November 22, 2016
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: Notice and Request for Comments on Federal Automated Vehicles Policy, NHTSA Docket No. 2016-0090, Fed. Reg. No. 2016-22993.
On behalf of the Self-Driving Coalition for Safer Streets (“Coalition”), I am pleased to submit these comments regarding the National Highway Traffic Safety Administration’s (“NHTSA”) Request for Comment on the “Federal Automated Vehicles Policy” (“Policy”) published in the Federal Register on September 23, 2016.
The Self-Driving Coalition for Safer Streets was established by Ford Motor Company, Google, Lyft, Uber, and Volvo Cars in April 2016 to work with lawmakers, regulators, and the public to realize the potential safety and societal benefits of fully self-driving vehicles. The Coalition focuses on the development and deployment of Society of Automotive Engineers (“SAE”) Level 4 or 5 fully automated vehicles. We believe that it is at the fully automated levels where we will see the greatest opportunity in safety and mobility. The Coalition is dedicated to working collaboratively with civic organizations, municipalities, and businesses to bring the vision of fully self-driving vehicles—that is, vehicles at SAE Level 4 or 5, that do not require a human driver—to America’s roads and highways. Self-driving technology has the great potential to enhance public safety and mobility (especially for the elderly and disabled), reduce traffic congestion, improve environmental quality, and advance transportation efficiency, and the Coalition’s mission is to promote the benefits of fully self-driving vehicles and support the safe and rapid deployment of these innovative and potentially life-saving technologies.
Regarding section I, the Vehicle Performance Guidance (“Guidance”), the Coalition acknowledges NHTSA’s enforcement authority in the area of vehicle safety, but submits that the proposed Guidance: (1) may be interpreted to require manufacturers to submit more than an initial Safety Assessment to the agency, that may lead to repeated resubmissions outside of significant updates to HAVs or disclosure of proprietary information; and (2) does not affirmatively discourage states from mandating the use of this guidance--adherence to which is purely voluntary at the Federal level--during NHTSA’s policy development.
NHTSA should pursue narrow rulemakings to amend several key Federal Motor Vehicle Safety Standards currently mandating human operator controls in L4-L5 HAVs which do not need a human operator. These standards have already been identified in USDOT Volpe Center’s March 2016 report entitled “Review of Federal Motor Vehicle Standards for Automated Vehicles” published earlier this year and we encourage NHTSA to pursue targeted rulemakings to amend several of these standards to provide for the self-certification of vehicles that would allow fully self-driving operation without the presence of, or capability to use, human operator controls in L4-L5 vehicles, where the use of such controls is not needed. The Coalition expects to provide a summary list to the agency, including standards like the requirement that service brakes be activated by means of a foot control in FMVSS No.135 (Light Vehicle Brake Systems), and request that NHTSA begin action on such a rule as soon as possible. Given the long duration of a rulemaking process, the lack of amendments to key standards would stand in the way of design, testing, deployment of vehicles that have been designed as fully self-driving from the ground up.
Congress should enact legislation to facilitate the rapid deployment of HAVs. For example, Congress should revise NHTSA’s exemption authority to allow for a greater number of vehicles to be allowed on the road for testing and deployment of HAVs. While the current exemption limits an exempted fleet for a manufacturer to up to 2,500 vehicles for up to 2 years, the need to accumulate results for HAVs in an expedited timeframe would require a significant increase in the fleet size and a longer exemption period for the fleet. This flexibility would provide multiple avenues for manufacturers and innovators to safely test and deploy a number of vehicle changes that would advance the safety of HAVs and passenger comfort and utility.
I. The Safety Assessment Using the Vehicle Performance Guidance Should be Modified to Avoid Imposing Unnecessary and Costly Burdens
The Coalition acknowledges the critical function that NHTSA carries out in establishing and enforcing vehicle safety in the United States, and we seek to work with NHTSA and DOT to carry out this important mission. In that spirit, the Coalition welcomes the agency’s effort to craft the “Vehicle Performance Guidance” (“Guidance”) as a way to establish a voluntary national framework for the rapid and safe deployment of HAVs on America’s roads. However, the Coalition has identified several areas where we believe that the Guidance can be clarified and improved, as we describe below.
The Guidance should discourage states from pursuing legislation based on the Guidance and encourage a single voluntary national framework focused on safety assurance.
The Coalition strongly supports efforts to promote nationwide consistency in the framework applied to HAVs. With such transformational potential for improving safety, mobility, and the environment, the self-driving car industry will be severely hamstrung by constant attempts at the state and local levels to push legislation—whether based on this Guidance or otherwise—with regulatory compliance burdens layered on. To be sure, there are some areas of HAV deployment that may require state action, such as licensing and registration, but states have already shown that they do not plan to remain limited to these areas when considering legislation on HAVs.
As it stands now, a few states have considered or may consider legislative proposals that focus on safety of HAVs in a sensible and narrow fashion, but some have attempted to codify the 15 categories in NHTSA’s proposed Safety Assessment, despite the fact that the Guidance is still subject to potentially extensive change that would create the regulatory patchwork at the state level that the Guidance aims to help foreclose. As a general matter, establishing a single national framework must remain the goal of NHTSA and all stakeholders who want to see the rapid and safe deployment of HAVs. A patchwork of state and local laws will most assuredly restrain the growth and deployment of HAVs and inhibit the numerous and significant benefits that would inure to consumers in this area.
To accomplish this goal, NHTSA should clarify its presentation of the Vehicle Performance Guidance and Model State Policy with an explicit request that states refrain from further legislating in the area of HAV safety. This would give the industry an important level of clarity regarding where the regulatory framework is headed and reduce uncertainty about state action that could frustrate uniform HAV testing and deployment.
The Safety Assessment process requires clarification.
The Coalition supports the overall goal of using the Safety Assessment letter to “assist NHTSA, and the public, in evaluating how safety is being addressed by manufacturers and other entities developing and testing HAV systems.” Furthermore, the Coalition supports the agency’s approach to make the Safety Assessment letter a voluntary submission. There are concerns, however, that confidential business information (CBI) may be sought for inclusion in a Safety Assessment or supplement. If such information is unintentionally released or maliciously breached, it could have enormous repercussions in the marketplace with respect to development and evolution of this industry. From this perspective, NHTSA should focus on limiting the information being requested from manufacturers to avoid unnecessary accumulation of sensitive industry data and the potential commercial harm companies could suffer if publicly released. To the extent, however, that NHTSA accepts voluntary submissions of CBI during the Safety Assessment letter exchange, it should consider methods by which to streamline its Part 512 process, as discussed below.
Relatedly, we note that the Policy recommends that manufacturers submit a Safety Assessment Letter for both testing and deployment of an HAV system. Furthermore, the notion of an entity submitting an updated letter whenever changes unrelated to the safety of the overall system are made (e.g. basic software updates, testing in a new geography, or moving an existing system to a new vehicle) would create additional burdens without providing any valuable information to the agency. We recommend that NHTSA clarify the conditions for a “significant update” that would appropriately prompt an updated letter. The Coalition recommends that the agency define “significant update” to mean a substantial change to the HAV system that affects its capability or safety (e.g. a wholly re-defined ODD or a step-level increase in the system’s level of automation)with a company making the determination whether the change is indeed substantial. In addition, the Coalition believes that separate letters submitted for testing and deployment are of little value to the agency without a significant update to the aforementioned HAV system. It should warrant nothing more than a basic amendment to any existing submission.
Safety Assessment Letter: Process, Applicability and Scope, and Structure and Content
The Coalition recognizes that one of the purposes for the Safety Assessment letter is to create a mechanism by which stakeholders can learn together as automated driving technologies evolve. We want to work together with NHTSA to help the letter process meet its goal. To that end, we make the following suggestions with respect to process, the letter’s applicability and scope, and its structure and content.
To achieve NHTSA’s goal of information-sharing with the public, we suggest the agency create a dedicated site. Additionally, the agency could consider ways in which to streamline the Part 512 request for confidential business treatment in order to make this voluntary process more efficient. Furthermore, we suggest NHTSA acknowledge receipt of a developer’s letter.
Applicability and Scope
The Coalition suggests that flexibility be a hallmark of the Safety Assessment letter process for three reasons. First, flexibility would reduce the burden on developers by allowing them to tailor their processes and streamline their submissions. For example, developers could use a single letter to cover features or systems that are common and migrate over multiple vehicle models. Second, flexibility would minimize the number of letter submissions during the testing phase but still provide a reasonable amount of information that could inform new submissions. More specifically, during the development and testing phase, the cadence of software updates may occur with a high degree of frequency and generate a large volume of data that may not be useful to NHTSA or other stakeholders. Third, flexibility would still recognize and permit developers to communicate the differences between SAE Level 2 and Level 4 technologies, which NHTSA staff have acknowledged publicly as an agency goal. We believe that the communication and treatment of systems needs to be different, since they are different, as is the way people interact with them.
Structure and Content of Assessment Letter
We recommend that developers’ letters should provide brief – yet sufficient – descriptions of how certain guidance elements in aggregate provide reasonable, practicable, and appropriate means by which to address the safety of an automated vehicle, system, or feature for applicable performance guidance areas. We submit that the Safety Assessment letter’s structure begin with a cover page, signed by a single authorized company official, that provides a description of what the letter is being submitted for (e.g. vehicle, system, or feature); the SAE J3016 automation level classification(s) of the vehicle(s), system(s), or feature(s); whether the letter is being submitted for testing or deployment; and a brief explanation of the automated application. The cover page would then be followed by the performance guidance areas outlined in Section One of the FAVP describing the processes used and listing supporting information. Furthermore, we recommend the agency consider changing the wording of the first choice available for each performance guidance area (i.e., “meets the guidance area”) to “product development effort for this feature is consistent with the guidance area.” Such a rewording provides clarification that in the instance there are no test protocols defined, the developer nevertheless has procedures in place that align with NHTSA’s intent in providing a given performance guidance area.
II.Model State Policy
The agency’s Model State Policy provides an important delineation of the traditional boundaries between federal and state jurisdiction in matters related to automobile policy. However, while initially focusing on areas of traditional state jurisdiction (such as licensing, traffic enforcement, and insurance), the Model State Policy also provides leeway for states to fill in gaps and build their own regulatory framework for HAVs. While this may not be NHTSA’s intention, the proposed Model State Policy does present this risk. For example, suggesting that states form a jurisdictional automated safety technology committee will naturally and inexorably lead state officials to venture beyond the strict state functions and recommend regulatory approaches to deal with motor vehicle safety issues within NHTSA’s domain.
NHTSA must play an important role in urging states and local entities not to rush into legislating simply because the subject matter is new and novel. The Coalition calls upon NHTSA to leverage its existing position as the federal vehicle safety authority to safeguard against overlapping regulation by state and local governments.
We further note that doing so would not only help the industry develop and mature to maximize the benefits to the consumer, but also would further NHTSA’s safety objectives. Should states and local governments move to enact disparate regulatory frameworks, it would reduce the ability of NHTSA to shape and influence the processes for safety assurance moving forward.
III.Only A Few of the “Modern Regulatory Tools” as Proposed in the Policy Would Facilitate the Safe and Rapid Deployment of HAVs but NHTSA Should Actively Pursue Them
The Coalition believes that the emergence of HAVs requires NHTSA, the industry and other stakeholders, to develop new approaches to solving the complex issues that accompany the design, testing, and deployment of HAVs. In this regard, the Coalition agrees with NHTSA that the goal should be “to ensure that premature, static regulatory requirements do not hinder innovation and diffusion of the dynamic technologies that are being developed in the industry.” However, we submit that some of the proposed new regulatory authorities and tools in section IV of the Policy run counter to this objective. Among other drawbacks, some of these new authorities and tools risk imposing prohibitive costs on manufacturing of self-driving vehicles and delaying the rollout of important safety features and updates.
We summarize our comments on these proposed new authorities and tools as follows:
Expanded Exemption Authority for HAVs
The Coalition supports NHTSA’s proposal to expand the exemption authority for HAVs to enable manufacturers and technology companies to expand and improve the pathway to test and ultimately deploy. As NHTSA has accurately stated, “One option that could facilitate the safe testing and introduction of HAVs would be to expand the Agency’s existing exemption authority. Current authority permits NHTSA to exempt not more than 2,500 vehicles per year for a two-year period, on the basis of equivalent safety.” Existing authority concerning “general exemptions” (49 USC 30113) provides some leeway for development and field evaluation of innovative features but its limitations on duration (two years) and vehicle numbers (2,500 in any 12-month period) do not provide for full deployment. 49 USC 30114 (“special exemptions”) is limited to research, investigations, demonstrations, training, competitive racing events, show, or display. The recently enacted section 49 USC 30112(b)(10) permits introduction of vehicles into commerce that do not comply with the FMVSS “solely for the purposes of testing or evaluation.” These limitations are simply insufficient to achieve the goal of rapid, safe deployment and to attain the real benefits to consumers described above.
To increase the exemption cap in a significant manner, the Coalition supports eliminating or raising the exemption cap to a level that will help facilitate meaningful commercial deployment of HAVs. To achieve this, the Coalition supports NHTSA’s proposal to be provided new authority to “grant incrementally increasing exemptions to the same manufacturer, progressively relaxing the numerical limits on annual production volume and exemption duration over time, or even eliminating those limits altogether (following an incremental one-step-at-a-time approach).” We strongly believe that Congress should act to provide this authority to the agency immediately, to fully enable the potentially life-saving safety innovations of L4 systems and mitigate the safety risks of legacy controls unnecessary for such systems. This new authority would allow the deployment of innovative safety technologies that meet or exceed the level of safety required by existing federal standards, while ensuring a prompt and transparent process. The Coalition does not see expanded exemptions as a long-term replacement for industry-wide standards. Rather, we see such exemptions as a necessary measure to ensure that safety innovations do not have to await completion of rulemaking actions, which can consume several years.
If granted by Congress, this new authority would supplement existing NHTSA authorities, to expedite the safe introduction of life-saving HAVs. Historically, many proven technological advances have been precluded from introduction in the United States by the current FMVSS. Current authority is simply insufficient to keep pace with safety technologies developed by the automotive industry overall. Among those technologies are adaptive beam headlights and side mirror-replacing sensors. The same limitations on NHTSA’s authority are creating significant impediments to the design, development, and introduction of fully self-driving L4 vehicles.
NHTSA Rulemaking to Amend Existing FMVSS for L4-L5 Vehicle Design
The Coalition submits that the voluntary approach envisioned in Section I of the FAV already creates a voluntary notification process to explain the safety basis for existing and planned technologies, and supplementing it with further HAV-specific processes would inhibit rather than facilitate innovation for this technology. However, there are a number of new areas that the Coalition strongly welcomes in this policy, which warrant significant attention, including (i) the agency’s efforts to improve existing interpretation, exemption, and rulemaking processes, and (ii) recognition by NHTSA that the current exemption process is insufficient to enable the full-scale commercial deployment of vehicles designed from the ground up to be fully self-driving and that additional steps need to be taken by NHTSA (e.g. targeted rulemaking) and Congress (e.g. revised NHTSA exemption authority) to facilitate the safe, large-scale deployment of fully self-driving vehicles in the United States. The remainder of the comments will describe how the Coalition believes that NHTSA and Congress can move forward to jumpstart these processes over the next year, beyond the voluntary Guidance and limits the industry faces under the current regulations.
NHTSA proposes to use its existing statutory authority to implement safety assurance tools that would include requiring manufacturers “to provide the agency with advance information and reporting about their efforts to ensure safe introduction of complex safety systems and HAVs, through systematic risk analysis, identification, classification, and reduction.” The purpose would be to ensure that manufacturers can demonstrate that they have “applied the NHTSA performance guidance, industry best practices, and other performance criteria and standards to assure the safe operation of motor vehicles.”
The Coalition acknowledges NHTSA’s efforts to require detailed assurance from industry stakeholders on their implementation of the Vehicle Performance Guidance and other standards and practices. However, given the Coalition’s concerns about the Guidance that forms the foundation for this proposed new tool, the Coalition cannot endorse this approach. More fundamentally, the safety assurance tool proposed in the Policy duplicates what is already in place for HAV manufacturers, who must demonstrate compliance with the Federal Motor Vehicle Safety Standards (“FMVSS”)—and expands the scope of the Vehicle Performance Guidance beyond what is authorized within the context of voluntary guidance. Additionally, mandating the Safety Assessment by rule cuts against current NHTSA practice of gathering impact analysis and data to support effectiveness, costs and benefits. Given the Safety Assessment letter process outlined in Section One of the Policy, the Coalition would suggest not proceeding with a rulemaking process to mandate Safety Assessment submissions.
Pre-Market Approval Authority for HAVs
Pre-market approval would take the United States in a fundamentally different direction and away from the self-certification approach that has been the hallmark and foundation of automobile regulations in this country for the past several decades. As such, the Coalition opposes any proposal to obtain pre-market approval authority for the deployment of HAVs. As the agency states, this authority would be “a substantially different regulatory approach than the self-certification approach established by Congress and used by NHTSA today.” The automobile industry has successfully and effectively incorporated new technologies and leveraged the flexibility of the self-certification approach to innovate in the area of safety for decades without the federal government demanding prior approval in the marketplace.
We further question the proposition that “public acceptance of and confidence in HAVs” would be meaningfully enhanced through the “affirmative approval by the federal government.” The self-certification of compliance with FMVSS, upon which manufacturers and regulators have relied for decades, coupled with NHTSA’s vigorous enforcement to address unreasonable risks to safety, has provided robust assurance to consumers and has instilled confidence in the safety and reliability of automobiles. More simply put, this is not a broken system. These concerns point strongly against instituting “a wholesale structural change in the way NHTSA regulates motor vehicle safety [that] would require both fundamental statutory changes and a large increase in agency resources.”
Post-sale Authority to Regulate Software Changes
In its Enforcement Guidance Bulletin published on September 23, 2016, NHTSA addressed software changes in the context of the agency’s existing enforcement authority by stating that “if software (whether or not it purports to have a safety-related purpose) creates or introduces an unreasonable safety risk to motor vehicle systems, then that safety risk constitutes a defect compelling a recall.” Thus, NHTSA’s current enforcement interpretation expressly eschews the need for additional authority.
We also note that the proposal in the Policy does not address what types of software changes are the target of the Agency’s focus. To be sure, the proposal ties the software to HAV capability, but does not distinguish between safety-related updates and all other forms of updates. These other updates may include those related to infotainment systems and telematics. The Coalition notes that any proposal to address authority regarding software changes, to the extent not already covered by existing authorities, should be tailored to the provision of critical safety updates rather than all software updates in general.
Variable Test Procedure to Ensure Behavioral Competence and Avoid Test Gaming
The Coalition recognizes the need for NHTSA to have flexibility in testing vehicles to account for innovation and the agency already states that it “believes it already has this authority.” However, the Coalition stresses that manufacturers need to be able to understand compliance requirements and to rely on test procedures that are objective and repeatable.
Functional and System Safety Authority
As a means of “bring[ing] NHTSA’s practices more into line with those other agencies (e.g., Federal Aviation Administration, Food and Drug Administration, Federal Railroad Administration),” the agency proposes to [presumably through rulemaking] “require manufacturers to report “serious risks” identified during the manufacturer’s Functional Safety analysis,” and may also require “manufacturers to modify their designs as necessary to reduce high-level risks to acceptable levels.”
The Coalition is concerned by these suggestions to the extent that the Agency might require manufacturers to modify their designs. Requiring reporting of “serious risks” identified through system safety processes and expecting NHTSA to require specific design modifications is not reasonable. A sound process that identifies such a risk would also result in mitigation of the risk. Moreover, having NHTSA require design modifications to mitigate specific risks would put NHTSA in the midst of ongoing system design and development, which is untenable for the agency and the industry. Any identified but unmitigated safety risk that rises to the level of an unreasonable risk to safety in the real world (i.e., after the design stage) may already have to be reported to NHTSA as a safety defect, so existing law suffices.
Regular Reviews for Making Agency Testing Protocols Iterative and Forward Looking
NHTSA proposes “an iterative and forward-looking process” for establishing and updating the FMVSS and other testing protocols for HAVs, in light of the rapid evolution of technologies in the space. In particular, NHTSA suggests conducting periodic analyses of how provisions are impacting innovation and the extent to which the FMVSS continues to be technology neutral, as well as using “sunset” clauses for certain provisions to ensure regular review and update.
We support the agency’s performing an analysis of the effects detailed test procedures for HAVs may have on innovation and conducting regular reviews of the agency’s test procedures to ensure they keep pace with innovation. On the specific proposals set forth in the Policy in this regard, the Coalition urges NHTSA to couple any proposed sunset clauses with periodic reviews of effectiveness.
Additional recordkeeping and reporting
We appreciate the opportunity to provide comments on NHTSA’s Federal Automated Vehicles Policy. The Coalition plans to provide further comments and specific proposals to assist NHTSA in developing and advancing the safe deployment of fully self driving cars in the United States and foster a consistent national framework for policy governing the industry. Should you have any questions regarding any of the comments above, please contact me at (202) 344-4747 or via email at firstname.lastname@example.org.
Hon. David L. Strickland
Counsel to Self-Driving Coalition for Safer Streets