Self-driving vehicles need new road safety rules

US government’s new paper on regulation offers important insights that can serve as a guideline for countries

As the race to get self-driving vehicles on the road continues apace, the US Department of Transportation (DOT) released its new federal guidance for regulation ‘Preparing for the Future of Transportation: Automated Vehicles 3.0’ on October 4. Given the US is undoubtedly the leader in the development of autonomous vehicle technology, should other countries learn from the US approach to their regulation? AV 3.0 offers important insights both for borrowing from the US approach and departing from it. Here’s why.

AV 3.0 is predicated on six principles. The first, safety, is premised on the idea that autonomous vehicles have the potential to reduce road accidents. Undoubtedly, these vehicles will also generate new safety risks. The goal of regulation is to enable technology development by reducing risks.

The second principle, tech neutrality, commits to eschewing a heavy-handed approach and allowing competition and public adoption to determine winners. The regulatory approach is to be neutral about technologies and develop flexible policies that enable innovation.

The third principle, modernisation, commits DOT to repealing outdated rules that inhibit the development of self-driving vehicles. DOT will redefine terms such as “driver” and “operator” to go beyond the human and recognise that an automated system could be the driver or operator of a vehicle. Principle four, consistency, seeks to eliminate divergent state laws that could impose transaction costs by developing national standards.

The fifth principle, proactive preparation, commits DOT to helping investors plan for a “flexible automated future.” The sixth principle, “freedom of the open road,” a particularly American idea, seeks to protect the ability of all road users to continue to enjoy the autonomy of driving their own vehicles alongside autonomous vehicles.

While the principles offer clarity on the regulatory approach, AV 3.0’s major shortcoming is its high reliance on the industry to do the right thing. For instance, the DOT merely “encourages” the industry to publicise their Voluntary Safety Self-Assessments in a bid to promote transparency. We already know that encouragement does not work — just consider the tragic death of Elaine Herzberg in an accident with a self-driving Uber vehicle in Tempe, Arizona, earlier this year.

AV 3.0 also recognises the futility of limiting testing to the ten previous ‘Automated Vehicle Proving Grounds’ it had identified in January 2017. DOT

admits that many locations are competing to perfect the technology and it does not wish to pick winners or losers.

The DOT also affirms that it may develop standards for self-driving vehicles that permit designs bereft of the currently ubiquitous steering wheels, pedals, and mirrors. This is to be welcomed.

Crucially, the DOT also announced a study of the workforce impacts of AV — it will be vital to understand how commercial truck drivers and taxi drivers will need to adapt to stay relevant in the AV era. Coevally, the impact of widespread adoption of automated vehicles will not be felt just by those who depend on driving for their livelihoods. A wide array of complementary labour

— automobile mechanics, fleet maintenance staff, insurance sales workers and others will find their skills may not match future requirements. Equally, the market may demand cyber security workers, programers, and other kinds of technicians. It is not just important for these changes to be properly understood — they have to be communicated clearly to affected groups.

And relying on industry to find solutions may not work — many commercial truck drivers and taxi drivers are selfemployed. Others work for small companies that don’t have training budgets or professional development programmes. Moreover, many taxi or truck drivers may not have the interest or aptitude to become cyber security workers or programers.

There is another important reason for the government to step in: there is a disparate impact on extremely vulnerable groups. In 2016, 91 per cent of taxi drivers in New York city were immigrants — about one-in-two drivers were from the Dominican Republic whereas onein-four drivers originated in Bangladesh. In London, 14,685 out of a total of 21,355 Black Cab drivers identified as white in 2017. Amongst the 117,857 other taxi drivers, white drivers only made up about 20,000. In other words, for the less lucrative non-Black Cabs, Asians and Blacks accounted for the vast majority. And drivers mostly came from poor countries such as Bangladesh and Pakistan. Similar profiles apply in countries such as Australia and the UAE. The advent of ride-hailing apps has already imposed crippling costs on vulnerable drivers. Automated vehicles will have an even more drastic impact — potentially eliminating many roles altogether.

To be sure, AV 3.0 offers needed clarity in several areas. Current vehicle safety standards are, unsurprisingly, written for vehicles driven by human beings. The DOT’s clear vision to eliminate conflicting regulatory requirements imposed by the different states that have the potential to inhibit automated commercial motor vehicles by “pre-empting” such rules is a welcome step.

In the end analysis, AV 3.0’s emphasis on industry-driven innovation offers lessons for other countries. The US DOT’s policy of extreme deference may not be appropriate. Certain basic safety requirements may have to be mandated before vehicles are allowed on to public roads. Further, encouraging companies to share safety information is insufficient — they must be mandated to make public information about safety records, failure incidents and accidents.

AV 3.0 is a missed opportunity — a greater emphasis on safety was warranted and the DOT’s hands-off policy may erode public confidence in the long run. Finally, government policy has to be formulated to protect vulnerable groups such as truck and taxi drivers.

Targeted, flexible, and accessible educational opportunities, tax breaks, and other steps are necessary to prevent vulnerable groups from being left behind. Absent these steps, society may progress on two different roads: one will have people speeding ahead into a driver-less future whereas the other will have people moving backward into an unemployed and poor past. Governments must step up to the challenge.

Current vehicle safety standards are, unsurprisingly, written for vehicles driven by human beings. These are ill-suited to automated vehicles and new rules will have to be developed