National Academies Press: OpenBook
« Previous: Chapter 5. State Legal and Regulatory Audit
Page 29
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 29
Page 30
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 30
Page 31
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 31
Page 32
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 32
Page 33
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 33
Page 34
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 34
Page 35
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 35
Page 36
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 36
Page 37
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 37
Page 38
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 38
Page 39
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 39
Page 40
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 40
Page 41
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 41
Page 42
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 42
Page 43
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 43
Page 44
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 44
Page 45
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 45
Page 46
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 46
Page 47
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 47
Page 48
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 48
Page 49
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 49
Page 50
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 50
Page 51
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 51
Page 52
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 52
Page 53
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 53
Page 54
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 54
Page 55
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 55
Page 56
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 56
Page 57
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 57
Page 58
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 58
Page 59
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 59
Page 60
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 60
Page 61
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 61
Page 62
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 62
Page 63
Suggested Citation:"Chapter 6. Prioritization and Harmonization Analysis." National Academies of Sciences, Engineering, and Medicine. 2018. Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan. Washington, DC: The National Academies Press. doi: 10.17226/25291.
×
Page 63

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 29 Prioritization and Harmonization Analysis 6.1 Overview A prioritization and harmonization analysis was conducted to 1) identify how and when regulations and laws will need to be modified to facilitate the implementation of C/ADS-equipped vehicles, and 2) identify areas in today's MVCs and regulations where harmonization may be important in supporting the deployment of C/ADS-equipped vehicles across the U.S. An understanding of commercial C/ADS application availability and market penetration as well as results from stakeholder outreach efforts and a legal and regulatory review were considered when developing the recommendations regarding potential laws and areas of law that may require modification across different time periods. Understanding Opportunities for Legal and Regulatory Modifications State legislatures, Governor’s offices, legal practitioners, and industry sectors engaged in C/ADS deployment recognize that current laws and regulations must be addressed in a comprehensive, yet flexible way to ensure safety and reap the anticipated societal benefits of C/ADSs, while simultaneously anticipating many unknowns. Unlike implementing traditional legal and regulatory changes, making simple citation modifications to add a new title brand or type of license plate, or instituting an adjustment in fees, fines, or driver sanctions, the changes to C/ADS-related laws are complicated and challenging. Modifying these laws will require changes to basic underpinning concepts and definitions, as well as an understanding of the C/ADS technology and its limitations Further, there exist a number of situations in which states, through their motor vehicle agency or DOT, may find their needs and those of industry best served by syncing their efforts in order to harmonize key regulatory or legislative provisions related to C/ADSs. There may also be areas in today’s vehicle codes and regulations where it may be important to harmonize for purposes of supporting the deployment of C/ADSs. In this respect, we assess the advantages, disadvantages, and practicality of harmonizing the approaches to these types of laws and regulations across the country. Findings (as noted throughout this document) suggest the need for states to get as "centric" as possible to the intersecting realities of technological imperatives, public policy, and stakeholder interests. The point where these realities intersect represents opportunities for legal and regulatory modifications are illustrated in Figure 5. Ideally, state legislative and regulatory schemes would fit neatly within the circles where public policy, technological imperatives, and stakeholder interests all intersect. However, it seems probable that the reality of legislation and regulation will be messier, straying in one direction or another on a given issue or at a particular moment in time.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 30 When assessing law and regulation change, the closer a legislative and regulatory approach comes to the theoretical center where technology, stakeholders, and current laws, regulations, and policies meet, the more likely that change is to meet public policy objectives, avoid or withstand external challenges, and meet the needs of a complex, evolving technology. Figure 5. Visual depiction of the centric approach to prioritization efforts. 6.2 Overarching Considerations Regarding Harmonization Individual state laws and regulations for C/ADS are vital, but equally important is ensuring that vehicles can operate seamlessly across state lines. Ensuring state laws do not substantially conflict can occur through a process known as harmonization, which is the process of minimizing redundant or conflicting standards that may have evolved independently (Pelkmans, 1987). Harmonization can create consistency of laws, regulations, standards, and practices, so that the same rules will apply across jurisdictional boarders. Harmonization is agnostic to the specific policy approach, and can mean moving all jurisdictions’ standards to the strictest standard, most permissive standard, or to a midpoint between the two poles or to a more uniform standard. Harmonized state regulations for C/ADSs offer a number of critical advantages over a diverse and discordant set of state regulations. However, while harmonization has some advantages, there are some drawbacks, possible alternative strategies, and several areas (identified in Serian et al. 2018) that do not need harmonization for the safe deployment and operation of C/ADSs. The following subsections explore harmonization of MVCs and other rules governing C/ADSs, considering benefits, drawbacks, examples, and alternatives to harmonization. Potential Benefits Associated with Harmonization As with most regulatory discussions revolving around C/ADSs, the primary rationale for harmonizing state MVCs is to make the vehicles and the roadways safer for all users, improve enforcement, and increase driver awareness and understanding from state-to-state. Furthermore, the potential for substantial reductions in motor vehicle crashes, fatalities, and injuries serves as the driving force behind the generally supportive posture the public sector has taken towards C/ADSs (Strickland, 2013). At the same time, states are motivated to remove any barriers to the operation of safe C/ADSs to encourage private sector investment and research into this technology. Consistent, clear, and predictable rules across disparate

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 31 political entities can reduce complexity for manufacturing and product development, potentially reducing costs to industry and consumers alike. Harmonization therefore touches issues that affect state policy makers and agencies, industry, as well as the traveling public. The business case for C/ADSs depends, in part, on state harmonization of MVCs. To the degree that similar regulations are in place across many or all states (for example, liability), some introductions of C/ADS technologies may be accelerated, increasing the market size. A-MaaS use cases, for example, rely on a “network effect” to function effectively (Rogers, 2016). Services like A-MaaS must build out a robust network of users to function, and states with consistent rules make it easier for companies to build a user base across state lines. This has been clearly demonstrated with the deployment of Transportation Network Companies from state to state and from city to city. The business case for C/ADSs in CMV operations can also depend on harmonized state MVCs and regulations. For example, many goods reach markets via CMVs, and ensuring their safe and efficient operation is important for other road users, the companies receiving and shipping the goods, and the users of the goods. Harmonized state MVCs can reduce the regulatory burden on CMV operators transporting goods across state lines, and may also be relevant as platooning CMVs enter the transportation network. From a broader economic perspective, harmonization offers a seamless regime that facilitates the free flow of goods and commerce across borders without regulatory barriers. This application has been demonstrated through such applications of the International Registration Plan. When there are disparate approaches to regulating specific aspects of vehicle equipment, manufacturing costs may rise (Canis and Lattanzio, 2014). Harmonization at the state and local levels can reduce these costs, which in turn is likely to benefit consumers through lower cost products and a more efficient technology rollout. This efficiency stems from the regulatory consistency that is provided for manufacturers of vehicles, as well as system developers regarding uniform vehicle functionality, external markings, and several other areas that could help reduce development and implementation complexity and accelerate market introductions. Another motivation for harmonization is facilitating data sharing between states, such as improving interstate consistency for vehicle records designations (e.g., titling and the National Motor Vehicle Title Information System and the Commercial Driver License System). Much of this type of harmonization is already underway. Note that when it comes to harmonization of areas such as liability and insurance requirements, this may need to be approached based on the type of consumer ADS application. Within this analysis, harmonization considerations are presented in in terms of consumer ADS applications (C/ADS-equipped passenger vehicles and A-MaaS; see Chapter 2 for definitions). Potential Drawbacks of Harmonization While harmonization offers many advantages for states, industry, and the traveling public, there are also downsides that bear consideration. Perhaps the most obvious concern a state might have regarding harmonization is a loss of control over policy decisions. Under the Tenth Amendment to the Constitution, states maintain powers not designated to the federal government (U.S. Const. amend. X).

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 32 Harmonization can reduce complexity by aligning a disparate set of rules or regulations. However, doing so can also cause problems if the harmonized rules are inappropriate, premature, or have other flaws. If an industry or product is not technically mature, for example, establishing rules prematurely could hamper innovation by restricting the product or industry in ways that may not improve safety or efficiency. If states harmonized to a premature or inappropriate set of rules, the problematic rule would shift from being a local or regional problem to being a national problem. For this reason, those involved in the harmonization process should carefully consider the costs and benefits to all stakeholders and interests, including manufacturers and the traveling public, both with regards to safety and cost. Beyond overt disadvantages, there are also basic questions of necessity when considering harmonization across all states. Harmonization is not a simple or speedy process, as demonstrated by the Driver License Compact, and typically takes several years to complete. Due to the slowness and difficulty of harmonizing regulations, states may wish to avoid harmonization for areas that are neither necessary nor beneficial. For issues that do not expressly need state harmonization, each state could ensure their laws and/or regulations are clear, in particular as they relate to core terms and definitions (e.g., drive, driver, operate). Additionally, states could require all vehicles follow the rules of the road, regardless of whether a human driver or the ADS performs the DDT. However, flexibility in interpretation and execution of the rules of the road is also important, in that it enables vehicles—both ADS and human-operated—to make reasonable decisions in a complex and dynamic operating environment. Harmonization may be unnecessary when regulating certain aspects of C/ADSs, as some manufacturers may develop the ability to automatically adjust to new parameters and requirements whenever crossing state lines or any jurisdictional boundaries. As an example, A-MaaS operations areas may gradually expand as regulatory regimes evolve. In much the same way that C/ADSs will need to adjust automatically to each roadway’s speed limit, developers could potentially include automatic adjustments for other non-hardware related regulatory provisions. Upward Versus Downward Harmonization As noted earlier, the simple act of harmonization should not be the end goal for states. There is a critical difference between upward and downward harmonization. In upward harmonization, states with less stringent regulatory regimes harmonize with states that have more stringent requirements. In downward harmonization, the equilibrium for agreed upon standards is driven to the lowest common denominator by the states with the most lenient regulations and requirements. Recent experience in the U.S. demonstrates the potential for downward harmonization. States briefly suspended regulatory efforts in 2015 after several years of various states passing C/ADS testing regulations. The pause in laws occurred when the industry responded to new C/ADS testing regulations by moving their testing and development work out of states with testing laws, such as California and Florida, to states without testing regulations that did not require testing companies to register or seek permission prior to conducting public testing. As such, upward harmonization toward stricter standards solely for the sake of stricter standards should not necessarily be viewed as the states’ objective. The

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 33 potential risk of downward harmonization, however, toward simply achieving the lowest common denominator, is likely to produce a regulatory regime that serves industry’s, rather than the traveling public’s, interests. As harmonization activities proceed, these factors must be carefully weighed. Implications of Not Harmonizing Much of the above discussion has centered on the benefits that come from states harmonizing their MVCs and standards for C/ADS, but it is equally important to note the risks for industry and the broader public if such harmonization, where needed, does not occur. Without harmonization, industry faces what has been termed a “patchwork quilt” of state regulations. The risk associated with the failure to harmonize standards when needed is that industry testing and eventual deployment of C/ADS will be unnecessarily delayed and/or potentially not broadly available, the safety benefits of C/ADSs will be lost, and the lives saved from roadway fatalities will be lessened. Regulatory inconsistencies will create roadblocks for C/ADS testing and operation if the differences include conflicting standards or excessive overlapping of state, county, regional, and local requirements. This risk is associated only with not harmonizing those areas we have recommended for harmonization. For areas identified as not requiring harmonization, the impact on industry is deemed much less detrimental with regard to testing, development, and deployment of C/ADSs. This is a result of the current lack of harmonization in these areas across states. We believe industry can comply with any existing related provisions without impacting their timelines for C/ADS deployment or the manufacturing process. For hardware, design, and other fixed vehicle attributes, harmonization through federal preemption is likely necessary. Standards that address or prohibit allowable hardware, hardware mounting, or other hardware aspects essential to level 4–5 ADS-equipped vehicles could prove prohibitive to their interstate operation, thus conceivably blocking market introduction. Alternatives to Harmonization Reciprocity Agreements Harmonization is not the sole mechanism for achieving seamless authority to operate a motor vehicle across state borders with different laws, regulations, standards, and practices. Reciprocity agreements represent a highly effective, and frequently utilized, alternative in which states sign agreements to mutually recognize each other’s licenses, share driver infraction records, or any other provision of vehicle codes that are necessary for operating vehicles in other states. Three examples of existing interstate compacts include 1) The Driver License Agreement, 2) the Driver License Compact, and 3) the Non-Resident Violator Compact (AAMVA, n.d.; National Center for Interstate Compacts, n.d.a; National Center for Interstate Compacts, n.d.b). Under these arrangements, states agree to mutually honor licenses issued by other states, share information regarding out-of-state infractions and driver safety, and allow for the processing of traffic citations across state borders. Most states take part in these arrangements: 44 out of 50 (88%) participate in the Non-Resident Violator

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 34 Compact, and 45 of 50 (90%) participate in the Driver License Agreement, for example. These agreements ease interstate travel without the harmonization of state vehicle codes or licensing requirements. The process for a state to join a reciprocity agreement or interstate compact such as these can involve state legislatures ratifying the agreement and the passage of legislation to codify the content of the reciprocity agreement. This process can be lengthy, and historically, some states have taken years to sign these compacts. An example candidate for reciprocity could include permit requirements for platooning level 1 driving automation system-equipped vehicles. However, with the recognition of the obstacles to interstate compact agreements, this is not a preferred recommended direction. Model Law / Best Practice Language Another approach that can make the harmonization process much easier is drafting model legislation or best practice language. A model law is a complete piece of legislation, pre-made and ready for introduction into the legislative process. Best practice language is less formal, and features pre-crafted tenets or aspects of legislation that lawmakers can adopt into law. This approach is effective because it enables unfamiliar legislators to start from a prepared text rather than requiring each individual state, and each legislative committee within each state, to craft original language. In discussions with state administrators, we found that stakeholders view model legislation as a solid foundation for state laws and cross-jurisdictional reciprocity. Similarly, state DOTs echoed the importance of uniform regulatory efforts related to infrastructure improvements, such as interconnected traffic signals and other V2I systems for CVs. Model laws or best practice language can be useful to lawmakers who are unfamiliar with C/ADS technology or who do not have the time to become experts in the intricacies of the technology or its full array of policy implications. However, while model laws can be helpful, they are far from perfect, as there are rarely issues as complex as those surrounding C/ADSs that can be addressed within the text of a singular, universal bill. Additionally, variation among existing state codes can make it difficult to apply model laws, as doing so would require closely scrutinizing existing code and replacing it with model language. As such, it is helpful for those crafting model laws to focus on providing model language primarily for the most important policy aspects of this topic. States, working with their legislatures, can then utilize the model language within the content of a bill, drafted according to their state’s statutes, codes, and regulatory agency structures. Associations such as AAMVA and AASHTO and the Uniform Law Commission are important resources and ideally should (with their members’ guidance and involvement) drive model law considerations. History teaches a lesson here as well. Model laws are difficult to get legislatively enacted in some states and can take years to achieve. The legislative process is complicated. Bills often change substantially from introduction to acceptance, which can reduce the effectiveness of model legislation. For these reasons, based on our interviews, many industry representatives favor the less prescriptive best practice language approach over that of model law when seeking harmonization; this is a lesson they learned through experience during past attempts to enact model legislation across the states.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 35 Guidelines Lastly, states should consider the use of less prescriptive and more flexible guidelines at the state level. NHTSA used this mechanism with the issuance of non-binding Federal Automated Vehicle Guidelines in 2013, again in 2016, and further updated under the current DOT leadership (NHTSA 2013, 2016, 2017). Based on our stakeholder feedback, recent experiences in the C/ADS regulatory area have led many to believe that it is likely too hard to “go straight to regulations” and that an approach centered on the issuance of clear guidelines may prove to be better. When responding to questionnaires, several stakeholders, including manufacturers and technology companies, cited the State of California’s in-depth regulations for the testing and deployment of C/ADSs as a reason for moving their testing operations to other, less prescriptive states. Guidelines addressing key areas of needed legal changes, such as definitions, developed at the federal level with input from state- and local-level government entities and other stakeholders (e.g., national and state associations, citizen groups) can provide state policy makers with a starting point for their individual efforts. Further, guidelines may serve as a catalyst for harmonization across states. Federal Preemption The impetus behind federal preemption has historically been avoiding a patchwork of state standards and regulations. The federal government has strongly indicated it will address standards for C/ADS hardware and software such that state level regulations (or harmonization) will not be necessary, and that federal laws and standards could preempt state policy (NHTSA, 2016, 2017). Under this scenario, the harmonization of certain aspects of C/ADS regulations and standards would become moot as NHTSA (or FMCSA for motor carriers) would set standards preventing states from enacting standards that differ from federal rules. Enforcement of federal standards would largely depend on the specifics of the policy in question; currently, many automotive safety and environmental regulations are enforced via self- certification, and this seems a likely avenue for future enforcement mechanisms (Canis and Lattanzio, 2014). As noted in the previous chapter, with the unanimous passage of the SELF DRIVE Act of 2017 by the House of Representatives, Congress has already begun to consider preemption of state and local regulations and laws pertaining to hardware and design features for C/ADSs. 6.3 Prioritization and Harmonization Recommendations As noted in Chapter 3, states should examine the fundamental legal concepts and terms that undergird their motor vehicle and transportation programs, and consider how these concepts and terms will be affected by the use of C/ADSs. This is a time to assess and then act, rather than reacting or acting for the sake of action. These latter approaches result in the crafting of piecemeal revisions and solutions to legal impediments and gaps as they arise, but fail to recognize the more strategic implications of specific law changes that go deeper than one piece of a state motor vehicle code. During the assessment process, states are advised to refer to NHTSA's guidance (2017) which encourages states "to allow DOT alone to regulate the safety design and performance aspects of ADS technology. If a State does pursue ADS performance-related regulations, the State should consult with NHTSA" (2017, p. 20). NHTSA notes that

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 36 "allowing NHTSA alone to regulate the safety design and performance aspects of ADS technology will help avoid conflicting federal and state laws and regulations that could impede deployment” (2017, p.18). Recommendations included within this chapter are based on past experience with C/ADS-related research, input from stakeholders, and the anticipated timeline of level 4-5 ADS-equipped vehicle deployment (Figure 3). All changes to laws and regulations should be girded by the state’s goals and objectives surrounding level 4-5 ADS-equipped vehicle deployment and an individual state MVC assessment. For additional reasoning behind these recommendations, see Chapters 4 and 5 of Serian et al. (2018). Consumer C/ADS Application Harmonization Considerations Just as there are varying levels of urgency for the recommended modifications and reviews that states should take to remove barriers to the operation of C/ADS-equipped vehicles on the roads, the same tiered urgency applies to the question of whether or when states should harmonize their vehicle codes to provide consistent standards across state borders. The importance of and need for harmonization is also often determined by the path to deployment of the various consumer C/ADS applications. C/ADS-Equipped Passenger Vehicle Harmonization Needs Harmonized regulations applying to personal C/ADSs will be essential to the long-term development of automated mobility. Some of these will be addressed via efforts to enable A-MaaS in the 2020 timeframe, with broad availability of personal C/ADSs coming shortly thereafter. In fact, initial personal level 3 ADS-equipped vehicles are now entering the market in Europe. The European model year 2018 Audi A8, for example, offers a level 3 ADS traffic jam assist function (Davies, 2018). Platooning C/ADS-Equipped CMV Harmonization Needs Unlike personal C/ADSs and A-MaaS, truck platooning may require changes to state MVCs, and truck platooning will eventually evolve from first generation SAE level 1 driving automation systems (with full driver involvement in steering and monitoring the road environment) to level 4 follower ADS-equipped vehicles. This harmonization analysis focuses on level 1 platooning where there are relatively few areas affecting initial deployment, but the areas that do exist are essential to opening the way for this important safety and energy conservation application. The primary area is creating appropriate following distance rules for automated longitudinal control rather than human control of brakes and throttle.9 A-MaaS Harmonization Needs A-MaaS involves the use of level 4 or 5 ADS-equipped vehicles offered for commercial transportation on-demand. This will function similarly to today’s transportation network companies, such as Uber and 9 Consistent with North American Council for Freight Efficiency's Confidence Report: Two-Truck Platooning (NACFE; 2017), this harmonization analysis focuses on "near-term two-truck platooning opportunities where human drivers have their hands on the wheel and the technology is assisting the driver in getting better fuel economy by reducing aerodynamic drag through safely following another vehicle at shorter distances than an unassisted driver could maintain" (p. 3).

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 37 Lyft, but without drivers. Several unique aspects of this service model will require additional modifications to state MVCs. For A-MaaS, aspects of inspection and vehicle maintenance will fall to the fleet operator, which may choose to adopt very high maintenance standards for purposes of quality control and ensuring high customer service levels. States will need to take care to avoid a situation where low quality A-MaaS services are able to shop around to find the state with the fewest inspection requirements and then operate their entire fleets under that state’s flag. At the same time, many major cities that are likely to be the primary focus for A-MaaS operations are part of larger, multi-state metropolitan areas, such as the New York City and New Jersey/Connecticut area or Washington, DC and its Maryland and Virginia suburbs. A-MaaS vehicles will need to operate seamlessly across such borders, even when not carrying passengers or safety drivers. Short-Term (2018–2020) Priorities Short-Term Recommendation 1. User Requirements Definitions and Driver Only Vehicle Codes States, if they have not yet done so, should review the fundamental terms “drive,” “driver,” “operate,” and “operator,” as well as any wording that arguably omits any restrictions on C/ADS-equipped vehicles. Any ambiguous terms should be clarified to provide consistency and reduce ambiguity. Additionally, policy makers should directly address the possibility that their vehicle codes can be interpreted to regulate only “drivers” (who are licensed and hence human), thus potentially exempting from legal oversight level 4–5 C/ADS-equipped dedicated vehicles (C/ADS-DVs) where the ADS is, in effect, the “driver.” Prioritization Reasoning When code terms are ambiguous, they should be accompanied by the clarification of intent and interpretation to provide consistency and clarity. This will also allow DMVs, DOTs and law enforcement agencies that interact with level 4–5 C/ADS-equipped vehicles, whether through licensing and regulation, provision of infrastructure, or law enforcement, interpretive guidance, to assist in developing new programs. While many states also have begun to develop definitions for different types and levels of C/ADS- equipped vehicles, this is an area that is clearly in need of consistent definitions, best practice definitions, or federal intervention. Although many states are using NHTSA's Federal Automated Vehicle Policy (2016) and subsequent A Vision for Safety 2.0 (2017), there are a number of instances in which states have developed their own definitions. In the past year alone, a wide range of varying state definitions have been legislated. While many of the definitions are similar, they are not sufficiently consistent and represent a clear threat to harmonization, state-to-state reciprocity, longer term vehicle titling and identification issues, manufacturer and technology company design and sales concerns, as well as A- MaaS launch and deployment. Moving forward, states should be mindful of a current NHTSA project—Assessment, Evaluation, and Approaches to Technical Translations of FMVSS and Test Procedures That May Impact Compliance of

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 38 Innovative New Vehicle Designs Associated with Automated Driving Systems—which is considering these definitional issues as well as issues associated with vehicle requirements. Harmonization Reasoning The terminology currently used in most state MVCs was recorded during a time when ADSs, or any computer-enhanced operations of a motor vehicle for that matter, were unheard of and technologically infeasible. If harmonization of state MVCs does not occur before other legal modifications, states may create conflicting, overlapping, and misaligned terminology and definitions for common terms that will make national standards even more difficult to construct. As noted above, conflicting definitions are beginning to emerge. Uniform definitions or best practices terminology, developed by either NHTSA or AAMVA, would greatly benefit the harmonization of these definitions across these fundamental aspects of the driving code. Harmonization for A-MaaS The impact of outdated and/or conflicting definitions on A-MaaS applications should be a focus for harmonization, as the A-MaaS business model is dependent on states harmonizing a number of changes to their MVCs. This harmonization is anticipated to take place from 2018–2020, likely before C/ADS- equipped passenger vehicle applications reach the market. A-MaaS allows companies to spread out the higher cost of purchasing a C/ADS-equipped vehicle across many users and to collect data to inform development and improvements in future systems. The harmonization recommendations noted above in terms of definitional clarifications apply strongly to A-MaaS as well. The growth of the A-MaaS market depends on clarifying that A-MaaS customers are passengers (as opposed to drivers) and have no responsibility for the safe or lawful operation of the ADS-equipped vehicle in which they are riding.  Prioritization Recommendation: In developing laws, regulations, and policies, ensure that the included definitions, especially those for level 3–5 C/ADS-equipped vehicles, are consistent with the SAE framework (as reflected in NHTSA, 2017). Harmonization Recommended: Yes. Best practices language. Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A- MaaS.  Prioritization Recommendation: States should consider working with NHTSA and AAMVA to develop uniform definitions for the terms “driver,” “operate,” and “operator” as best practices or uniform definitions. This effort should recognize the possibility that state MVCs can be interpreted to regulate only “drivers” (who are licensed and hence human) and thus effectively exempting from legal oversight level 4–5 C/ADS-equipped vehicles (where the C/ADS, when engaged, could be considered the “driver”). Harmonization Recommended: Yes. Guidelines (policy decision).

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 39 Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A- MaaS. Short-Term Recommendation 2: Truck Platooning-Related Code Provisions Policy makers should audit their highway/transportation state laws and regulations to identify those areas that impact the use of truck platoons. Areas possibly requiring modification include lane restrictions, service requirements, size, weight, and following distance—following distance being a frequently addressed area (Serian et al., 2017). Policy makers should also determine if local governments will have the ability to regulate truck platoons in ways that differ from the rest of the state. A number of stakeholders noted the need to address state regulations related to move over laws, following distance and tailgating, passing of other vehicles, convoys, and vehicle size and weight laws. However, some stakeholders saw no need for any legal or regulatory changes. As highlighted in Wagner et al. (2018), perspectives vary based on how jurisdictions interpret their current law and regulation wording. As with operator and driver definitions, some states have begun to define what a truck platoon is, where it can operate, and how approvals will be considered. Prioritization Reasoning Most states within our sample impose following distance requirements that are incompatible with the deployment of truck platoons. If truck platoons are to be encouraged on a state’s highways, modification of these following distance requirements will likely be necessary. Within state DMVs and DOTs, truck platoons are generally treated as groups of individual trucks, sometimes operating as a caravan. However, within the language of existing state codes, other terms theoretically could be interpreted to cover truck platoons. To take one example, in the UVC and many states codes, there is a repeated reference to a “combination of vehicles,” a term that is almost never defined. While there appears to be strong consensus that the term does not cover truck platoons, the law itself is somewhat ambiguous on this point. Indeed, because of this ambiguity, one state (Michigan) clarified in 2016 language that a truck platoon is not considered a “combination of vehicles” [See MI § 643a(10)]. A clearer definition of truck platoon will then help pave the way for other needed legal adjustments to the regulation of platoons. Since trucks in platoons operate independently but in relative close proximity, state regulators may need to consider the aggregate length, weight, and possible noise restrictions as they apply to a set of trucks operating as a platoon, depending on the outcome of relevant engineering analyses. Consideration for modifications should also address lane restrictions (if needed) or service requirements on platoons from a harmonization perspective. Truck platoons are likely to operate across state boundaries and travel over both state and local roadways, emphasizing the need for consistency in definitions from state to state. Truck platooning (especially platooning with level 1 or level 2 driving automation system-equipped CMVs) seems to have fewer legislative barriers than other level 4–5 C/ADS-equipped vehicles. Some examples identified in Loftus-Otway and Gallun (Updated 2018) include the following.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 40 Harmonization Reasoning Truck Platooning Driving Automation System-Equipped CMVs Truck platooning is in need of short-term harmonization to amend state MVCs to remove unintentional barriers to deployment of platooning level 1 or level 2 driving automation system-equipped CMVs. The interstate nature of commercial motor carriers demands uniformity (to the degree possible) among state vehicle codes related to platooning. As such, this represents a call for states to work together to harmonize their updated regulations to permit the use of this technology seamlessly across state lines. Currently, state laws regarding truck following distances are a patchwork of numeric minimums at various distances and language such as “safe and prudent” following distances that require interpretation as to the legality of platooning. Given that platooning is expected to be commercially available in the near future, consistent treatment of platooning to create a nationwide regime that recognizes a mode of longitudinal control dependent on sensing and connected braking to maintain safe shorter distances is valuable in the short term (North American Council for Freight Efficiency, 2016). At the same time, a number of technology developers have noted that their systems will have the ability to automatically adjust to new parameters and requirements whenever crossing state lines or any jurisdictional boundaries. As previously noted, the legal classification of a truck platoon is generally not specified in state codes. The provision of additional guidance or amended laws to provide a clearer definition of the classification of truck platoons is important to avoid confusion and lend regulatory certainty to technology developers and the trucking industry. Consideration of Aggregate Length, Weight, and Noise Limits as well as Lane Restrictions for Platoons Now is the time to conduct the necessary analyses to determine if there are any negative length, weight, and/or noise effects due to heavy CMVs operating as a platoon. Any requirements stemming from these analyses would ideally be harmonized from the start. However, if any restrictions are levied on truck platoons with regard to length (number of tractor-trailer combinations in the platoon), weight, or noise, embedded software systems may be able to adjust operations as needed when crossing state lines (e.g., through the use of embedded mapping software and protocols). Or, in the case of specific bridges identified by a state as vulnerable to a platoon traversal, software could dissolve the platoon temporarily prior to crossing the bridge. Therefore, lack of harmonization with regard to operational restrictions on truck platoons should not delay the early growth of platooning deployment; however, deployment of platooning will grow more robustly if these factors are harmonized (again, if engineering analyses actually determine there are factors to be addressed). It could be argued then, that individual early- deployment states could take the lead in working out best practices, with harmonization coming in the 2020 timeframe.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 41  Prioritization Recommendation: If truck platoons are to be encouraged on a state’s highways, modification of following distance requirements will likely be necessary, particularly in states that impose prescriptive following distances.  Prioritization Recommendation: The legal classification of a truck platoon is generally not specified in state codes. Policy makers should consider providing guidance or amend laws to provide a clearer definition of the classification of truck platoons. Harmonization Recommended: Yes. Definitions surrounding truck platooning could benefit from best practices definitions and consistency from state to state. States should consider working with AAMVA, AASHTO, the Commercial Vehicle Safety Alliance, and FMCSA to develop consistent platooning recommendations. Consumer C/ADS application(s) affected first: Truck platooning driving automation system- equipped CMVs.  Prioritization Recommendation: Since trucks in platoons operate independently but in relative close proximity, state regulators may—if supported by engineering analyses—need to consider the aggregate length, weight, and possible noise restrictions as they apply to a set of trucks operating as a platoon.  Prioritization Recommendation: In addition to a vehicle code review, policy makers should audit their highway/transportation state laws and regulations to identify those that impede the benefits of truck platoons. Some areas possibly requiring modification include lane restrictions, service requirements, size and weight, and following distance, which has been addressed frequently with regard to driving automation systems. In addition, policy makers need to determine if local governments will have the ability to regulate truck platoons in ways that differ from the rest of the state. Another important consideration for policy makers is harmonization across local boundaries, state boundaries, and international boundaries so as to not impede commerce. Harmonization Recommended: Useful, but not essential. Definitions surrounding platooning could benefit from best practices definitions and consistency from state to state. States should consider working with AAMVA, AASHTO, the Commercial Vehicle Safety Alliance, and FMCSA, to develop consistent platooning recommendations. Consumer C/ADS application(s) affected first: Truck platooning driving automation system- equipped CMVs. Short-Term Recommendation 3: Vehicle Identification and Title Brands This recommendation addresses two areas highlighted by stakeholders and revealed in Wagner et al. (2018): 1) how a vehicle will be identified as a level 4–5 ADS-equipped vehicle (an OEM and NHTSA role), and 2) how vehicles may need to be branded or identified for titling and/or registration. Each state

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 42 currently has varying title brands and inconsistent definitions for brands. This is an area that should be considered in the short-term to allow states moving forward at the onset of C/ADS development to adopt consistent title and registration branding. Prioritization Reasoning As the term denotes, “brands” determine the designation of the SAE J3016 automation level associated with a driving automation system-equipped vehicle, and are associated with the vehicle identification number. These brands follow the vehicle from birth (manufacturing) to death (salvage or destruction). State law enforcement personnel stakeholders specifically noted that vehicle identification markings are a top priority and should be nationally uniform. Resource and jurisdictional stakeholders concurred that that titles (ownership documents) should be branded with an ADS code. Further, it was recommended by Serian et al. (2017) that a uniform title brand be suggested by NHTSA for use across all states. Harmonization Reasoning Each state currently has varying title brands and inconsistent definitions for brands. This is an area that should be considered in the short-term to allow states moving forward at the onset of C/ADS development to adopt consistent title and registration branding.  Prioritization Recommendation: States should consider modifying statutes that define vehicle brands based on SAE J3016 automation level, and include new title brands for level 3 and higher ADS-equipped vehicles. State accommodations for title brands should follow the SAE levels, but also make provisions for aftermarket applications and brand revisions. Harmonization Recommended: Yes. Guidelines (policy decision). Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A-MaaS. Short-Term Recommendation 4. Data Privacy and Data Security While NHTSA and other federal regulators are engaged in the various privacy issues presented by CVs, which may also be automated, states should also consider statute changes to ensure public confidence and clarity on data collection and use. Prioritization Reasoning Loftus-Otway and Gallun (Updated 2018) highlighted current federal provisions and recommendations for states to consider surrounding privacy, particularly with regard to event data recorder data and third- party access to data. This is an instructive area for states to review as they consider what types of modifications may be necessary and the expected role of the federal government in this area. Level 3–5 ADS-equipped or CVs carry the promise of enhanced efficiency, increased safety, and other benefits. As the automation increases, however, and cars move towards levels 4–5, more and more data will be generated, collected, stored, transmitted, and shared. This data will include vehicle system data and accessory system data, as well as PII from owners’, drivers’, and passengers' personal devices. Vehicles

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 43 and their systems will be at risk for unauthorized access, hijacking, and theft, putting PII at risk for unauthorized access and use.  Prioritization Recommendation: First and foremost, state policy makers should ensure that any privacy-sensitive data collected from vehicles through connected infrastructure or otherwise is not publicly accessible—for example through open records statutes—in ways that could compromise the privacy of individual drivers (e.g., by being linked to specific cars). Second, states should consider whether this same data could be used by state enforcement officials in ways that compromise Fourth Amendment protections against unconstitutional search and seizures. Beyond these two areas of legal activity, new programs may ultimately be needed for specific protections— analogous to data protection in electronic data recorder laws—against data capture and privacy violations for other types of equipment in and associated with C/ADSs. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Short-Term Recommendation 5: User Attentiveness Provisions States may require the driver to be fully attentive during the DDT, but when these conditions apply, they limit the use of level 3–5 C/ADS-equipped vehicles. There are also several different requirements that regulate user behavior and demand a heightened level of attentiveness. Although the requirements are dispersed throughout the reviewed vehicle codes, some will require modification. Prioritization Reasoning All state codes appear to allow C/ADSs, particularly at level 3 and below when there is a human person seated behind the controls. A few states actually require that a human driver be attentive and control the vehicle at all times or devote full time and attention to the operation of the vehicle. Other state codes note that “drivers” must be in reasonable control of the vehicle. States should review alertness requirements and user attentiveness provisions in their MVCs, as they may deter the use of automation and negate some C/ADS user benefits. States with prohibitions against inattentive users will most likely need to modify related statutes. Modifications should include definitions or clarifications for operating and attentiveness on the part of a human within a driving automation system-equipped vehicle when lower levels (level 2 and below) of automation are deployed, and where they may be required to take over operations of a vehicle, be seated in the driver's seat, and/or awake. Driver attentiveness statutes may also have detrimental effects on the launch and operation of A-MaaS providers, as these services are predicated on the assumption that passengers will have no required level of attentiveness nor be required to assume control of a level 5 C/ADS-equipped vehicle. These statutes may also limit the operation of A-MaaS vehicles without any passengers awaiting instructions for the next pick up or while en route to pick up a passenger.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 44  Prioritization Recommendation: States should review all statutes that related to driver attentiveness and inattentive driving and consider modifications to these anti-distraction provisions depending on the level of driving automation system deployed. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Short-Term Recommendation 6: Rules of the Road States should identify and determine whether the rules of the road apply to C/ADSs and make appropriate modifications to motor vehicle laws. Some states have indicated in newly developed statutes that all provisions of rules of the road and accompanying penalties apply to level 4–5 C/ADS-equipped vehicles. Additionally, when these rules apply to “drivers,” clarification is needed as to who or, in the case of level 4–5 C/ADS-equipped vehicles, what that “driver” is to ensure that C/ADSs are not exempted from rules of the road requirements. Prioritization Reasoning In addition to standard rules of the road (speed limits, passing restrictions, etc.), rules of the road often apply standards of due care, such as yielding to a blind person, right of way for funeral processions, and yielding to school buses or ice cream trucks. While level 4–5 C/ADSs may have sensory means to address these issues, states still need to consider these issues at all levels. Visibility is a trigger for a number of rules of the road requirements and states ultimately may need to modify or clarify these visual cue requirements to tailor them more specifically to the capabilities of C/ADSs  Prioritization Recommendation: States should review all rules of the road with an eye to the “human” element and implement provisions that apply to all drivers and vehicles as appropriate. Visual cues should also be reviewed. Benchmarks may need to be modified or adjusted by policy makers to accommodate the sensory abilities of C/ADS-equipped vehicles operating at level 3 or above. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Short-Term Recommendation 7: Local Restrictions The issue of local restrictions is one of the overarching priority considerations discussed earlier in this report. The UVC and some states provide explicit powers to local authorities to override the state laws in their MVCs. The sharing of powers and responsibilities between the states and localities with respect to C/ADSs may be a major impediment unto itself. Most of these challenges, however, lie beyond MVCs. Given the importance and scope of this issue, we spotlight this type of provision, both for rules of the road and platoons. State-local cooperation, authorized by the law, may need refining or modification in the future in some states. Moving forward, states should begin reviewing existing statutes that allow for local control/local restrictions in light of level 3 and above C/ADS-equipped vehicle deployment.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 45 Prioritization Reasoning Local controls over roadways, with regard to both who can operate on them and the rules of the road, is an area that is likely to require the examination of local restrictions statutes. If these requirements need to be adjusted for C/ADSs, then state laws may need to be modified in ways that will involve significant policy choices. Alternatively, if new programs or directives are needed to ensure closer harmonization between states and localities, then this is an area for wholly new legislative or regulatory activity. Various levels of authority can impede commerce and have implications on harmonization for future C/ADS deployment. Examples of state-enacted local restrictions may be found in North Carolina and Illinois (North Carolina House Bill 469, Illinois Public Act 100-0352).  Prioritization Recommendation: Some states’ codes provide local governments with authority to regulate traffic and impose local restrictions in additional to state restrictions. To the extent that local provisions would be stricter than state code restrictions, states should work with local governments to determine if the local controls should remain in place with deployment of level 3 and higher C/ADS-equipped vehicles. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Short-Term Recommendation 8: Aftermarket Technologies The application of existing state laws to aftermarket conversions of conventional vehicles into C/ADS- equipped vehicles remains unclear. Existing laws should be revised or clarified. In addition, aftermarket modifications should be classified and DMVs notified via a state-determined process if C/ADSs are installed on a vehicle. States should be aware that NHTSA governs vehicle safety equipment, and state revisions to this area of law are subject to federal preemption. Prioritization Reasoning There is little legislative activity in this area. Nevada recently passed AB 69, which does extend immunity from liability for damages caused by modifications by an unauthorized third party to the original manufacturer or developer of a C/ADS. (An Act Relating to Transportation, 2017). And while the possibility of aftermarket C/ADS downloads has been discussed, none have yet been mass produced for the market. However, this is an area that is likely to be part of the ongoing C/ADS conversation and states should prepare early on for these possible modifications. Harmonization Reasoning ADS-Equipped Passenger Vehicles and A-MaaS Urgency for harmonization is also noted regarding aftermarket C/ADSs, in which passenger vehicle owners convert a level 0 or level 1–2 driving automation system-equipped vehicle into a level 3–5 C/ADS-equipped vehicle through the addition of sensors and other automated driving technology (Serian et al., 2017). Wagner et al. (2018) notes that under both the UVC and some state laws, consumer

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 46 modifications of vehicles are regulated. Most of these laws, however, seem to engage only at the inspection stage, if at all. This means that most aftermarket modifications remain largely unregulated. This market could potentially become unsafe unless governments establish clear boundaries regarding modification of existing vehicles with C/ADS capabilities. From a harmonization standpoint, the possibility of unsafe aftermarket modifications endangers all road users, regardless of the state in which the vehicle operates. Therefore, early harmonization is important to avoid a “vehicle automation outfitter” setting up shop in the state with the least stringent rules on such modifications. Here, upwards harmonization is needed to establish a high-quality standard.  Prioritization Recommendation: States should begin to consider how they will address, legislate, or regulate aftermarket C/ADSs. Current laws and regulations may need to be revised or clarified with respect to whether and how they regulate aftermarket driving automation system- related technologies installed on a vehicle. Harmonization Recommended: Yes. Guidelines (policy decision). Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A- MaaS. Short-Term Recommendation 9: Unattended Vehicles Prioritization Reasoning States should consider clarifying the meaning of laws that prohibit unattended vehicles, especially for level 4–5 C/ADS-equipped vehicles. While mass deployments of passenger vehicles operating at these levels are beyond the mid-term timeframe, this legal area will still need to be considered for A-MaaS vehicles in the short-term. Harmonization Reasoning C/ADS and A-MAAS States and A-MaaS providers will both benefit from the harmonization of the meaning and application of laws that prohibit unattended vehicles. Unattended vehicles represent a public nuisance in that they can block traffic and block use of the roadway shoulder by emergency first responders. To address this concern, states generally include in their MVCs the prohibition of drivers leaving their vehicles unattended (except for parking). Under this law, however, an empty A-MaaS C/ADS-equipped vehicle operating without any passengers or any driver could be considered “unattended” because the engine is running and the brake is not applied. The ability for an A-MaaS vehicle to operate while empty, presumably on its way to pick up a passenger, is central to the development of this business model. This effort should be undertaken in the short term and harmonized to establish uniformity and clarity of the term “unattended” and its applications, as A-MaaS-focused efforts to harmonize will eventually apply to C/ADS-equipped passenger vehicles (e.g., for self-parking) as well.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 47  Prioritization Recommendation: Clarify the meaning of laws that prohibit unattended vehicles, especially for level 4–5 C/ADS-equipped vehicles, including A-MaaS fleet vehicles. Harmonization Recommended: Yes. Best practice language Consumer C/ADS application(s) affected first: A-MaaS. Short-Term Prioritization Recommendation Summary In the next 2 years (2018–2020), the following state MVC provisions are suggested for consideration for modifications (Table 6). Note, items below are not ordered chronologically. Table 6. Short-Term Prioritization and Harmonization Modification Recommendations RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING DEFINITIONS AND DRIVER ONLY VEHICLE CODES Conduct a critical review of fundamental vehicle code terms “drive,” “driver,” “operate,” and “operator,” and develop necessary clarification in terms, intent, and interpretation. Harmonization recommended C/ADS-equipped passenger vehicles and A-MaaS Best practice language Address the possibility that vehicle codes can be interpreted to regulate only “drivers” (who are licensed and human) and exempt level 4–5 C/ADS-equipped vehicles from legal oversight. Harmonization recommended C/ADS-equipped passenger vehicles and A-MaaS Guidelines (policy decision) PLATOON-RELATED ISSUES Consider the need to modify following distance requirements for platoons on a state’s highways. This is particularly important in states that impose prescriptive following distances. Additionally, provide guidance and clarify the legal classification of truck platoons. Harmonization recommended Platooning ADS- equipped CMVs Best practice language; receive guidance (e.g., from the Commercial Vehicle Safety Alliance)

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 48 RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING Develop restrictions as needed if technical scan/engineering analyses identify any negative length, weight, and/or noise effects due to trucks operating as a platoon. Further, audit state laws and regulations that may impose lane restrictions or service requirements on platoons to develop harmonization across the state. Useful but not essential Platooning ADS- equipped CMVs Best practice language; receive guidance (e.g., from the Commercial Vehicle Safety Alliance) VEHICLE TITLING AND REGISTRATION Memorialize, from the time of manufacture to junk or salvage on title and registration documents, that the vehicle is driving automation system-equipped. Consider memorialization of aftermarket technologies. Harmonization recommended C/ADS-equipped passenger vehicles and A-MaaS Guidelines (policy decision) PRIVACY PROTECTIONS Assess state policy protections for privacy- sensitive data collected on vehicles through connected infrastructure and vehicle transmission and also the implications of open records laws and the applicability of current state privacy protection statutes. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. USER ATTENTIVENESS Modify prohibitions against inattentive drivers depending on level of driving automation system deployed. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. RULES OF THE ROAD – APPLICABILITY TO C/ADS Identify how and whether the rules of the road apply to different levels of driving automation systems. Ensure that level 4–5 C/ADS- equipped vehicles are not exempted from rules of the road requirements. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 49 RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING RULES OF THE ROAD – LOCAL RESTRICTIONS Modify local controls over roadways for who can operate on them, the rules of the road, and consider issues of state level preemption. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. AFTERMARKET MODIFICATION Revise or clarify existing laws with respect to whether and how they regulate aftermarket driving automation system-related technologies installed on a vehicle. Harmonization recommended C/ADS-equipped passenger vehicles and A-MaaS Guidelines (policy decision) UNATTENDED VEHICLES Clarify the meaning of laws that prohibit unattended vehicles, especially for level 4–5 C/ADS-equipped vehicles, including A-MaaS vehicles. Harmonization recommended A-MaaS Best practice language Mid-Term (2021–2025) Priorities The mid-term timeframe of 2021–2025 will see significant change in terms of vehicles at different SAE J3016 levels of automation. We anticipate an increasing number of level 3 C/ADS-equipped passenger vehicles, the introduction of level 4 C/ADS-equipped A-MaaS vehicles, and the early introduction of level 4 C/ADS-equipped passenger vehicles. The most significant and rapid change is expected to happen during this period. The following law and/or regulation code changes are recommended in advance of this timeframe. Mid-Term Recommendation 1: User Qualifications, Testing, and Driver Education States should expect to modify licensing qualifications and requirements and determine if additional user testing and education is necessary. This effort goes well beyond the issuance of a digital or physical credential with a “C/ADS-equipped” designation. States will also need to look at any underlying reasons citizens have NOT been able to obtain a license to drive and if those issues can be resolved in level 4–5 C/ADS-equipped vehicles. These reasons might include medical competency provisions, sight restrictions, or other physical restrictions. Regulations governing medical competency boards should also be reviewed to determine possible changes in representation and skill sets. By this point, states should

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 50 have addressed the underlying fundamental definitions of “drive,” “driver,” “operate,” and “operator,” making a review of these types of provisions less daunting. Testing components should also be reviewed during this time, and states might consider adding level 4–5 C/ADS component language to existing driving tests to help ensure that the user understands the C/ADS’s functional limits. Some of these terms may include C/ADS, DDT, minimal risk condition, operator and driver, and ODD limits. Education requirements will similarly need to be examined for possible inclusion of new provisions and units to help users understand the appropriate times to engage and disengage a C/ADS, as well as how to operate during a handoff situation. Similar consideration should be given to how to interact with level 4–5 C/ADS-equipped vehicles. Prioritization Reasoning Changes in this area will be incremental and informed by the technology. In reality, user testing components are likely to be driven by policy and not law in most states, and user education is clearly state based. So, while states should review their laws and regulations for testing and education, the much more pressing component is user qualification, particularly adjustments to allow those who could not “drive” before to “drive” now. Further, any change to issuance requirements, testing, or training will be on a state-by-state basis and is likely to intersect with federal mandates or requirements for non-commercial drivers in this area. The emphasis needs to be on reciprocity and best practices or model minimum standards developed by the states in concert with AAMVA. Harmonization Reasoning C/ADS-Equipped Passenger Vehicles and A-MAAS Many A-MaaS companies promise their services will launch in the 2021 to 2025 timeframe (Ford, 2017; Bosch Media Service, 2017; Townsend, 2016). As such, the transition from testing with a safety driver in the vehicle to level 4 C/ADS-equipped A-MaaS vehicle will require several changes to state MVCs and will benefit from state harmonization. One such case can be found in the provisions of the code that specify who may and may not operate a vehicle. Among the most heavily promoted benefits to come from self-driving cars is improved access and mobility for all (Duncan et al., 2015; Douma, Lari, & Andersen, 2017; National Association of City Transportation Officials, 2016). Today, state MVCs limit or prohibit entirely certain individuals from obtaining licenses and operating motor vehicles through provisions requiring vision and operating tests. To fully reach the potential benefits of C/ADSs (including C/ADS-equipped passenger vehicles), states will need to address the barriers that currently prevent many people from simply riding as passengers in level 4 or 5 C/ADS-equipped vehicles.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 51  Prioritization Recommendation: Determine who can operate C/ADSs at different SAE J3016 levels of automation and adjust the law for driver licensing requirements.  Prioritization Recommendation Develop driving tests (or amend existing tests) keyed to the different SAE J3016 levels of automation. Harmonization Recommended: Useful, but not essential. Reciprocity agreements, best practice language. Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A-MaaS. Mid-Term Recommendation 2: Reasonable Articulable Suspicion (Implied Consent) States should consider when a “reasonable articulable suspicion” of alcohol or drug use is appropriate in settings where the system is controlling a level 4–5 C/ADS-equipped vehicle. For example, if it is possible for law enforcement to confirm at the scene that the ADS was properly engaged at the time of a violation or crash, there can be no presumption of a “reasonable articulable suspicion” of impaired driving. Amendments to implied consent would seem logically to lead to parallel allowances for drinking for C/ADS operators under conditions that allow for safe operation for high levels of automation. Yet the latter decision would involve more comprehensive standards and public decisions about when drinking is and is not allowed in operating vehicles. Adjustments to implied consent for higher levels of automated driving will not lead to the automatic legalization of alcohol consumption while driving. Prioritization Reasoning Until the legal ambiguities over who is a driver/operator are resolved and modified in statutes and regulations where appropriate, there could be confusion and a deadlock for law enforcement in administering alcohol and drug tests. For example, if a police officer pulls over a swerving car and the user maintains that the level 3–5 C/ADS-equipped vehicle was operating with the ADS properly engaged and hence by law they are not the “driver” or “operator,” is there a general presumption that implied consent applies or must the police officer rebut this defense with added evidence of some kind? Conversely, could the implied consent provision be interpreted too aggressively in ways that stigmatize user who are operating the vehicle safely, particularly if alcohol restrictions have been partly lifted for some SAE J3106 vehicle levels?

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 52  Prioritization Recommendation: Consider when a “reasonable articulable suspicion” of alcohol or drug use is appropriate in settings where the ADS is engaged in a level 3–5 C/ADS-equipped vehicle. Harmonization Recommended: No. Assumes definitions are modified to clarify users are considered passengers when traveling in a level 4–5 C/ADS-equipped passenger vehicle or A-MaaS vehicle. Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A-MaaS. Mid-Term Recommendation 3: Prohibitions Against Use of Alcohol and Legal Drugs The UVC and all states prohibit a person from driving while intoxicated or under the influence of drugs or narcotics. The UVC directs that a person shall not drive any vehicle while “under the influence of alcohol” or “any drug to a degree which renders such person incapable of safely driving” or that exceeds prescribed blood concentrations set in the statute (UVC § 11-901). In most states, these general prohibitions are reinforced by legislative limits on alcohol and drugs in the driver’s body. Nevada’s statutes are illustrative: it is unlawful for a person to drive if he or she “[h]as a concentration of alcohol of 0.08 or more in his or her blood or breath; or . . . is found by measurement within 2 hours after driving . . . to have a concentration of alcohol of 0.08 or more in his or her blood or breath” (NV Rev. Stat. § 484C.110). In most states, alcohol for commercial vehicles and for minors are generally stricter, sometimes approaching a zero-tolerance level (See, e.g., CA Veh. § 6-51 [commercial vehicles]; § 11-905 [zero tolerance for minors]). Prioritization Reasoning States will need to clarify alcohol and drug use and regulation within the various SAE J3016 levels of automation (including in states where marijuana has been legalized). For example, for higher levels of automation, states may decide to allow operators to consume alcohol in advance of or during driving. If alcohol consumption is allowed for some levels of automation, this legal modification will also impact the implied consent requirements discussed above. Implied consent for higher levels of automation may need to be adjusted or even eliminated entirely.  Prioritization Recommendation: Clarify alcohol and drug use and regulation within the various SAE J3016 levels of automation (including in states where marijuana has been legalized). Develop offenses, fines, and sentencing terms at varying levels of automation. Adjust implied consent requirements accordingly. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 53 Mid-Term Recommendation 4: Motor Vehicle Liability Prioritization Reasoning Some crashes, incidents, and harms will be the result not of human driver error, but rather flaws in the engaged ADS when it is in control of the vehicle. Some state laws currently place full legal responsibility for any damages and harms on the human driver and preclude any manufacturer liability. Policy makers should consider whether it is more equitable to place primary responsibility on vehicle manufacturers and technology companies when the ADS is engaged and in control of the vehicle, or at least include partial responsibility for manufacturers. There is also a parallel need to define the process for determining who was in charge of driving at the time of the crash (likely by modifying event data recorder regulations, typically a federal responsibility).  Prioritization Recommendation: Determine responsibility for crashes, incidents, and harms that may not be the results of human operator errors, but rather flaws in the ADS as operating at the time of the incident. Harmonization Recommended: Useful, but not essential. Best practice language. Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles (later) and A-MaaS (now). Mid-Term Recommendation 5: Due Care Standard Policy makers should clarify how the “due care” standard applies when a vehicle is operating at SAE J3016 levels 3–4. Presumably, there are “due care” considerations applied to the human driver with respect both to whether to engage the ADS and whether to override or disable the ADS in specific settings. Since these types of “due care” actions currently have no reliable benchmark in human experience, states should define what these terms might mean in different technological and operational scenarios. Prioritization Reasoning Throughout the UVC and state MVCs, users are regularly instructed that they must exercise “due care,” act “reasonably” and with “prudence” and avoid “carelessness” to comply with the rules of the road. Illustrative examples of state statues include the following: • Nebraska simply requires that “[a]ny person who drives any motor vehicle in this state carelessly or without due caution so as to endanger a person or property shall be guilty of careless driving” (NE Rev. Statutes § 60-6,212). • In California, the “driver of a motor vehicle when reasonably necessary to insure safe operation shall give audible warning with his horn.” This “reasonable and prudent” directive is unavoidably benchmarked against human judgment.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 54 It seems likely that C/ADS programmers can match and exceed good human judgment in most cases, but in other, more complex settings (e.g., “traveling upon any narrow or winding roadway, and when special hazards exist”) the C/ADS may not fully grasp the situation as a prudent human would (NE Rev. Statutes § 60-6,212). In some cases, human decision-making abilities, coupled with multi-sensory cues, assist in making this “reasonable” judgment required by law (UVC § 11-304). Similarly, passing streetcars or driving on street car tracks invoke some state requirements that the driver “proceed only upon exercising due caution for pedestrians” exiting and around the streetcar (UVC § 11-1401—1404). C/ADSs raise still more challenges with regard to this human judgment requirement by presenting a possible bifurcated “reasonable person”—one who is human driving a non-C/ADS-equipped vehicle and the other driving, for example, a level 3 C/ADS-equipped vehicle with the ADS engaged. What a reasonable human might do rounding a sharp curve may be different than what a reasonable human operating in a level 3 ADS-equipped vehicle would do given the dramatic differences in vehicle capacities. This fact only further underscores how these judgment-based provisions may complicate how compliance is assessed at different SAE J3016 levels of automation.  Prioritization Recommendation: Modify or adjust benchmarks to accommodate the decision- making abilities of level 3–5 ADS-equipped vehicles, especially for the “due care” standard. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Mid-Term Recommendation 6: User Distraction Provisions Prioritization Reasoning Most states prohibit driver distractions, which can take the form of video screens, texting, earphones, and even grooming. None of these driver distraction provisions impede the use of C/ADSs and none create safety hazards. Given the features of ADSs, some of these provisions may need to be modified as the fleet evolves into SAE J3016 levels 4–5. In particular, level 5 C/ADS-equipped vehicles operating as part of an A-MaaS fleet will assume no passenger responsibility for attentiveness nor have any expectation that the passenger will retake control of the vehicle.  Prioritization Recommendation: Modify anti-distraction provisions to enhance the utility of level 4–5 ADS-equipped vehicles for those who would benefit from their use. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 55 Mid-Term Recommendation 7: Unfair Criminal and Civil Sanctions on Users Prioritization Reasoning Drivers can be charged with criminal acts for violating certain traffic laws, but in the case of level 4–5 C/ADS-equipped vehicles, some of these criminal charges are less straightforward. Certain provisions may assign criminal liability to a human operator in situations where the vehicle could be completely to blame, or at least largely responsible, for resulting harms or violations. For example, in vehicles operating at SAE J3016 levels 4–5, C/ADS malfunctions may cause the vehicle not to register a visual or audible signal from law enforcement, creating criminal liability for the human operator. Criminal acts that could be a result of C/ADS malfunctions need to be reviewed in MVCs as do assumptions of liability against OEMs, suppliers, technology companies, and A-MaaS network companies. Harmonization Reasoning With many OEMs and new entrants in the manufacturing sector targeting this time range for deployment, states will need to harmonize their vehicle codes in advance. This will involve addressing several new areas of policy and considering major operational issues for C/ADSs, such as liability and faults in the ADS as engaged at the time of the event of interest. Harmonization is thus recommended for level 3–5 C/ADS-equipped passenger vehicles.  Prioritization Recommendation: Amend statutes governing criminal and civil liability to leave open the possibility that a level 3–5 C/ADS-equipped vehicle with a properly engaged ADS could also be responsible in whole or in part for a resulting violation. Harmonization Recommended: Yes. Best practices language. Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A-MaaS. Mid-Term Recommendation 8: Crash Reporting and Rendering Aid Prioritization Reasoning States should consider reviewing and modifying rendering aid statutes. These “report and render aid” provisions can create two separate challenges for level 4–5 C/ADS-equipped vehicles. The first challenge is that compliance requires a human. Thus, in cars without an active operator or even able occupants, compliance with the requirements may be difficult, if not impossible. The second challenge returns to the problem of definitions. The render aid requirements focus their mandates on the terms “drivers” and “operators.” It is possible that if a level 4–5 C/ADS-equipped vehicle operating without a human aboard causes an accident, it could be exempt from these provisions since there is no “driver” (human present in the car). Again, while most of the related deployment modifications apply to level 4–5 C/ADS-equipped passenger vehicles, this provision also applies to A-MaaS vehicles, which are expected to be deployed in the mid-term.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 56  Prioritization Recommendation: Modify rendering aid statutes for level 4–5 C/ADS-equipped vehicles to address the fact that a human driver may not be present in the vehicle. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Mid-Term Recommendation 9: Vehicle Requirements Prioritization Reasoning Most states impose specific types of requirements on the physical features of vehicles. While most of these vehicle requirements are open-ended and do not appear to impede C/ADSs, some do appear to pose potential impediments in the future (e.g., horns and other audible warning devices, steering wheels, mirrors, brake pedals). Level 4 and level 5 C/ADS-equipped vehicles, especially those that are C/ADS- DVs, may not conform to these requirements. As a result, certain vehicular requirements in state codes may become outmoded or unduly prescriptive in limiting the types of C/ADSs allowed in the state. As part of a detailed MVC audit, policy makers may want to consider identifying and modifying soon to be obsolete requirements so that more applicable terms are used (e.g., referring to “steering assemblies rather than “steering wheels”/“wheels” or “braking systems” rather than “pedals”). The aforementioned NHTSA project, Assessment, Evaluation, and Approaches to Technical Translations of FMVSS and Test Procedures That May Impact Compliance of Innovative New Vehicle Designs Associated with Automated Driving Systems, is considering issues associated with vehicle requirements. States should be aware of this research effort and monitor NHTSA for findings related to the project. Harmonization Reasoning C/ADS-Equipped Passenger Vehicles and A-MaaS Many features of the modern motor vehicle could become obsolete in C/ADSs, including items related to vehicle controls, such as brakes and a steering wheel (NHTSA, 2016b). Not only will these devices become unnecessary in level 4–5 C/ADS-equipped vehicles, they will likely be intentionally absent for level 5 vehicles utilized by A-MaaS services, as service providers will seek to prevent passengers from taking control of the vehicle. As such, states should ensure they harmonize vehicle codes and standards to enable A-MaaS vehicles to operate across state lines. Harmonization would also enable manufacturers produce a fleet that can operate throughout the U.S. without customizing vehicle design and operation on a state-by-state basis. A-MaaS-focused efforts to harmonize in this area apply to level 4–5 C/ADS- equipped passenger vehicles as well.  Prioritization Recommendation: Begin to identify obscure requirements that reference specific items (use “steering assemblies” rather than “wheels” and “braking systems” rather than “pedals”) to address over the longer-term.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 57 Harmonization Recommended: Yes. Reciprocity agreements and/or federal preemption (likely a policy decision). Consumer C/ADS application(s) affected first: C/ADS-equipped passenger vehicles and A-MaaS. Mid-Term Prioritization and Harmonization Recommendation Summary In summary, the following state motor vehicle code provisions are suggested for modifications in the short term (Table 7). Note, items below are not ordered chronologically. Table 7. Mid-Term (2021–2025) Prioritization and Harmonization Modification Summary RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING DRIVER LICENSING Determine who can operate driving automation systems at different levels of driving automation and adjust the law for driver licensing requirements. Useful but not essential C/ADS-equipped passenger vehicles and A-MaaS Reciprocity agreements, best practice language DRIVER TESTING AND EDUCATION Develop driving tests (or amend existing tests) keyed to varying levels of driving automation systems. Useful but not essential C/ADS-equipped passenger vehicles and A-MaaS Reciprocity agreements, best practice language IMPLIED CONSENT Consider when “reasonable articulable suspicion” of alcohol or drug use is appropriate in specific ODD with a properly engaged level 3–5 C/ADS-equipped vehicle. No harmonization recommended C/ADS-equipped passenger vehicles and A-MaaS Assumes definitions are modified to clarify users are considered passengers when traveling in a level 4– 5 C/ADS-equipped passenger vehicle or A-MaaS

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 58 RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING PROHIBITIONS AGAINST USE OF ALCOHOL AND LEGAL DRUGS Clarify alcohol and drug use and regulation (including in states where marijuana has been legalized) within the various levels of driving automation. Develop offenses, fines, and sentencing terms for lower level violations at varying levels of driving automation. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADS MOTOR VEHICLE LIABILITY – USER AND OWNER LIABILITY FOR DAMAGES Determine responsibility for crashes, incidents, and harms that may not be the result of human error but rather flaws in the C/ADS as engaged at the time of the event of interest. Useful but not essential C/ADS-equipped passenger vehicles (later) and A-MaaS (now) Best practice language RULES OF THE ROAD – DUE CARE STANDARD AND HUMAN JUDGMENT Modify or adjust benchmarks to accommodate the decision-making abilities of level 3–5 C/ADS-equipped vehicles operating at level 3 or above, especially for the “due care” standard, which is tethered to human judgment. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADS USER DISTRACTIONS Modify anti-distraction provisions to enhance the utility of C/ADS-equipped vehicles for their drivers (while the ADS is unengaged) or passengers (while the ADS is engaged). No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADS UNFAIR CRIMINAL AND CIVIL SANCTIONS ON USERS (REASONABLE ARTICULABLE SUSPICION) Amend statutes governing criminal and civil liability to leave open the possibility that when properly engaged, the ADS in a level 3–5 C/ADS-equipped vehicle could also be responsible in whole or in part for a resulting violation. Useful but not essential C/ADS-equipped passenger vehicles (later) and A-MaaS (now) Best practice language

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 59 RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING CRASH REPORTING AND RENDERING AID Consider the need for modifications to “rendering aid” statutes for level 4–5 C/ADS- equipped vehicles. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADS VEHICLE REQUIREMENTS Consider culling obscure requirements that reference specific items (e.g. use “steering assemblies” rather than “wheels” and “braking systems” rather than “pedals”). Harmonization recommended C/ADS-equipped passenger vehicles and A-MaaS Reciprocity and/or federal preemption (likely a policy decision) Long-Term (2026 and beyond) Priorities Beyond the short- and mid-term changes needed, states should undertake several longer-term efforts after 2026, once the C/ADS market begins to reach greater penetration of the broader fleet. These include efforts to fully normalize and structure the operation of C/ADSs, both for personal use and under an A- MaaS service, such as through the adaptation of vehicle inspection requirements. Long-Term Recommendation 1: Vehicle Inspection Prioritization Reasoning Some agency inspection laws and accompanying regulations will likely need to be modified to accommodate the new technological features of C/ADS-equipped vehicles, such as the absence of steering wheels and brake pedals. Inspection laws and regulations may also need to be amended to include new requirements, such as mechanisms for disengaging a level 3–5 C/ADS-equipped vehicle, to ensure the safety of these vehicles on state roadways, and for any aftermarket C/ADS technologies applied to a vehicle. Note that, based on our 15-state review, no issues were found with auto emission regulations or laws.  Prioritization Recommendation: Modify agency inspection legislation/regulations to accommodate the new technological features of C/ADSs. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 60 Long-Term Recommendation 2: Consumer Protection (i.e., Lemon Laws) Prioritization Reasoning States should consider needed modifications to lemon laws. These laws, originally designed to protect consumers, may not be sufficient to ensure adequate protection from C/ADS product defects. For states that have adopted these laws, some modifications may be necessary to account for the fact that problems with the programming or automation may not be evident over the relatively short period during which manufacturers are legally held responsible for making repairs. Harmonization Reasoning As states begin to consider their lemon laws, best practice language may be useful in helping states to account for new technologies in order to ensure adequate consumer protection from product defects.  Prioritization Recommendation: Modify lemon laws to account for new C/ADS technologies to ensure adequate consumer protection from product defects. Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs. Long-Term Recommendation 3: Occupant Safety and Protection Prioritization Reasoning Occupant safety requirements may need to be revised to take full advantage of C/ADS sensor capabilities (e.g., sensing the weight of each passenger to determine appropriate safety restraint use; disengaging when belts are not in place so that the vehicle will not operate in conflict with safety requirements laws). Presently, at least current state child restraint requirements (i.e., the age of a child determines the required restraint) may not be programmable in a C/ADS. However, the same or improved protections for occupants may be accomplished by alternative sensory-based requirements (e.g., passenger weight determines restraint type). Legal responsibility of drivers for meeting occupant safety provisions for level 4–5 C/ADS-equipped vehicles may also need to be assessed to determine contributory negligence provisions within revised tort laws.  Prioritization Recommendation: Revise occupant safety requirements to take full advantage of C/ADS sensory capabilities (e.g., seatbelts and child restraints). Harmonization Recommended: No. Not currently harmonized at the state level; no specific reason to harmonize for C/ADSs.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 61 Long-Term Prioritization and Harmonization Modification Recommendation Summary In summary, the following state motor vehicle code provisions are suggested for modifications in the long-term (Table 8). Note, items below are not ordered chronologically. Table 8. Long-Term (2026 and Beyond) Prioritization and Harmonization Modification Summary RECOMMENDATION HARMONIZATION RECOMMENDED? CONSUMER C/ADS APPLICATION AFFECTED FIRST MEANS OF ADDRESSING VEHICLE INSPECTION Modify agency inspection legislation/regulations to accommodate the new technological features of C/ADS. No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADS CONSUMER PROTECTION LAWS Modify lemon laws to account for new driving automation system-related technologies to ensure adequate consumer protection from product defects. Useful but not essential Best practice language OCCUPANT SAFETY AND PROTECTION Revise occupant safety requirements to take full advantage of driving automation system- equipped vehicles' sensory capabilities (e.g., seatbelts and child boosters). No harmonization recommended Not currently harmonized at the state level; no specific reason to harmonize for C/ADS 6.4 Complex Interplay of Deployment Transition and Interoperability Over time, the composition of the vehicle fleet will shift from almost entirely human-operated vehicles to almost entirely C/ADS-equipped vehicles. The pace of this evolution is difficult to predict, but the timeline envisioned in this study sees this transition occurring over the next 30 years, with level 1 vehicles predominating now and level 5 C/ADS-equipped vehicles perhaps beginning to dominate by the end of the 30-year period or just beyond. In this assessment the implications of the interoperability issues that this mix of vehicles will create for state MVCs were considered. It could be argued that state MVCs should evolve apace with the evolving vehicle fleet. However, this approach would inevitably cause lags between the emergence of C/ADS-

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 62 related issues and appropriate state MVC changes. A better approach is the application of research such as this, further study of the impact of C/ADSs on various jurisdictions, and the timely enactment of comprehensive legislative schemes to systematically address complex issues as they emerge. Oliver Wendell Holmes famously observed that "the life of the law has not been logic: it has been experience" (Holmes, 2009, p.1). But the body of experience and case law related to C/ADSs will develop more smoothly if logical, comprehensive legislative schemes are in place from the outset. Absent such schemes, case law will develop haphazardly, driven by narrow factual situations and political concerns. As a practical matter, this will impede the implementation of C/ADSs and delay the realization of their ultimate benefits. The demands imposed upon a state MVC by a 10% C/ADS-equipped vehicle market penetration rate versus an 80% market penetration rate are not appreciably different. A comprehensive state policy and legislative approach is needed in either scenario. Therefore, the priorities outlined in this Assessment should be considered regardless of the anticipated penetration rate. In terms of priorities, the progression of varying levels of C/ADS fleets are not expected to be linear. As noted in the timeline, level 5 C/ADS-equipped A-MaaS vehicles will be sharing the road with level 2 passenger vehicles and level 1 C/ADS-equipped truck platoons. Given this scenario, in the event of a crash involving a level 1 platoon vehicle, a level 2 passenger vehicle, a level 4 C/ADS-equipped vehicle, and a conventional vehicle, ALL laws relating to ALL vehicles would need to be on the books. States recognize that there will likely be decades of transition from conventional vehicles to level 4–5 C/ADS-equipped vehicles. Accordingly, planning and designing for mixed environments is essential. From a DOT perspective, this effort will require rethinking planning processes. The same will hold true for DMVs and law enforcement. 6.5 Prioritization and Harmonization Conclusions States need to begin modifying their laws and regulations now, at least in basic areas such as definitions, and in rapidly advancing areas. However, while moving forward is important, any action must also be tempered by a state recognition that some issues are not yet clear. That lack of clarity is primarily rooted in three key issues: 1. As of this writing, the federal direction for C/ADS-equipped vehicle oversight has yet to be set, and it is unknown what federal preemption will include or what actions Congress will take. 2. The timeline for deployment of different levels of C/ADS-equipped vehicles is also unknown, but is not expected to be linear. 3. Many states are waiting for model laws before taking action, but that effort (with the exception of AAMVA’s continued work on best practices) is not imminent and has been met with concerns from the motor vehicle community.

NCHRP Web-Only Document 253, Vol. 5: Developing the Autonomous Vehicle Action Plan 63 Despite these issues, enough states are now moving ahead to provide good law and regulation examples to consider. Drawing on these states’ laws and regulations as examples, and using the checklist and timetable of the top expected legal and regulatory modifications (See Chapter 3 of this report), state agencies, lawmakers, and other state stakeholders can begin to prioritize what modifications will be most needed in their states and plan accordingly. The eventual deployment of C/ADS-equipped vehicles will require a regulatory structure that can cross state lines as easily as these vehicles will. These recommended changes and harmonization efforts are based on their urgency, as well as their necessity for the deployment of each consumer C/ADS application. As outlined, many of the most urgent recommendations for harmonization relate to elements of state MVCs that define key terms and their implications. For example, who is a “driver” or an “operator” of a motor vehicle and what exactly does it mean to “drive” or “operate” the vehicle on public roads? Similarly, states are encouraged to harmonize changes that address previously basic scenarios. Situations where a vehicle is left unattended will soon take on a very different meaning when A-MaaS services become available, as these vehicles must be able to operate without any human passengers or operators present between trips. Additionally, for truck platoons to properly operate, issues related to minimum following distances and other core aspects will need to be addressed and harmonized so that the full benefits of this technology can be realized. Meanwhile, states should not underestimate the challenges associated with any harmonization effort, and as such, should not expend energy or time harmonizing their MVCs for matters that don’t require uniformity. We advise against harmonization for recommended changes related to such issues as distracted driving, driving under the influence, privacy protections, lemon laws, and occupant protection devices, among others. These recommended changes, while no less important, simply fail to rise to the level of requiring states harmonization, as many are not currently harmonized for level 0 vehicles, to little or no ill effect.

Next: Chapter 7. C/ADS-Equipped Vehicle Action Plan »
Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan Get This Book
×
 Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Web-Only Document 253: Implications of Connected and Automated Driving Systems, Vol. 5: Developing the Autonomous Vehicle Action Plan provides technical background on developing Volumes 1 through 4 of NCHRP Web-Only Document 253. It includes further background on terminology and definitions used in the suite of reports; legal and regulatory reviews and needs assessment; state legal and regulatory audit; a prioritization and harmonization analysis; and a Connected and Automated Driving Systems (C/ADSs) analysis.

Appendix A accompanies Vol. 5. The appendix serves as a standalone, sort-able spreadsheet delineating activities at the federal level and in each state, with links to legislation or policy materials.

View all volumes of NCHRP Web-Only Document 253:

  • Vol. 1: Legal Landscape
  • Vol. 2: State Legal and Regulatory Audit
  • Vol. 3: Legal Modification Prioritization and Harmonization Analysis
  • Vol. 4: Autonomous Vehicle Action Plan
  • Vol. 5: Developing the Autonomous Vehicle Action Plan
  • Vol. 6: Implementation Plan

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!