Automated Driving Regulation: Three Big Issues as the Federal Government Edges Forward

Mobility stakeholders are eager to learn how the Biden Administration will approach automated driving systems (ADS), including to what extent this approach may differ from that of the Trump Administration. We may learn in the not-too-distant future, thanks in significant part to two (genuine) parting gifts from the prior administration: a detailed summary of the federal government's recent efforts and a new—and still open—rulemaking docket in which interested parties can comment on the path forward.

Comments submitted to date reflect deeper divisions permeating regulatory debate, including the respective roles of federal and state governments and the reliance on voluntary and international standards. Another important question under discussion is whether the time has come for the federal government to move from regulating by exemption to promulgating binding rules.

As the deadline for comments is April 1, 2021, we must wait to see how the Biden Administration reacts to these submissions. In the interim, additional interested parties—including ADS developers, automotive and equipment manufacturers, trade associations, consumer groups, state and local transportation authorities, and drivers, pedestrians, and bicyclists—will likely weigh in on these and other divisive issues.

Two Parting Gifts

The U.S. Department of Transportation (DOT) released its Automated Vehicles Comprehensive Plan on January 11, 2021, to provide a broad overview of the federal government's recent actions to "support[]the safe integration of ADS into the surface transportation system." Laying the groundwork for future action, the Comprehensive Plan detailed the DOT's three goals "to prioritize safety while preparing for the future of transportation":

• (1) Promote collaboration and transparency;
• (2) Modernize the regulatory environment; and
• (3) Prepare the transportation system for ADS integration.

The DOT also provided examples of its progress, including engaging with stakeholders, publishing guidance, granting exemptions, continuing research, and funding demonstration projects. One notable omission remains: The federal government has yet to promulgate regulations governing, or even a regulatory framework for, ADS-equipped systems.¹

This takes us to the other significant development at the tail end of the Trump Administration: On November 19, 2020, the National Highway Traffic Safety Administration (NHTSA) released for public comment an Advance Notice of Proposed Rulemaking (ANPRM) on a "Framework for Automated Driving System Safety."

Broadly speaking, a Notice of Proposed Rulemaking (NPRM) leads directly to a new federal rule. In the ordinary course, an agency describes a proposed rule, stakeholders comment on that proposal, and the agency finalizes the rule incorporating any modifications deemed necessary in light of public comments received.

By contrast, an ANPRM is still a preliminary document much like a "Notice of Inquiry." An ANPRM is most useful to an agency considering regulatory action that requires more information before proceeding to an NPRM. Against this regulatory backdrop, it is important to understand that this ANPRM does not necessarily indicate NHTSA is anticipating a wholesale overhaul of the Federal Motor Vehicle Safety Standards (FMVSS) or promulgation of other performance requirements to address ADS.

It is also clear when reviewing the ANPRM itself that NHTSA intended it as a preliminary step towards more clarity in federal regulation. NHTSA explained it "is considering the creation of a governmental safety framework specifically tailored to ADS." Specifically, NHTSA sought comment on the selection and design of "the structure and key elements of a framework," including "appropriate administrative mechanisms to achieve the goals of improving safety, mitigating risk, and enabling the development and introduction of new safety innovations."

NHTSA further emphasized its current view that the agency should focus on "four primary functions" of ADS:

• Sensing (how the ADS receives information about its environment);
• Perception (how the ADS detects or categorizes road users, infrastructure and conditions);
• Planning (how the ADS analyzes the situation, plans, and makes decisions); and
• Control (how the ADS executes dynamic driving functions).

Finally, the ANPRM listed 25 specific questions on which it seeks public input from stakeholders, divided into four categories:

• Questions about statutory authority;
• Questions about a safety framework;
• Questions about administrative mechanisms; and
• Questions about NHTSA research.

Stakeholder Responses to NHTSA's 25 Questions

Initially, NHTSA asked for comments by February 1, 2021, but extended the deadline to April 1, 2021. Already, more than 650 comments have been submitted—likely a reflection of the last-minute extension. Now, several weeks remain before the deadline for comments.

After analyzing the comments submitted, we highlight some of the statements we expect will garner the most attention by the agency and stakeholders. It would not be surprising to see future comments responding to those already submitted. It is all fair game for rebuttal should other stakeholders disagree. While not an exhaustive list, some of the more interesting comments address three of NHTSA's four categories of questions.

1. What Is NHTSA's Authority Over ADS?

Although the questions NHTSA posed about the scope of its (and DOT's) authority were primarily focused on whether the agency could promulgate rules or use enforcement actions to police ADS development and implementation, the most notable comments to date have instead addressed two related issues.

First, some commenters have discussed the appropriate division of responsibility between the federal and state governments in this space. The National Conference of State Legislatures (NCSL), for example, states that the "federal government must not encroach into this space as it could inadvertently create significant roadblocks for the deployment of autonomous vehicles and erode the agency's spirit of cooperative federalism."

These issues of federalism are perennial in heavily regulated industries, with many states wanting to retain authority and many in industry endorsing the view that the only thing worse than federal regulation is regulation by 50 states plus countless territories and localities.

Second, commenters have pressed NHTSA to implement only performance-based safety standards, rather than to develop design-based standards. The University of California, Davis Policy Institute for Energy, Environment and the Economy (UC Davis), for example, stresses that "Performance-based standards, rather than design-based standards, such as safety benchmarks, will be more flexible and allow the industry to evolve in a manner that preserves safety and consistency."

The comments also stress that the agency has and should utilize its authority to promulgate such standards: "[C]ertainty and clarity will allow innovation, while ensuring that minimum levels of safety are achieved. Binding (rather than voluntary) performance-based ADS safety standards within the FMVSS will likely result in more consistent and improved safety outcomes." This comment fits well with NHTSA's longstanding view that it regulates best through performance-based rather than design-based standards.

2. What Should NHTSA's Safety Framework Look Like?

NHTSA posed 13 questions about what its safety framework should look like, including specific questions about whether the agency had identified the correct core elements, what NHTSA should be doing "at this early point in the development of ADS," and how its safety framework could both improve safety and enhance consumer confidence. Comments here raised a number of interesting points.

First, at least one commenter highlighted that NHTSA should specifically address connected vehicles in its framework for the development of ADS. The Maryland Department of Transportation (MDOT), which generally supported NHTSA's proposed framework, stressed that connected vehicles (CVs)—vehicles that use not only internal sensors but also vehicle-to-vehicle and vehicle-to-everything technologies—"must be integrated in a national safety framework for ADS, as [CV technology] provides an important redundancy for ADS and enables additional safety-related services beyond what in-vehicle sensors can provide." This comment raises fascinating and difficult issues for NHTSA, given the limited scope of the agency's jurisdiction and the regulatory authority of the FCC and others over radiofrequency spectrum.

Commenters also stressed that NHTSA's safety framework needs to be built in part on voluntary and international standards, not just through the exercise of regulatory fiat. Underwriters Laboratories, for example, noted that it "is pleased that NHTSA recognizes the value of existing consensus standards and has requested comment on a framework that envisions a relationship to the work done by the private sector," and "applauds DOT's specific reference to UL 4600, Standard for Safety for the Evaluation of Autonomous Products."

The Insurance Institute for Highway Safety (IIHS) similarly stressed that "ISO2148, ISO26262, UL4600 or some combination of the three could be used as the basis for" NHTSA to "shape ADS safety ahead of its deployment onto our nation's roads." Further, Mapless AI noted the significance to "safety suppliers" of "the voluntary application of ISO 26262, ISO 21448, and ANSI/UL 4600 enhanced by future NHTSA research." UL 4600 envisions a standard for ensuring a valid safety case rather than requiring a particular technical approach.

The ISO standards represent strides in designing safe vehicles requiring rigorous engineering rules. Because the technology inherent in vehicles equipped with ADS exceeds the scope of traditional safety standards, NHTSA will likely consider these or alternative state-of-the-art safety case approaches; but the incorporation of external standards into federal regulatory schemes is fraught with complexity.

A number of commenters suggested that NHTSA focus on clarifying the terminology used about ADS. UC Davis suggests, for example, additional "clarity on AV safety terminology would be beneficial. Several terms to include are safety, security, as well as unreasonable risk, as they apply to AV operation." This is a perennial problem, and one with significant implications not only for safety but also for public awareness and acceptance of ADS.

Finally, commenters stressed the need to include in the developing safety framework not only occupants of vehicles equipped with ADS, but also people outside the vehicle. NHTSA mentioned pedestrians, bicyclists, and others in the ANPRM—but only in passing. The California State Transportation Agency (CSTA), stressed that "[t]he determination of which FMVSS features may be exempted as non-applicable should be made through consideration for the safety of all persons, not just occupants of the tested vehicle, and ensure at least a comparable level of safety as a traditional vehicle."

Echoing the CSTA comment, the League of American Bicyclists emphasized that NHTSA should develop "a federal safety framework for ADS … where the safety of people outside of a vehicle is at least as important as the occupant of a vehicle." More than 550 individuals' comments included near-verbatim statements echoing the League's concerns.

3. What Administrative Mechanisms Should NHTSA Prioritize?

Some comments addressed NHTSA's questions about how it should proceed administratively, given the varying mechanisms on which the agency could rely to implement its developing safety framework.

First, commenters questioned NHTSA's ongoing reluctance to promulgate binding rules. IIHS, for example, stressed that it "remain[s] disappointed with NHTSA's dithering about how to approach the regulation of ADS safety. As our past comments suggest, there are several things NHTSA could require of ADS-controlled vehicles that would put them on the road to becoming safer than human-driven vehicles without stifling innovation. NHTSA's obsession with 'removing regulatory hurdles' and 'not stifling innovation' is inconsistent with the agency's mission 'to save lives, prevent injuries, and reduce economic costs due to road traffic crashes.'"

According to IIHS, "[t]he compelling need for early safety regulation is to obligate ADS developers to favor safety over competing demands of the technology (e.g., cost, speed, style)." We expect no one to disagree that safety is paramount, but we do expect disagreement about whether the time has come for binding regulations, given the rapidity with which ADS technologies are changing.

Second, some stressed that the agency should rely on third-party verification systems. The National Society of Professional Engineers (NSPE) in particular urged "NHTSA to include a third-party verification process as part of a federal safety framework. A third-party verification process should establish that the ADS technology under review meets a minimal level of safety, as determined by an assessment of risk. This can be done through the submittal of risk assessments audited by a professional engineer …."

Finally, as discussed above several commenters suggested that NHTSA incorporate existing voluntary industry standards into its framework.

Our Final Thought

We expect to see many more comments in the coming weeks. Depending on those comments and the Biden Administration response, we will likely garner additional clarity on the development of the federal regulation of ADS in the coming months.

For stakeholders, this rulemaking is an important opportunity to educate NHTSA and the broader DOT on how to move beyond the current safety standards, which focus on the physical design of vehicles, to broader questions including whether and how to regulate ADS performance. The comments submitted so far have added to this dialogue but many issues warrant further attention.

Should NHTSA, for example, develop standards based on disengagement data, or is disengagement an unreliable metric? Similarly, how should future rules address the period during which vehicles equipped with varying levels of automation occupy public roads at the same time? Whether taking the short or long view, such issues will likely affect how NHTSA proceeds and, thus, shape the regulation of ADS-equipped vehicles in the United States.

Future mobility is dynamic and ADS-equipped vehicles represent innovation and opportunity, but not without immensely complex issues. Mobility stakeholders still considering submitting comments to NHTSA may want to focus those comments on higher-level issues and tie answers to NHTSA's specific questions to a broader vision of the short- and medium-term future, not only of federal regulation but also of ADS technology, state and local regulation, and public acceptance.

As NHTSA recognizes, ADS presents an immense opportunity—but the path from today to the day when a significant number of ADS-equipped vehicles operate on public roads depends as much on public acceptance as it does on revisions to the Federal Motor Vehicle Safety Standards.

FOOTNOTES

1  On March 30, 2020, NHTSA published a Notice of Proposed Rulemaking (NPRM) to consider updates to the Federal Motor Vehicle Safety Standards (FMVSS) that would apply to automated vehicles that do not require a human driver. Under the existing safety standards, the front left seat is the driver's position, and there are various requirements related to the driver's position such as standards for seat belts and airbags. While the rulemaking did not cover all applicable FMVSS requirements, the NPRM represented a significant step by NHTSA towards a regulatory climate amenable to ADS-equipped vehicles that do not require a human driver.