Industry insiders, including government officials, cite an interesting analogy as a justification for their initial jump into the regulation of driverless vehicles that was first promulgated on September 20, 2016. The point made is that, had current regulations been in effect when the "Model T" hit the streets, we would have experienced far fewer collisions.
The first wave of USDOT regulations for driverless vehicles has four basic components:
- Vehicle Performance Guidance for manufacturers, developers and others, outlining a recommended 15-point program for safe design, development, testing, and deployment of self-driving vehicles prior to commercial sale.
- Model for state policy, spelling out what responsibilities states will still have (such as licensing, registration, traffic laws, insurance and liability, etc.), while carving out the Federal role for autonomous vehicles.
- The policy will use current regulatory rules to aid the development of safe vehicles by allowing flexibility in design and allowing limited exemptions for testing "nontraditional vehicle designs." The National Highway Traffic Safety Administration will also issue a bulletin clarifying its vehicle recall authority for cases when it identifies product defects.
- NHTSA will identify 12 potential "new tools, authorities, and resources" for ensuring safety. Some may be created under current law and some may require congressional action.
As the Chicago Mayor and City Council have already promulgated regulations prohibiting driverless vehicles from operating within the City, we are already headed for an institutional collision, irrespective of the magnitude and severity of vehicle collisions.
USDOT's introductory "message" cites a few important concerns regarding autonomous vehicles:
- Will they fully replace the human driver?
- What ethical judgments will they be called upon to make?
- What socioeconomic impacts flow from such a dramatic change?
- Will they disrupt the nature of privacy and security?
We should applaud USDOT for so openly acknowledging even these concerns. They are serious concerns. The most obvious of them was addressed in Part 1 of this series: Massive unemployment. But this is only the tip of the socio-economic iceberg. The next two questions I would next ask are: (1) Who will reap the economic rewards from this innovation? And, (2) "What will one of these vehicles cost at the purchasing and operating level?" Sub-questions follow almost relentlessly. For example:
- Will this technology make even more sense with electric vehicles than ICE-powered vehicles?
- If so, how will this shift affect the relationship between capital costs and operating costs?
As an illustration of the second sub-question here, operating costs for an electric-powered vehicle will be significantly less, while the capital costs for it are much higher -- due largely to the cost of the batteries. Similarly, the operating costs of an autonomous vehicle (or HAV as USDOT calls it) will be lower because there will be no drivers' salaries or fringe benefits to pay. However, the technology promising to replace the driver will almost certainly make the vehicle's purchase more expensive. So it would be important to have at least a crude vision of both these cost savings and expansion in order for a potential purchaser to calculate a somewhat-informed trade-off. Insofar as predicting the comparative liability exposure, I doubt even an ordinary crystal ball would be much help. Instead, predicting this comparison is more like hitting a 150-mph curveball while highly intoxicated.
Also, just as I stated in Part 1 of this series, USDOT's initial "message" includes the statement that, "The rise of new technology is inevitable." So I did not simply pull the title for this series out of the air. Nor is it simply my personal opinion. Instead, it seems to be the consensus of most regulatory agencies and public transportation industry magazine editors and publishers. It is almost certainly the dream of many suppliers, vehicle manufacturers and operating agencies and companies -- capital costs notwithstanding.
So as the installments before this one, and those to follow will do, this installment will attempt to help NBT readers make intelligent, responsible, ethnical and fair choices. In this world, selecting the "best buy" in a vehicle will no longer work as the center of the transportation industry universe. With the introduction and expansion of the poorly-named autonomous vehicle, decisions about planning, operations, marketing, safety and liability (among many others) will likely take center-stage. Those agencies and companies that "get t his right" will likely be the survivors of this transition. The class of triumphant profiteers is already obvious. And If Secretary Foxx is employing the term "soon" the way I use it, we better wise up quickly. Because these vehicles will not crawl into the public transportation landscape. They will zoom into it.
One of the less-subtle hints in the USDOT policy is that NHSTA, rather than FMCSA, will become the center of the regulatory universe. Along agency lines, this shift appears to make sense as the vehicles become their own drivers. As an odd coincidence, the schoolbus community is already used to NHTSA serving as the center of its regulatory universe, alongside state departments of transportation and education. The motorcoach community, activities of which are largely governed by the FMCSA, is not used to this relationship. Welcome to the future.
As someone pretty familiar with NHTSA's Federal Motor Vehicle Safety Standards (FMVSS), I cannot even guess at how these regulations will be restructured to capture the certification requirements for these new technologies. Will they be integrated into the four existing "layers" of the Standards (i.e., the 100 series, 200 series, 300 series and 400 series)? Or will NHTSA create an entirely new series to cover technologies designed to shift former driver responsibilities to the vehicle manufacturer? Further, an OEM may currently "self-certify" its vehicles as being compliant with these standards. I can only wonder if NHTSA will allow the same latitude to manufacturers of HAVs.
Within the mere quartet of USDOT's four concerns cited near the top of this article, the growth of self-driven vehicles should explode the value of stocks in the liquor industry, and remove the most-common argument against the legalization of marijuana: The first year of marijuana's recreational use in the State of Colorado translated into a doubling of vehicular collisions. Straying no further away from transportation, this single point illustrates the range, diversity and enormity of side-effects that technologies so new and dramatic are likely to generate. The closest comparison I can think of is how the emergence of the shipping container radically reduced shipping costs (associated with loading, unloading and breakage), and effectively led to an explosion of outsourcing. So a society where robots provide our mobility will be a vastly different society. And managing it will involve a radically new and different set of both procedures and priorities. USDOT is perceptive and honest in its initial inclusion of concerns about privacy and security. Again, the motorcoach world of the future will not revolve around the "best buy" in a vehicle. Nor will it revolve around notions like customer service -- a centerpiece of the current motorcoach industry's collective heart and commercial goals. What it will revolve around is a wild guess.
Among the challenges that will change radically are those of safety and liability. The USDOT document's Executive Summary cites the fact that 94% of U.S. roadway collisions in 2015 involved human choice or error. So as I pointed out in Part 1 of this series, the robots can crash into quite a few things and still have an exponentially better safety record than their human-driven counterparts. What will this mean as a liability matter? With these statistics, if your driver-operated bus or coach collides with a "highly automated vehicle," which party do you expect to soak up the blame in the inevitable lawsuit? If my hunch is correct, transportation agencies or companies may eventually be unable to even afford the risk of deploying humanoid-driven vehicles. If they cannot, because they cannot find or afford insurance for them, what will we do with all our non-autonomous vehicles? Sell them to Third World countries in a fire sale? And how will the insurance industry adjust to these realities, since their greatest concern, understandably, is liability?
Even the transportation vocabulary is changing. USDOT already has acronyms for vehicle-to-vehicle collisions (V2V) and vehicle-to-infrastructure collections (V2I). Vehicle design is obviously changing. One highly-visible article I recently reviewed contained a drawing of the HAV with rear-facing seats -- a configuration that actually makes sense, since seatbacks will distribute frontal impact forces (the highest in most accident scenarios) far better than any seatbelt could. But if HAVs remove fatigue from the operating environment (and the term "duty cycle" is reduced to largely a maintenance concern), it makes sense to equip motorcoaches with seats that morph into cots -- much like those which European passengers trains have had for decades. At the moment, most of the preliminary designs shared with the public have focused largely on style, while many of the early tinkering efforts associated with these vehicles' operation have been almost childlike. (Stay tuned for the follow-up article about cameras and sensors.)
In the normal world, those who think "outside the box" take huge risks. And while they are occasionally rewarded in huge ways, they are rarely rewarded in any way: Ninety-five percent of "start-up" companies fail within their first five years. In contrast, and particularly during the early emergence and development of HAVs, those who think best outside-the-box are far more likely to be the winners. Those who do not will be left behind. When this almost certain reality collides with an industry that rarely ever designs a transportation system (instead letting software optimize the chaos), and conducts virtually no monitoring, the transition to HAVs is likely to turn the industry inside out.
So new and unpredictable is the emergence of HAVs and their almost endlessly-widespread consequences, that USDOT is almost coy in "qualifying" its initial policy statement about them: "We are issuing this Policy as agency guidance rather than in a rulemaking [effort or format] in order to speed the delivery of an initial regulatory framework and best practices to guide manufacturers and other entities in the safe design, development, testing, and deployment of HAVs" (Executive Summary). Huh?
At the same time, the nature and pace of these changes almost demands a quick response, and such a response could never be formulated through a typical laborious three-stage rulemaking process. So while USDOT may be covering its institutional derriere with this statement, this gov-speak is understandable. The unpredictability of this technology, much less the range, diversity and severity of side-effects, is overwhelming. So despite the ambiguity of opening statements, we have to commend USDOT for at least opening our minds.
The remainder of USDOT's Federally Automated Vehicle Policy is divided into four sections:
- Vehicle Performance Guidance for Automated Vehicles
- Model State Policy
- NHTSA's Current Regulatory Tools
- New Tools and Authorities
Future installments of this NBT series will try to address some points related to these policy areas. And if USDOT's guess about the speed of this transition is even roughly correct, stay tuned and pay close attention. As some almost always announces at the beginning of race car event, "Drivers, Start your Research."