How society and the law will need to adapt to autonomous vehicles
Epochal changes in the movement of people and goods over the millennia — like the creation of the wheel or the invention of the steam engine — have resulted in dramatic societal changes. Most recently, the diesel engine gave us cheap and ubiquitous travel options and connected our communities in ways no-one initially comprehended.
The next revolutionary technology in mobility will be connected and Highly Automated Vehicles (HAVs). While the technologists are making daily strides, the policy and law makers need to keep pace. In the development of autonomous vehicles we need to consider the unintended consequences and the likely impacts on our societies.
As pilot tests are rolled out, concerns have been expressed about how the law may, or may not, be an impediment to the effective and efficient integration and rollout of these vehicles and technologies to public highways around the world.
And as lawmakers, we don’t want to stand in the way of progress. While considering how to legislate for anticipated change we have a duty to adopt a ‘goldilocks’ type approach where we put in place ‘just-right’ standards that are robust enough to placate fears about HAVs, while still encouraging innovation.
We need to set up the licensing and insurance structures for these vehicles and ensure that basic safety, security and liability laws protect citizens from unintended consequences. The complexity around these issues is further compounded by the fact that there are least five different levels of automation:
At levels 0–2 the human driver monitors driving and the environment but at levels 3 through 5 the automated driving system monitors the driving environment. So lawmakers are having to consider a blended mix of technologies that may (or may not) occupy our road network alongside the current non-automated driving fleet.
This is no simple challenge, but we’re seeing various approaches around the globe of lawmakers trying to get to grips with these complexities, putting in place standards on many legal fronts that allow for pilot tests to begin.
One such example is the UKs Vehicle Technology and Aviation Bill 2016–2017 introduced on February 22, 2017.
This bill outlines the liability of insurers for automated vehicles including contributory negligence standards, rights of recovery for unauthorised alternations or the failure to update software, and the rights of the insurer to claim against a person responsible for an accident.
This is the kind of work that needs to be done to provide a groundwork and some certainty to equipment manufacturers, software developers and the public at large before they purchase these vehicles.
It also starts to put a framework around aftermarket software additions that provide connectivity or further automation to older non-HAV type vehicles.
As ever, there are a number of different regulatory frameworks in operation. There are three major classes of regulatory framework being put into place around the world currently:
1. A prescriptive type approach (adopted in California, Finland and the Netherlands) where laws are created with specific rules around the deployment of these vehicles, vehicle licensing, driver licensing, pilot testing permit authorization, and minimum liability insurance amounts required to be held.
2. A policy type approach (France, Florida, Michigan), that puts in place a framework but no prescriptive legislation. By adopting this approach you exert less control over manufacturers wishing to test vehicles on public roads and do not put in place any geographical restrictions.
3. Ad-hoc pilot testing approach approved without any formative legislation, ordinances or rules in place. An example of this approach can be seen in Uber’s pilot testing of automated taxis in Pittsburg Pennsylvania and the UK’s funding of four cities at £10 million each to test driverless cars
Each of these approaches creates its own set of challenges, and it is probable that as vehicle fleets evolve over time continued legislation will still be needed from central, state and local governments. Along the way we will encounter both anticipated and unexpected consequences and it’s likely that some legislation will be rendered obsolete.
A first out-of-the-gate issue that a jurisdiction should consider for HAVs is what does your current law forbid, curtail, or is silent upon?
For example, must the car have a steering wheel and pedals to be legal to drive on the road?
It may sound banal, but if your legislation requires this (and some states do have this requirement in their regulations), or an international treaty has some exclusionary type language then an automated vehicle that does not have these items would not be allowed to drive on the road.
The Vienna Convention on Road Traffic 1968 for example, was argued to have been a major reason why the European Union did not develop a more robust approach to HAVs.
Article 8 of that convention required that every driver shall at all times be able to control his vehicle. Member states of the EU along with the United States submitted an amendment to the United Nations in 2015 to allow drivers to take their hands off the wheel of self-driving cars, as well as proposing amendments to Article 8 and 39 of the convention which were also considered to hamper the advancement of new technologies aimed at improving safety.
These amendments were entered into force in March 2016, although the European Parliament noted that SAW J3016 levels 3 and above were still incompatible with the Vienna Convention.
How the law and automated vehicle technologies interact will be governed by factors we can anticipate, like liability, cyber security, hacking, and privacy. But there are subtler nuances that will begin to emerge as fleets evolve and acceptance of these technologies grows.
Examples could be:
• What or who determines the handover from automated vehicle to driver and vice versa? Will the vehicle rely on automation for traffic jams, highway driving, or parking assistance or will it occupy the entire driving space?
• What role will government be required to play in infrastructure provision beyond the traditional traffic signals, signs and striping? Will vehicle to infrastructure technologies be provided by the public sector?
• When are HAV crashes more likely to occur? Will it be at handover because of user error, after maintenance or vehicle updates, or because of hacking?
• How will developments affect legal approaches to discourage distracted driving? Will autonomous vehicles render these laws obsolete?
• What will determine probable cause to stop an individual in an HAV? How will law enforcement develop probable cause for an individual in a vehicle that may have no impairment that could be used to justify a stop?
• Driver licensing: who can be licensed and what endorsements may need to be amended?
• Can the visually impaired now drive? Will the current exemption requirement for a commercial truck driver with diabetes be necessary?
• Will platooning distances, for trucks and personal vehicles need to be amended? What is a safe distance, especially in a mixed fleet scenario with some HAVs and some traditional human driven vehicles? How will platooning for trucks square with the federal bridge formula?
Lawmakers need to anticipate these and a host of other scenarios in a world of HAVs. We cannot sit idly on the sidelines and wait for an occurrence to happen.
In a review of Texas Law and connected and automated vehicles myself and Wendy Wagner recommended the need for policymakers to consider whether the testing and deployment of connected and automated vehicles in the State would benefit from more formal, legal oversight.
Bryant Walker-Smith has also recommended a strategy checklist for government promotion of automated driving. This includes preparing governments, infrastructure, conducting a legal audit, and developing a point person in one agency with authority and credibility to coordinate amongst various local and state government agencies.
As connected and automated vehicles become part of our ‘new-normal’ for transportation a few issues are guaranteed to occur. Challenges in creating regulatory structures are predictable. Unintended consequence will occur, and laws will become obsolete. The law however, will evolve as it has over the millennia to the new challenges and technologies that we will utilize through legislation and in the common law jurisdictions through case law.
On March 12th, Lisa Loftus-Otway will be participating in a session at this year’s SxSW conference entitled ‘Assimilating self-driving cars into our society’