arising in the final seconds before an accident, but the information would not be entirely released without the consent of the driver. This individual could be a mix of the traditional car enthusiast who appreciates technology add-ons to their conventional car but chooses not to relay this information to their insurance company. This individual potentially poses a greater risk than the individual who operates a completely autonomous vehicle but less so than one who chooses to drive a conventional vehicle.
As opposed to relying on information based on demographics steeped in age and driving experience, in a reality of autonomous vehicles, insurance companies could create a spectrum taking into consideration the decreased risk of collisions with autonomous vehicles and the ability to access analytic information necessary to appropriately apportion liability. The value an individual places on the freedom to manually operate a vehicle and their concerns of providing this analytic information to their insurance company will determine where they fall on the spectrum.
Conclusion
There is a significant chance that in the next decade or so, autonomous motor vehicle will be a reality. This poses significant questions as to the proper scope of liability and the particular actors that can and should be held liable. However, this development in technology also poses significant opportunities such as reduced motor vehicle accidents, as well as a wealth of information to better assess risk and apportion liability, on the part of the insurance companies.
1 Self-Driving Cars: The Next Revolution. Centre for Automotive Research.
2 Gabriel Weiner & Bryant Walker Smith, "Automated Driving: Legislative and Regulatory Action" (2014) THE CTR. FOR INTERNET AND SOC.
3 Transport Canada. Canadian Motor Vehicle Traffic: Collision Statistics 2013.
4 U.S. CENSUS BUREAU, STATISTICAL ABSTRACT OF THE UNITED STATES: 2012, at 693 tbl. 1103 (2012).
5 The University of Michigan Transportation Research Institute has created a fake city called Mcity to test autonomous vehicles for Ford, General Motors, and Fiat`s Chrysler. This fake city is a physical simulation of a dense, complex urban environment. It has many real situations packed into a 32-acre area - city blocks, suburban streets, suburban arterials, rural roads, freeways and ramps, roundabouts, traffic circles and complex skewed intersections. The University of Michigan Transportation Research Institute provides a video simulation of the fake city in action on their YouTube channel: .
6 How self-driving cars will cut accidents 90 percent.
7 Cohen, Roy Alan. "Self-Driving Technology and Autonomous Vehicles: A Whole New World for Potential Product Liability Discussion." (2015) Defense Counsel Journal 82, no. 3.
8 Rupp, J. and King, A., "Autonomous Driving - A Practical Roadmap," (2010) SAE Technical Paper.
9 Google, Google Self-Driving Car Project.
10 Tavia Grant, "Driverless Trucks Could Mean 'Game Over' for Thousands of Jobs", The Globe & Mail (July 26, 2015)
11 Google, supra at note 9.
12 Ibid.
13 Ibid.
14 Samuel Gibbs, "Uber`s first self-driving car spotted in Pittsburgh", The Guardian.
15 Ibid.
16 Nigel Andretti, "Disruptive Uber Linked to Self-Drive Car Programme", MotorMe (May 22, 2015) .
17 Daimler, "World Premiere on U.S. highway: Daimler Trucks Drives First Autonomous Truck on Public Roads" Daimler (May 05, 2015) .
18 Ibid.
19 Globe & Mail, supra at note 10..
20 Ibid.
21 Ibid.
22 Ibid.
23 Ibid.
24 Ibid.
25 Highway Traffic Act, R.S.O. 1990, c. H.8
26 Fahad Shuja, "The Roadmap for Autonomous (Self-Driving) Vehicles in Ontario, Canada" Ontario Good Roads Association (July, 2015)
27 District of Columbia B19-0931: Defines "autonomous vehicle" as "a vehicle capable of navigating District roadways and interpreting traffic-control devices without a driver actively operating any of the vehicle`s control systems," requires a human driver "prepared to take control of the autonomous vehicle at any moment," restricts conversion to recent vehicles, and addresses liability of the original manufacturer of a converted vehicle. Final version removed previous provisions requiring autonomous vehicles to operate on alternative fuels and imposing a vehicle-miles-traveled tax in lieu of DC motor fuel tax. Passed Congressional review (April 2013).
28 California SB-1298: As amended, defines "autonomous technology," "autonomous vehicle," and "operator"; finds that the state "presently does not prohibit or specifically regulate the operation of autonomous vehicles"; requires rulemaking before 2015; permits current operation under certain conditions; imposes additional oversight on the operation of vehicles without a human in the driver's seat; and requires that the "manufacturer of the autonomous technology installed on a vehicle shall provide a written disclosure to the purchaser of an autonomous vehicle that describes what information is collected by the autonomous technology equipped on the vehicle."
29 Michigan SB 0663: declares that the original manufacturer of a vehicle is not liable for damages resulting from another person's conversion or attempted conversion of the vehicle into an automated motor vehicle, or the modification of installed equipment, unless the defect from which the damages resulted was present in the vehicle when it was manufactured. Similarly addresses liability of subcomponent system producers for equipment installed by those producers to convert vehicles into automated motor vehicles.
30 Merchant, Grary and Lindor, Rachel "The Coming Collision Between Autonomous Vehicles and the Liability System" (2012) 52 Santa Clara L. Rev. 1321.
31 Alex Davis, "A Crummy Highway is Delaying your Self-Driving Tesla" (July 17, 2015) Wired.
32 Thomas Reuters, "Tesla Model S Car Hacked, Shut Off While Driving", CBC (August 6, 2015).
33 Mark Clothier et al., "Fiat Chrysler Offers Patch After Hackers Commandeer Jeep", Bloomberg News (July 22, 2015).
