November 22, 2016

BY ELECTRONIC SUBMISSION

Docket Management Facility
U.S. Department of Transportation
1200 New Jersey Avenue, SE
West Building Ground Floor, Room W12-140
Washington, DC 20590-0001
Re:

Request for Comment on Federal Automated Vehicle Policy
Docket No. NHTSA 2016-0090

Dear Sir or Madam:
The Consumer Technology Association (“CTA”) is pleased to submit these comments on
the Federal Automated Vehicle Policy recently released by the National Highway Traffic Safety
Administration (“NHTSA”).
CTA is a national trade association representing the entrepreneurs, technologists, and
innovators who are molding the future of the consumer technology industry. Our more than
2,200 member companies include many that are working to transform the driving experience
through the development of automated driving technologies and self-driving cars. These
advances are leading the way to expanded mobility, increased fuel efficiency, reduced traffic
congestion, and, above all, enhanced safety on our nation’s roads and highways.
CTA research shows that U.S. consumers are eager to experience the benefits of new
vehicle technologies. Our recent report, Self-Driving Vehicles: Consumer Sentiments, found that
60% of U.S. consumers are interested in purchasing an automated vehicle, and 70% would be
willing to test drive one. Further, 93% of U.S. consumers who have used existing driver
assistance technologies, such as adaptive cruise control, collision avoidance systems, and
parking assist, report satisfaction with these features – a positive indicator of consumer
acceptance of future innovations.
The Federal Automated Vehicle Policy is an important milestone on the road to offering
the automated driving technologies and, ultimately, the self-driving vehicles that consumers are
seeking to help keep their families and employees safe, and improve transportation mobility for
the elderly and disabled. CTA commends NHTSA for its foresight in taking the first steps toward
establishing a policy framework to advance these rapidly emerging life-saving technologies in
the U.S. In particular, we commend NHTSA for releasing its proposals as voluntary guidelines
and seeking public comment to inform its views.

 We also appreciate the emphasis on collaboration between government and industry
that is reflected in the policy. We are encouraged that, with this collaborative focus, NHTSA will
engage the expertise of both the traditional automotive industry as well as technology providers
across the automated vehicle industry. As CTA’s membership spans the broader industry, we
can be helpful in this regard. We firmly believe that a collaborative approach is essential to
ensure that new technologies are introduced in a way that both advances safety and puts the
U.S. at the forefront of automated vehicle innovation.
Safety and innovation are complimentary concepts. In fact, innovation and technology
development are proven drivers of improved motor vehicle safety. Since 1965, the number of
fatalities per 100 million vehicle miles traveled has experienced a four-fold decline,
notwithstanding substantial growth in our country’s population and the number of vehicles on
the road over that period. 1 This achievement is due in large part to technology advances like
seat belts, air bags, anti-lock braking systems, and electronic stability control.
More than any of these technologies, increasingly automated vehicles and, ultimately,
self-driving cars have the potential to drastically reduce the number of fatalities, rapidly moving
toward the “Vision Zero” goal established by NHTSA, the Federal Highway Administration, and
the Federal Motor Carrier Safety Administration in partnership with the National Safety Council.
Self-driving vehicles are expected to dramatically reduce crashes: If 90% of vehicles were selfdriving, U.S. traffic deaths are projected to decrease by 21,700 and save $447 billion per year. 2
It is further projected that, when 100% of vehicles are self-driving, U.S. traffic deaths would
decrease even more precipitously and save $1.3 trillion per year. 3
Realizing this vision requires recognition of the important role that technology expertise
plays in advancing safety, in conjunction with a policy approach that promotes innovation and
discovery. The comments that follow highlight aspects of NHTSA’s automated vehicle policy
that would foster innovation, and others that could impede the technology development process
of these life-saving innovations. Our comments also seek clarity about the application of certain
aspects of the NHTSA policy. We appreciate NHTSA’s attention to these comments and look
forward to partnering with the agency on future policy initiatives in this area. If we work together,
NHTSA, CTA’s member companies, and others in the motor vehicle and technology industries
can develop a policy framework for the U.S. that propels the nation to the forefront of automated
vehicle innovation and allows consumers and businesses to enjoy the transformative benefits of
life-saving automated driving technologies, while aiming for the highest level of safety.
1. Clarity and Consistency Foster Innovation
From an innovation standpoint, clarity and consistency are fundamental to any
successful policy framework. Clear, comprehensible guidelines make it easier for entrepreneurs
and innovators to understand their obligations, which in turn will improve compliance. Consistent
1

NAT’L HIGHWAY TRAFFIC SAFETY ADMIN., DOT HS 812 318, TRAFFIC SAFETY FACTS: RESEARCH NOTE 2 (2016),
https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812318.
2
Preparing a Nation for Autonomous Vehicles: Opportunities, Barriers and Policy Recommendations, Eno Center for
Transportation, at 8 (Oct. 2013), https://www.enotrans.org/etl-material/preparing-a-nation-for-autonomous-vehiclesopportunities-barriers-and-policy-recommendations/.
3
The ‘Internet of Things’ Is Now Connecting the Real Economy, Morgan Stanley, 82 (Apr. 2014),
http://www.technologyinvestor.com/wp-content/uploads/2014/09/internet-of-Things-2.pdf. Specifically, $488B savings
from accident avoidance, $507B productivity gain from autonomous cars, $158B fuel savings, $138B productivity gain
from congestion avoidance, and $11B fuel savings from congestion avoidance.

2

 policies and rules across jurisdictions also promote innovation, as they relieve the burden of
interpreting and applying multiple, varying laws and regulations. CTA supports the components
of the Federal Automated Vehicle Policy that set clear, rational “rules of the road” for vehicle
testing and development, and promote regulatory uniformity across jurisdictions.
a. Clarification of Federal and State Responsibilities
NHTSA’s clarification of the state and federal responsibilities with respect to automated
vehicles is particularly welcome in this regard. CTA strongly agrees with NHTSA’s conclusion
that responsibility for regulating motor vehicles and motor vehicle equipment should rest at the
federal level. This approach, which enables industry to focus its adherence on a single set of
rules, has both legal and practical support. Most importantly, the National Traffic and Motor
Vehicle Safety Act (“Safety Act”) clearly vests NHTSA with primary regulatory authority over
motor vehicles and motor vehicle equipment. 4 A national policy framework advances technology
innovation and promotes safety by simplifying and clarifying compliance to one set of rules. As
the policy states, “a manufacturer should be able to focus on developing a single [highly
automated vehicle] fleet rather than 50 different versions to meet individual state
requirements.” 5
For example, clarification of the state and federal roles is important with respect to the
regulation of human and non-human drivers. As the NHTSA policy recognizes, advances in
technology mean that motor vehicle equipment will increasingly assume “driving” tasks that
have traditionally been associated with human drivers. 6 CTA agrees that responsibility for
“licensing” non-human drivers should rest at the federal level, since the automated driving
function is inextricably linked to the performance and safety of motor vehicle software and
equipment. State licensing activities in this area could lead to inconsistent technical
requirements and other restrictions that could limit opportunities for testing and integration of
autonomous vehicles.
NHTSA’s model state policy can play an important role in promoting innovation by
encouraging clear and consistent state laws and regulations. As NHTSA acknowledges, broadly
consistent state rules and regulations are important to facilitate the introduction and operation of
automated vehicles – “drivers” should not have to “reinvent the wheel” every time they cross
state lines. 7 CTA encourages NHTSA to ensure that future state implementation of the model
policy serves this intended purpose. We also agree with NHTSA’s recommendation that states
“evaluate their current laws and regulations to address unnecessary impediments to the safe
testing, deployment, and operation of HAVs, and update references to a human driver as
appropriate.” 8
We also strongly encourage NHTSA to clarify that it is not asking states to adopt the
federal 15-point safety assessment at this time; implementation by states before the federal
policy is clear in its intent and obligations will lead to confusion and thwart innovation of lifesaving technology. We note the policy’s statement that “[t]he goal of State policies in this realm
4

49 U.S.C. § 30103(b)(1) (“When a motor vehicle safety standard is in effect under this chapter, a State or a political
subdivision of a State may prescribe or continue in effect a standard applicable to the same aspect of performance of
a motor vehicle or motor vehicle equipment only if the standard is identical to the standard prescribed under this
chapter. . . .”).
5
Nat. Highway Traffic Safety Admin., Federal Automated Vehicles Policy (2016), at 7 (hereinafter “NHTSA Policy”).
6
Id. at 38.
7
See id. at 39.
8
Id.

3

 need not be uniformity or identical laws and regulations.” 9 However, if state laws are not
uniform, they are per se different and therefore inconsistent. Spread across more than 50
jurisdictions, even minor inconsistencies can add up to a substantial compliance burden for the
motor vehicle industry, making it critical that clarity in the federal policy precedes state adoption.
Further, it is critical that application of the model state policy upholds the flexible,
collaborative approach indicated in the federal policy. We note that although the NHTSA Vehicle
Performance Guidance is explicitly “not mandatory,” 10 the model state policy encourages states
to require that manufacturers follow it – rendering the guidance effectively mandatory if states
adopt this recommendation. 11 As noted above, this inconsistency will create confusion about
compliance obligations, and is likely to thwart the legal testing of highly automated vehicles
(“HAVs”) that is already underway. The model state policy should mirror the federal policy in
treating the Vehicle Performance Guidance as establishing voluntary best practices.
b. Application of Current Law and Regulation
Understanding the application of the Federal Motor Vehicle Safety Standards (“FMVSS”)
to automated driving technologies and self-driving cars is crucial to the development, adoption,
and mass deployment of these life-saving innovations. CTA appreciates NHTSA’s explanation
of how its current regulatory tools can be applied to meet the FMVSS. In particular, NHTSA’s
goal of acting on interpretation requests within 60 days and exemption applications within six
months is a positive step toward helping the regulatory process address the rapid pace of
technology innovation.
To accelerate the testing and deployment of automated vehicles, CTA strongly supports
NHTSA’s proposal to expand the statutory exemption authority to increase the number of
vehicles and extend the timeline for exemptions. 12 The NHTSA policy describes challenges
created by the current statutory limits that CTA members have experienced in practice: “[t]he
current production volume limit . . . make it difficult to generate sufficient data for analysis (by
manufacturers, government, and other researchers) that could enhance safety. The limited
duration of exemptions require frequent and repeated application renewals, and cause
uncertainty as to the availability of the exemption over a longer period, which makes planning
difficult” (sic). 13 Expanding NHTSA’s exemption authority would allow manufacturers and other
entities to gather the data they need to improve safety and performance, while preserving the
agency’s oversight authority through the terms and conditions of individual exemptions. We fully
support this proposal and urge NHTSA to pursue it with congressional leaders.
The letters of interpretation issued to Google 14 and BMW 15 are evidence of the advances
in innovation that NHTSA’s current regulatory tools can accommodate. Absent the promulgation
of FMVSS that pertain specifically to automated driving technologies, these tools offer the
clearest way for industry to understand the applicable federal safety requirements. We
encourage NHTSA to work with industry on additional ways the agency could accelerate its
9

Id.
Id. at 11.
11
See id. at 41-42 (“The application should include the manufacturer’s or other entity’s safety and compliance plan for
testing vehicles, which should include a self-certification of testing and compliance to NHTSA’s Vehicle Performance
Guidance for the technology in the test vehicles . . . .”).
12
NHTSA POLICY, at 75; 49 U.S. Code § 30113(b).
13
Id. at 76.
14
Nat. Highway Traffic Safety Admin., Google – Compiled Response to 12 Nov 15 Interp. Request (Feb. 4, 2016).
15
Nat. Highway Traffic Safety Admin., BMW of North America – Response to 5 Oct 15 Interp. Request (Jan. 4, 2016).
10

4

 processes and timelines, as well as further expansion of exemptions to enable world-leading
automated vehicle testing and deployment in the U.S.
c. Adoption of Consistent Definitions
Finally, CTA commends NHTSA for conforming its definitions of certain terms to reflect
the meanings that are commonly understood in the industry. In particular, CTA is pleased that
NHTSA has adopted the SAE International (“SAE”) definitions for levels of automation. Using
definitions developed by an international industry-driven, standards-setting organization like
SAE promotes innovation by establishing a common “language” for policy and commercial
initiatives. This approach promotes global consistency, as it facilitates the movement of ideas
and technology across borders. The adoption of the SAE levels of automation is especially
important to clearly distinguish the types of automated systems to which the NHTSA policy is
intended to apply at both the federal and state levels.
2. Clarification of NHTSA’s Vehicle Performance Guidance is Necessary to Promote
Safety and Innovation
The Vehicle Performance Guidance forms the substantive core of the NHTSA policy. We
commend NHTSA for having the foresight to make this guidance voluntary, as well as noting the
importance of transparency and collaboration. While the guidance identifies the key safety
considerations for testing and deployment of automated vehicles, there are multiple areas
where it is not clear. In addition, some aspects of the NHTSA policy address issues that have
already been considered, or are currently being addressed, by federal agencies and industry
groups.
As described above, clarity serves the twin purposes of improving safety compliance and
fostering an environment that is conducive to innovation. CTA strongly encourages NHTSA to
further refine and simplify the Vehicle Performance Guidance, including working with industry to
collaboratively define rational obligations to achieve these goals. As part of this effort, we
encourage NHTSA to engage the expertise of the broad automated vehicle technology industry
to better understand the development and testing processes for software, hardware, and other
equipment discussed in the agency’s guidance. It is especially critical that each of the 15-Point
Safety Assessment and other obligations do not create overly burdensome requirements that
are inconsistent with the realities of technology development and testing, as this will
inadvertently thwart life-saving innovations in the U.S.
a. Clearly Define Industry Obligations by Carefully Specifying the Scope and
Applicability of the Vehicle Performance Guidance
As presently conceived, the scope of the Vehicle Performance Guidance is quite broad
and in many aspects unclear and inconsistent with the realities of automated vehicle technology
innovation. It covers both test- and production-level vehicles, as well as nearly every participant
in the motor vehicle industry – including “traditional vehicle manufacturers and other entities
involved with manufacturing, designing, supplying, testing, selling, operating, or deploying highly
automated vehicles.” 16 The breadth of the guidance raises important questions about the
obligations themselves, as well as the manner of and responsibility for compliance.

16

Id. at 11.

5

 i. Clarify Compliance Expectations
First, without repeatable, objective test methodologies it is not clear how to demonstrate
compliance with certain elements of the guidance. NHTSA should seek input from industry to
help define test methodologies that are both technology-neutral and sufficiently flexible to allow
for innovation over time. Without this clarity, how can manufacturers and other entities be
assured that they have satisfied NHTSA’s expectations with respect to open-ended assessment
categories like “Human Machine Interface” or “Ethical Considerations”? What does it mean to
say that a particular vehicle “meets” or “complies” with these elements? With respect to
functional safety, what do “all necessary and appropriate steps” (to debug software) mean?
Which entity will be responsible or liable for consequences of non-compliance? While CTA
notes NHTSA’s intention to clarify the Vehicle Performance Guidance through future action,
additional explanation and specificity would help industry understand compliance expectations
in the interim, albeit voluntary. Going forward, it also will be important for NHTSA to collaborate
with industry regarding the obligations for which it may make sense for the agency to seek
regulatory authority, whereas others that should remain voluntary.
In addition, CTA also would appreciate clarification of the specific systems, technologies,
and equipment to which the guidance is intended to apply. CTA has previously expressed the
position that the regulation of certain stand-alone software or mobile device applications would
exceed NHTSA’s statutory authority. 17 This is but one example of questions regarding the
extent of NHTSA’s proposed authority over software, hardware and other equipment. Further
specification is necessary to help clarify whether the agency’s proposed regulatory and
enforcement activities fit within the bounds of the Safety Act.
ii. Clarify the “Materiality” Threshold for Updates to Safety Assessments
Similarly, the guidance calls for submission of an updated Safety Assessment in the
event of any “material” software or hardware updates – specifically “[i]f these software or
hardware updates materially change the way in which the vehicle complies (or take it out of
compliance) with any of the 15 elements of the Guidance.” 18 Notably, there is a marked
difference between still being in compliance vs. being out of compliance. CTA encourages
NHTSA to work with industry to determine whether continued compliance is indeed “material” in
this context and sufficient to necessitate an update, or whether being out of compliance should
be the determinant. Although the guidance posits some illustrative examples, additional clarity
about the materiality threshold is necessary for manufacturers and other entities to better
understand their obligations in this area. Here also, the rule must be rational in the testing
context; this implies a high bar for what is considered a “material” update, thereby setting
expectations that align with the practicalities of the technology innovation process.
This clarification is particularly important to foster open innovation in the research and
development process. Development of advanced automotive technologies is incredibly iterative,
with changes being made regularly, sometimes multiple times daily, to improve and strengthen
systems under development. This process would be significantly disrupted by excessive
application of the Safety Assessment recommendation and could drive innovation and testing to
other countries. Consequently, it is important that the “significant update” rule be rationally

17

Consumer Technology Association, Response to Request for Public Comments on NHTSA Enforcement Guidance
Bulletin 2016-02: Safety Related Defects and Emerging Automotive Technologies (May 2, 2016).
18
NHTSA Policy, at 17.

6

 enforced, with expectations that comprehend the practicalities of the iterative technology
innovation process.
iii. Clarify the Procedures for Submitting Safety Assessments
Finally, clarification of certain “nuts-and-bolts” procedural matters would greatly
streamline research and testing of automated vehicles. Importantly, the policy does not specify
the point at which testing or deployment can begin following submission of the recommended
Safety Assessment. We suggest that NHTSA clarify that testing or deployment can begin upon
the agency’s confirmation that it received the Safety Assessment. This approach preserves
flexibility and market competition, and avoids creating a compliance mechanism that could
resemble non-voluntary pre-market approval, as well as open the door to subjective nontechnology neutral outcomes.
b. Duplication of Federal Regulatory Efforts
The cybersecurity and privacy elements of the guidance address issues that have
already been, or currently are, the subject of considerable discussion and analysis among other
federal agencies and industry. NHTSA’s recent enforcement guidance bulletin already explains
the agency’s approach to the cybersecurity challenges presented by connected and
autonomous vehicles. 19 Layering additional considerations on top of the concise and complete
analysis in the guidance bulletin only serves to create additional complexity for industry with no
apparent value added. Moreover, multi-year, multi-stakeholder processes involving industry,
academia and government like the NIST Cybersecurity Framework already provide clear,
detailed guidance in this area.
This is also true of the consumer privacy obligations outlined in the NHTSA policy. Since
2014, the industry’s obligations in this area have been guided by the Privacy Principles for
Vehicle Technologies and Services developed by the Alliance of Automobile Manufacturers and
the Association of Global Automakers. Absent the identification of any deficiency of these
principles, CTA is uncertain why the new standards in the Federal Automated Vehicle Policy are
necessary.
3. NHTSA Should Avoid Undue Interference with Research, Discovery and
Development
In issuing the Federal Automated Vehicle Policy as agency guidance rather than through
a formal rulemaking, NHTSA expressed an interest in ensuring that “premature, static regulatory
requirements do not hinder innovation and diffusion of the dynamic technologies that are being
developed in the industry.” 20 However, certain components of the policy would have precisely
that negative effect on innovation, causing the U.S. to fall behind in the global race to automated
vehicle leadership. Regardless of whether agency action is at the guidance or rulemaking stage,
it should not impede the industry’s efforts to develop innovative new safety technologies and
other advances. CTA urges NHTSA to steer clear of guidance, actions or regulation that could
lead to these unintended consequences. The following are examples of particularly concerning
guidance that would cause undue interference with life-saving autonomous vehicle innovation.
19

See NHTSA Enforcement Guidance Bulletin 2016–02: Safety-Related Defects and Automated Safety
Technologies, 81 Fed. Reg. 65705-9 (Sep. 23, 2016) (hereinafter, the “NHTSA Bulletin”) (describing the application of
NHTSA’s enforcement authority to software and related security concerns).
20
Request for Comment on “Federal Automated Vehicles Policy,” 81 Fed. Reg. 65703, 4 (Sept. 23, 2016).

7

 a. Pre-Market Approval
CTA is most concerned about NHTSA’s suggestion of the possible introduction of premarket approval for HAVs. Shifting from the current system of manufacturer self-certification to
pre-market approval would have a chilling effect on innovation by profoundly altering the
dynamics of product development in the motor vehicle industry, as well as impacting the
hallmarks of open competition and technology neutrality that drive innovation and investment in
the U.S. marketplace. The experience of other federal regulators suggests that this would be the
wrong step for NHTSA to take, and one that would be unwarranted given the proven track
record of the agency’s current policy approach.
i. An Interruption of the Innovation Cycle
Technology advances quickly. Ten years ago, highly automated vehicles were largely
theoretical. Today they can already be found as test vehicles on the streets of our country’s
cities and towns, as well as numerous other countries around the globe. As with all innovations,
these advances have been made possible by a process of industry investment in continual
research, discovery, development, and refinement. Observing how technology performs both in
controlled environments and, most importantly, under real-world conditions is a key element of
this process. More opportunities for observation enable learning that informs improvements to
the next generation – enhancing functionality, efficiency, and safety, among other features.
Pre-market approval would unduly constrain this cycle of open innovation. Bringing a
product from invention to market under such a system would require considerably more time
and expense for government and industry alike. One result would be to delay the process of
proof of concept to refinement to prototype and then to production, leading to the slower
introduction of new life-saving products into the U.S. marketplace relative to other leading
countries. In addition, the significant investment of resources and time demanded by the premarket approval process could discourage investment in novel technologies. It also would
impact market competition and risk non-technology neutral outcomes where the agency
indirectly picks winners and losers by affecting competing stakeholders’ time-to-market. These
considerations could outweigh the benefits of pursuing new automated vehicle innovations in
the U.S., and lead to a chilling effect on innovation and competition in this country despite
NHTSA’s best intentions.
Ultimately, pre-market approval would drive automated vehicle technology development
away from a process that is driven by entrepreneurs and innovators, toward one directed by
regulators. It goes without saying that this would not create an ideal environment for innovation
to thrive in this country. Given that advances in technology are fundamental to motor vehicle
safety, CTA urges NHTSA not to pursue regulatory tools like pre-market approval that would
significantly hinder life-saving innovations in the U.S.
ii. Regulatory Experience in Other Industries Does Not Support Pre-Market
Approval
The examples of other federal pre-market approval processes cited by NHTSA illustrate
the immense challenge of translating this approach to the motor vehicle context. NHTSA
focuses on the type certification process employed by the Federal Aviation Administration
(“FAA”), noting that the FAA’s “challenges seem closest to those that NHTSA faces in dealing

8

 with HAVs.” 21 However, the subtext of NHTSA’s summary of its discussions with the FAA is that
the pre-approval model is not suited to the scale and diversity of the motor vehicle marketplace:
In discussions with NHTSA about usefulness and feasibility of NHTSA’s requiring some
type of pre-market approval as a precondition to the manufacturing and selling of HAVs,
FAA noted that there were significant differences between the industries and products
FAA regulates and those NHTSA regulates in terms of the number of manufacturers,
number of models, and number and frequency of new model introductions. For example,
the FAA deals with only a few manufacturers and only rarely needs to approve an entirely
new model of an airliner. NHTSA further notes that the motor vehicle industry’s longestablished practice of introducing and producing motor vehicles on a model-year basis
might create challenges for the industry due to potential delays in the beginning of
22
production of vehicle models caused by the length of the approval process.

These concerns are borne out by the facts. In 2015 alone, manufacturers introduced 46
new motor vehicle models in the United States. 23 Those new entrants joined 221 models
already on the market, many of which were updated for the 2015 model year. 24 These statistics
– which don’t even include the dozens of light trucks, trailers, motorcycles, and other vehicles
introduced each year – raise the prospect that NHTSA could one day be faced with reviewing
and approving hundreds of vehicle models each year. As noted above, such an undertaking
would require significant investments of time and resources. Even then, there is no guarantee
that regulators would be able to keep pace with the frequency of product development and
innovation in the motor vehicle industry. Indeed, an appendix to the NHTSA policy notes that the
FAA’s pre-market approval process for the Boeing 787 aircraft required more than 200,000
hours of work over a period of eight years. 25 The FAA experience speaks to the delays and
uncertainty that pre-market approval could create. Simply put, regulators like the FAA and
NHTSA are not poised to stay apace of technology – and trying to force the automotive and
technology ecosystem into such an ill-fitting process will only harm U.S. innovation and global
competitiveness.
The FAA recognized this. Notably, when faced with a product whose characteristics
more closely parallel motor vehicles, the FAA explicitly chose not to require pre-market
approval. Like the motor vehicle industry, the emerging market for unmanned aircraft systems
(UAS) is highly diverse, with a number of different manufacturers and dozens of product lines.
As with motor vehicles, UAS operations implicate a number of public safety considerations.
Although the FAA considered pre-market approval for UAS as part of its rulemaking process for
commercial operations, it ultimately decided against imposing such a requirement. The FAA’s
reasoning reflects a concern about the impact of pre-market approval on innovation. In
particular, the FAA noted that “Unmanned aircraft technologies continue to evolve at a rapid
pace . . . . rapidly evolving technologies could face obsolescence by the time the certification
process is complete.” 26 The FAA concluded that its operational rules for UAS were sufficient to
mitigate any safety risks. 27
Similar considerations should inform NHTSA’s evaluation of its pre-market approval
proposal for HAVs. The number of participants and product offerings in the motor vehicle
21

NHTSA Policy, at 95.
Id. at 73.
23
See Total Number of Existing and New Car Models Offered in the U.S. Market from 2000 to 2016, STATISTA.COM
(Oct. 31, 2016), https://www.statista.com/statistics/200092/total-number-of-car-models-on-the-us-market-since-1990.
24
Id.
25
NHTSA Policy, at 95-96.
26
Operation and Certification of Small Unmanned Aircraft Systems, 81 Fed. Reg. 42064, 42182 (June 28, 2016).
27
Id.
22

9

 marketplace is simply too great to subject to pre-market review without compromising the speed
of introducing new life-saving technologies to market. In addition, we note that pre-market
approval regimes in the U.S. and elsewhere generally accord a measure of liability protection to
covered manufacturers and other entities. The NHTSA policy does not discuss the interaction
between a potential pre-market approval system and legal liability. Would NHTSA take on the
liability risk if it gives pre-market approval to an HAV? A complete understanding of this
interaction is necessary to fully review and analyze this proposal.
iii. Existing Regulatory Tools are Adequate
NHTSA’s discussion of significant new authorities like pre-market approval overlooks the
fact that its existing regulatory tools are more than capable of addressing the challenges
presented by automated vehicles. Indeed, the agency has explicitly acknowledged the flexibility
of its current regulatory approach. 28 The current system of manufacturer self-certification
adequately balances the interests of government and industry. The FMVSS allow NHTSA to set
high standards for the design and operation of motor vehicles. At the same time, selfcertification enables manufacturers and other entities to bring new innovations to market in a
timely, competitive and technology-neutral manner when they have demonstrated the requisite
level of safety – not when a particular feature has made its way through a bureaucratic review
process. This proven process allows NHTSA to focus its expertise and resources on defining
desired performance outcomes, rather than trying to specify or certify the technology employed
to meet that objective.
According to a comprehensive analysis conducted by the Department of
Transportation’s Volpe Center, “there are few barriers for automated vehicles to comply with the
FMVSS, as long as the vehicle does not significantly diverge from a conventional vehicle
design.” 29 For novel vehicle designs, the expanded exemption authority discussed above would
help facilitate testing and integration while NHTSA considers the promulgation of FMVSS that
are specifically tailored to new vehicle configurations.
Together with the enforcement framework described in NHTSA’s recent guidance
bulletin, this approach offers exactly the kind of nimbleness and flexibility that motivated the
agency’s search for “modern” regulatory tools. 30 The agency’s response to emerging
technologies that do not easily fit this framework should be to first proactively engage the broad
automated vehicle industry, and then if necessary, propose new FMVSS through the notice and
comment proceeding process; NHTSA should not discard an approach that has well served the
safety interests of the public for decades.
Importantly, for NHTSA’s current regulatory tools to function as intended, the federal
government’s primary role in motor vehicle regulation must be clear. Conflicting and unduly
burdensome state regulations will otherwise limit opportunities to develop, test, and deploy
automated vehicles. This is why it is so important that, as discussed above, state policies be
implemented in a way that respects the flexible approach reflected in the NHTSA policy.

28

See NHTSA Bulletin, at 65707 (noting that “NHTSA’s statutory enforcement authority is sufficiently general and
flexible to keep pace with . . . innovation.”).
29
John A. Volpe National Transportation Systems Center, Review of Federal Motor Vehicle Safety Standards
(FMVSS)
for
Automated
Vehicles
–
Preliminary
Report
(2016),
http://ntl.bts.gov/lib/57000/57000/57076/Review_FMVSS_AV_Scan.pdf.
30
Id. at viii.

10

 iv. Other Proposed Tools are an Imperfect Solution
Finally, CTA notes that the hybrid certification/approval process described in the NHTSA
policy would not overcome the challenges described above. Pre-market approval of any kind
would necessarily introduce delay and uncertainty into the product development process,
making it more difficult to develop and deploy new technologies in a competitive manner both in
the U.S. and relative to other developed countries leading in automated vehicle innovation.
With respect to other regulatory tools NHTSA might pursue, CTA encourages the
agency to clearly distinguish the way in which the “cease-and-desist” authority described in the
Federal Automated Vehicle Policy would differ from its existing recall authority under the Safety
Act. The exercise of both authorities would appear to have essentially the same effect in terms
of the removal of unsafe vehicles from public operation. Clarification of the distinguishing
features of “cease-and-desist” authority would help industry better understand its intended
purpose.
b. Post-Sale Authority to Regulate Software Changes
CTA also urges NHTSA to reconsider its proposal to seek post-sale authority to regulate
software changes. 31 Similar to the concerns raised above with respect to pre-market approval,
regulatory review of new software developments could delay the deployment of life-saving
innovations. For example, Tesla’s ability to make “over the air” software updates has allowed it
to address emerging operational and safety issues in real time, minimizing the risks to
consumers. Delaying such updates to allow time for regulatory consideration and approvals,
including through agency simulations of the effects of software changes, runs the risk of
compromising safety – precisely the opposite of what this proposed regulatory tool is intended
to achieve. In addition, an assessment and review process for software updates also raises
competitive and technology neutrality considerations, as it places government in the role of
“gatekeeper” with respect to which vehicles incorporate the most up-to-date technologies and
which technologies get implemented. The negative effects on innovation and competition far
outweigh any incremental benefits.
NHTSA’s discussion of potential software simulations also raises the question of agency
competencies, as does its proposal to monitor the vehicle design and production process to
ensure functional and system safety. 32 NHTSA’s core competency is as a regulator, not in the
testing or design of software. As such, it would be inappropriate for NHTSA to take on the role
of evaluating software updates or determining whether manufacturers should redesign their
software, as suggested by the NHTSA policy. Industry is in the best position to test and modify
software designs, and has clear incentives to do so in a way that upholds safety.
c. Application of Vehicle Performance Guidance to Lower Levels of Automation
CTA disagrees with the application of NHTSA’s Vehicle Performance Guidance to
vehicles at lower levels of automation (i.e., SAE levels L1-L2). The NHTSA policy acknowledges
a “clear technical distinction” between these vehicles and HAVs in which an automated system
performs some or all of the dynamic driving task, with good reason: vehicles equipped with
driving automation features classified as L1 and L2 require a human driver to complete the

31
32

NHTSA Policy, at 76-77.
Id. at 78.

11

 dynamic driving task in real time. 33 The need for human participation in the dynamic driving task
means that the regulatory and safety issues presented by less automated vehicles are
fundamentally different than for HAVs. Subjecting both categories of vehicles to substantially the
same policy guidelines blurs this important distinction.
As described above, clarity is essential for innovation. Although the NHTSA policy
identifies the categories of the guidance that pertain to less automated vehicles, the actual
application of each category is much less defined. For example, the policy acknowledges the
“complex human factors issue surrounding SAE Level 2 systems,” but does not explain with
specificity how those factors influence compliance with the Vehicle Performance Guidance. 34
Instead, the policy “encourages the automotive industry to work with NHTSA to develop
appropriate methods and metrics to understand and quantify effective human factors
approaches.” 35 CTA respectfully submits that the time for the guidance to apply to less
automated vehicles is after this research has been completed. Applying the guidance without a
full understanding of the technical ramifications creates uncertainty that serves neither the
interests of public safety nor innovation.
This uncertainty is compounded by the fact that vehicles equipped with driving
automation features classified as L1 and L2 are already on the roads. However, the extent to
which the Vehicle Performance Guidance applies retroactively is unclear. For example, although
the policy indicates that the recommended Safety Assessment for HAV systems already being
tested and deployed should be submitted within four months of the completion of the Paperwork
Reduction Act process, the expectations with respect to less automated vehicles are
undefined. 36 The best way for NHTSA to resolve these and other ambiguities would be to
acknowledge the significant technical distinctions of less automated technologies and issue a
clarification that takes account of the unique characteristics of technologies in this category.
d. Data Sharing Provisions
Finally, NHTSA advises manufacturers and other entities to “develop a plan for sharing
[their] event reconstruction and other relevant data with other entities” including the agency
itself. 37 CTA agrees that data analysis and machine learning are key to enhancing the safety
and performance of HAVs while companies are testing their technology. We also note that the
value of various event and outcome data varies. Some data can be more useful than other data,
and industry – on the front lines of this testing – is in the best position to determine what
information is most useful and appropriate to share and how to share it. For this reason, we very
much appreciate the discretion that the agency has provided regarding data sharing according
to industry-led plans that can take account of the varying value of test data, as well as the
commercial, proprietary, privacy, and intellectual property considerations. NHTSA should be
very cautious in considering going beyond this approach to impose any mandatory data sharing
requirements via Vehicle Performance Guidance safety assessments, rulemaking, or other
means, as this would have a deleterious effect on innovation.
NHTSA also should recognize that even anonymized or “de-identified” data can
constitute proprietary and competitive information. This is particularly true of data generated
through the research and development process, which can indicate the direction of new
33

Id. at 32.
Id.
35
Id.
36
Id. at 16.
37
Id. at 18.
34

12

 discoveries and current and future investments. A blanket requirement of compulsory data
sharing with third parties (and even NHTSA) could enable less innovative market participants to
benefit in an anti-competitive manner from the data and the investments of others. This would
distort incentives for developing and testing automated vehicles in the U.S. In addition to
distorting incentives to invest in innovation, this could prevent the best-performing technologies
from distinguishing themselves in the marketplace, potentially putting U.S. industry at a
disadvantage globally. These considerations are compounded by the liability and intellectual
property concerns inherent to the research and development process. For all of these reasons,
CTA urges NHTSA not to propose any affirmative data sharing obligations.
***
Thank you for the opportunity to provide these comments. We look forward to working
with you as the Federal Automated Vehicle Policy evolves into a world-leading policy framework
for promoting life-saving HAV innovation, adoption, and deployment. Please do not hesitate to
contact us with any questions you may have.
Sincerely,
/s/ Gary Shapiro
Gary Shapiro
President and CEO
Consumer Technology Association
1919 S Eads St
Arlington, VA 22206
/s/ Jamie Boone
Jamie Boone
Director, Government Affairs
Consumer Technology Association
1919 S Eads St
Arlington, VA 22206

13