A Performance-Based Highway Geometric Design Process (2016)

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.

181 8.1 AASHTO Curve Model The current AASHTO model for design of horizontal curves does not produce cost-effective solutions across the range of project types and contexts. As horizontal alignment is among the most influential elements of road design, a revisiting of the AASHTO curve model with potential for developing a more robust and science-based approach is a high research priority. The following are research areas and questions that should be addressed: • Is a comfort-based model the appropriate one-for-all or even any set of conditions? What other approaches or models may apply? • What might a design model based on crash-risk factors associated with horizontal curves look like? • How should risk as measured by traffic volume be included in horizontal curve design? • Under which conditions should heavy trucks or vehicles other than passenger cars be the basis for curve design? • To what extent should the cost of constructing or reconstructing a horizontal curve be part of the design model or approach? • What interrelationships among vertical geometry, cross-section, and roadside characteristics should be incorporated in a curve design model? A straw-man approach illustrating these issues was noted earlier in this report to demonstrate a potential new overall approach to curve design. Appendix C contains a paper study of the rela- tive risk of crashes and cost effectiveness analysis for two-lane rural highway curvature, based on the AASHTO HSM models. This research area warrants a substantial multi-year study aimed at producing a revised, uni- fied, context-sensitive approach to horizontal curve design for the 2025 edition of the AASHTO Green Book. 8.2 AASHTO SSD Model Sight distance is the length of roadway ahead that is visible to the driver. This distance at a given speed determines the amount of time available for the driver to react to what is seen. The AASHTO SSD model consists of two components: perception/reaction distance and breaking distance. The perception/reaction distance is the travel distance at the design speed for the reaction time taken as 2.5 seconds. SSD is to be provided at all locations along any road. The SSD model is applied to all roadways regardless of the context. The distance varies based on the speed, but the perception/reaction time does not change based on context. C h a p t e r 8 Research and Knowledge Needs to Fully Implement the Revised Process

182 a performance-Based highway Geometric Design process The DSD for a stopping maneuver uses the same model but varies the perception/reaction time based on the context. In urban areas where there is more distraction the time is longer. This component varies from 3 seconds to 9.1 seconds for rural and urban roads. AASHTO would ideally revisit the subject of SSD and sight-distance controls in general for the urban environment. The research should focus on: • Risk-based approaches incorporating traffic volume, road type, and likely presence of conflicts; • Relationship of available sight distance to relevant crash types and frequency; and • Applicability of SSD to design of urban streets vs. an alternative approach based on typical crossing conflicts. 8.3 O&M Understanding Related to Geometric Design Elements For most agencies, the typical project development process takes into consideration the capi- tal costs of a project. The cost to maintain the facility is not usually a significant consideration. In a few cases, the feasibility of future maintenance may be considered, but the cost of that future maintenance is typically not. It is important to consider future maintenance costs and activities during the planning/design stage of any significant infrastructure investment, including a high- way project, as decisions made early in the life cycle can have a material impact on future main- tenance costs. Further Research and Study Areas that would benefit from further research and study include: • Costs and benefits of critical design options to establish dimensional criteria and guidance, including examples. • Statistical analysis of maintenance personnel incidents and development of substantive crash prediction methods to guide the geometric design of alternatives that affect work zone safety, enforcement, and incident management options. • Better understanding of the short- and long-term risks concerning the operation and main- tenance of facilities that use ITS technology. 8.4 Challenges with New Process The implementation of revised geometric design processes faces numerous challenges as seen in the experiences associated with changes that accompanied load resistance factor design (LRFD) for bridges, CSS, and the AASHTO HSM. The challenges that accompanied LRFD implementation offer constructive insights considering this transition also centered on a new design methodology or process consistent with the focus of this NCHRP research. Ultimately, the states’ transition to LRFD took more than a decade despite encouragement by FHWA and AASHTO. While there are numerous reasons for the extended transition, the first is found in the process to prepare the new manuals, which involved AASHTO leading their development, review, and approval process. Each state subsequently underwent a similar review in order to update its manuals, guidelines, and procedures. Although this sort of development process takes time, it is acknowledged that this was not the only reason that accounted for the extended transition. In addition to the development of the new LRFD manuals, the transition was drawn out due to institutional inertia (that is, agency or personal resistance to change) focused around a

research and Knowledge Needs to Fully Implement the revised process 183 fundamental disagreement over the need to change (“the original way is working fine”). It has been observed that individual resistance was aggravated because the new process at times resulted in differing designs. Additional issues centered around the need for access to additional train- ing and software. Training was required for junior and senior staff so that they could become familiar with and capable of applying the new process. Universities also needed to redevelop curriculum around the new process. The initial lack of software using the new processes also delayed the transition to LRFD. Without software, the design and review process would have been lengthy and costly to complete, especially for complex designs. Therefore, the availability of software was important to embracing the new LRFD approach. Similar challenges were seen when the CSS process and the HSM were introduced into the industry. Likewise, challenges are expected to occur when implementing changes to the geomet- ric design process. This chapter reviews several of the design processes discussed in the other chapters of this report, summarizes the key challenges expected (restraining forces), and presents several suggested practices (driving forces) to help mitigate these challenges. 8.4.1 Summary of Potential Design Processes Ten alternative design process concepts are presented below. • The complete streets concept focuses on creating roadways and related infrastructure that pro- vide safe travel for all users. • The concept of CSD, better known as CSS, places priority on ensuring that highway projects fit the context of the area through which they pass, puts project needs as well as the values of the highway agency and community on a level playing field, and considers all trade-offs in decision making. • The concept of performance-based design incorporates a design process that considers explicit consideration of performance measures, typically operational and safety performance measures. • The concept of practical design focuses on addressing only those improvements that are needed and eliminating those improvements that are not absolutely essential, thereby reduc- ing the overall cost of a project. • The design matrix approach includes three levels of design for highway projects: basic, modi- fied, and full. • The safe systems approach takes a holistic approach in that the responsibility for road safety is shared between all facets of the transportation system (that is, roadway infrastructure, road- way users, and vehicles). • The concept of travel time reliability focuses on designing a roadway in such a way that maxi- mizes the travel time reliability of the roadway. • The concept of VE is a systematic process of project review and analysis by a multidisciplinary team to provide recommendations for improving the value and quality of the project. • The concept of designing for 3R projects includes a set of geometric design criteria for 3R proj- ects that are less restrictive than the geometric design criteria in use for new construction and reconstruction. • The concept of designing for very low-volume local roads (≤ 400 ADT) recognizes that VLVLR represent a different design environment than higher-volume roads. Each process has its own focus, with the developers observing an existing issue and subse- quently developing a revised process to address their problem. Nonetheless, each process likely had its own challenges when responsible agencies were developing and implementing these approaches, or when other agencies sought to adopt the design concept. Acceptance and pro- motion of the final revised design process should identify and address challenges in the establish- ment and introduction of a revised process. The following sections provide an overview of likely

184 a performance-Based highway Geometric Design process challenges and barriers for any process change. Depending on the final process recommenda- tion, there may be unique challenges and barriers that will be further identified and addressed at that time. 8.4.2 Challenges to Implementing a Revised Geometric Design Process Challenges and barriers are expected for any revised process an industry or agency may under- take. Restraining forces may be concentrated within a small group of individuals or could be concerns shared by a large portion of the target population. Specific to a revised geometric design process, challenges have been identified and categorized in the following four areas: • Organizational and Institutional; • Risk Management; • Scalability for Owner, Roadway, and Project Size; and • Professional. A summary of the key challenges and barriers expected within each category is provided in the following sections. 8.4.3 Organizational and Institutional The most significant challenges and barriers to a revised process will likely occur in the area of organizational or institutional resistance. Reasons will span a wide range of viewpoints, but many will have a foundation in the cost of changing the way business is done. Therefore, it is important these concerns be identified and understood before establishing a plan to implement a revised design process. • At this stage, the revised design process is envisioned to use new and evolving processes that better quantify the safety performance of design alternatives, such as the HSM predictive models. As with any new model, there are potential data and data systems needs that an agency will be unable to meet (for example, a road and intersection inventory of elements used by models). Agencies may need to invest significant resources in developing and main- taining such data and systems. There is also the issue of maintaining this information on local and low-volume roads, where there is already limited information available. Initially, agencies may need to prioritize and identify which data elements are the most crucial to their design process and focus resources in these areas. • Agencies need staff experienced in performing operational and safety analyses that are envi- sioned to support a revised geometric design process. Having staff with these skill sets partici- pate in the project development process will be important to a successful design. Agencies will need to invest in training a complement of staff that can support the project development. For state DOTs, this might include having staff that can support local agency projects. • Design engineers may need to learn and accept new design criteria and models for their use. This could range from updating existing criteria based on outdated models, or potentially new analyses or models as well as processes for selecting criteria for projects based on the project’s context and purpose. • Institutional inertia, or an agency’s resistance to the change, can stem from multiple mis- perceptions about the revised processes. A common misperception would be that the revised approach increases the time and cost of the design process without any evident benefit or gain to the agency. There are also learned biases regarding design standards and design approaches that still exist (that is, the nominal safety mindset which is that a design is safe when all design criteria are met) despite that CSS, Complete Streets, and practical design have been in use for several years.

research and Knowledge Needs to Fully Implement the revised process 185 • The transition process will also be a challenge for most agencies. Needs regarding training, documentation, and process quality control will have to be addressed before the first project can be developed using the revised process. There will then be a point at which an agency must commit to using the revised process on all new projects, trusting that the implementation plan has been successful. • Most transportation agencies have different sections or bureaus that deal with the different functions. The design section is often separated from the traffic section that focuses on the operations and the planning section that develops the concept of the improvement being designed. The improved process breaks down these silos and provides for better communica- tion between the various functions. 8.4.4 Risk Management Agencies must be aware of risk management, starting in project development and through con- struction and O&M of their system. A revised geometric design process should not only allow design engineers to understand the potential implications of their decisions, but provide a process to docu- ment and defend decision making so not to put the state in a situation of unnecessary tort risk. • It was earlier noted that the revised processes may create additional need for data, analysis, or software. Especially early in the transition when agencies might have limited data, experience, or access to software, the revised process requires a framework for allowing design engineers to make decisions that can be defended in a tort case. • A revised geometric design process may result in new or different documentation to support the defense of an agency. Agencies should coordinate with legal staff to help determine which decisions need documentation and the type of documentation to maintain. • Through the decades, there have been many improvements to vehicles and their performance. While some of the geometric models are based on driver comfort, it might be possible to update the models taking into consideration the performance of modern vehicles. However, this will need to be weighed against the existing fleet’s characteristics, including the antique and classic cars still on the road. Therefore, a potential risk management issue is that the pro- cess should continue to design to the lowest common denominator or take advantage of the advances in automobile performance. • Prior to the recent efforts and research to expand the scientific knowledge on geometric ele- ments and safety performance, operational elements had been central to the design itself. If this were to remain the same with the suggested revised design process, then a potential liabil- ity issue could remain since the process doesn’t make use of the latest knowledge. Therefore, it is key for the future of risk management that the revised design process make full use of the expanding knowledge in highway and traffic safety. 8.4.5 Scalability for Owner, Roadway, and Project Size Changes to develop a data-driven process that relies on operational and safety analysis to inform the decision-making process could be a challenge for local agencies. Scaling the process not only to fit the context of the road environment, but also to the agency responsible for imple- mentation, is important. • Cities and counties with relatively smaller programs in comparison to the state DOT may initially struggle with having the technical expertise and software. At the outset of transition- ing local agencies to the revised process, the state DOT may need to support training, provide access to any developed software, and assist in project development for pilot projects. • Rural counties, townships, and small towns that predominantly operate lower-volume sys- tems may have challenges performing operational and safety analyses that are part of the

186 a performance-Based highway Geometric Design process revised geometric design process due to either a lack of available information, technical expertise, or access to new software. However, AASHTO already publishes a design policy for very low-volume rural local roads. Therefore, the resulting process may be able to identify a process simplified for lower-volume systems or establish a set of minimum criteria based on safety research. 8.4.6 Professional A geometric design process that results in significant changes could have a ripple effect at the academic and professional level. It will be important to identify and consider what changes could influence the profession in this category. • The revised geometric design process is envisioned to be accompanied by substantial re-write of AASHTO policies. This would be a multi-year effort, as was the case with the LRFD process. In addition to updating the AASHTO policies, states would need to review and update their own manuals, guides, and policies impacted by the revised design process. Updating state and local level documents will require additional time and cost as part of the implementation. • A revised process, if having substantial changes from the existing process, would require uni- versities and colleges to update their curricula to instruct the next generation of design engi- neers on the revised processes and train them to use the tools. Similarly, professional licensing and testing requirements may need updating. • A revised approach will require a significant effort of education targeted to both existing pro- fessionals and college students. Existing professionals may feel threatened or be resistant to change. For those accustomed to designing by pulling values from tables, the new process of designing to provide the best performance may be more complicated. • Existing college curricula does not always prepare students with all the knowledge to be an informed roadway designer. Some college instructors lack the informed background or expe- rience to help students become accomplished designers. Graduating students currently need training on the design tools as they enter the profession. This will continue to be the case with the revised process. 8.4.7 A State’s Experience Implementing a Flexible Design Process The Minnesota DOT undertook an initiative to provide for design flexibility in its project development process. The Design Flexibility Engineer at the Minnesota DOT shared several experiences and perspectives on this transition (AASHTO 2004). The transition process the Minnesota DOT relies on establishes a focus on education and outreach, provides technical expertise to design staff in project development, and updates policies and criteria. With this approach, it is important to note that the first two elements, when well executed, makes it easier for the design community to accept and embrace updated policies and criteria that constitute the design process. In order to accomplish these goals, two keys to the approach included: 1. Identify a staff person to work full-time to institutionalize the revised process. 2. Understand the design criteria, policies, and decisions that have the largest financial impact on the project cost. Also identify those which are the easiest to implement from either a tech- nical or design community acceptance perspective. As shown in Figure 64, where criteria, policies, or decisions have high financial impact on the project outcome, but are easy to address either technically or from a design community acceptance viewpoint, then addressing these areas first should provide the greatest return on the

research and Knowledge Needs to Fully Implement the revised process 187 agency’s investment. Several examples that may fit this category, depending on a state’s design approach, include: • Designing long turn lane lengths in constrained environments, especially if volumes are low enough to allow the intersection to operate properly with shorter turn lanes. • Avoiding excessive earth work (economical cross sections). For example, widening shoulders to standards on low-volume roads that have limited number of crashes, instead of using nar- row shoulders with edge line or shoulder rumble strips, which may provide a similar safety performance. • Interchange ramp designing, such as selecting large radius for loop ramps or designing long ramps, which may increase right-of-way purchases. One example of success that has occurred in Minnesota is the viewpoint on alternative inter- section designs. In Minnesota, the public and the design community both initially had reser- vations regarding the restricted crossing U-turn (RCUT) as an option for rural expressways. Through outreach and education efforts, including working with other states, this intersection type is now viewed as an alternative to constructing interchanges at locations experiencing severe crashes. The safety performance of the RCUT has so far proven superior to the traditional through-STOP intersection, but could be constructed at a fraction of the cost of an interchange. 8.4.8 Lessons Learned from Recent Process Changes Based on experiences through the implementation of LRFD, CSS, and the HSM, the follow- ing are offered as suggestions for a successful transition (AASHTO 2014, Neuman et al. 2002, AASHTO 2010). Other positive driving forces may exist and will be used in the transition, but the following have proven to be critical to success. Identify a Champion. Eventually, champions will need to be identified at numerous levels within each state DOT, including a liaison for technical transfer to the local and low-volume road authori- ties. However, the first step is a champion at the national level to communicate the need and benefit to key agencies, organizations (that is, FHWA and AASHTO), and executive managers at the state Fi na nc ia l I m pa ct High Low Easy HardEase of Implementaon (Technical or Acceptance) Largest Return on Investment Figure 64. Understanding financial impact and ease of implementation for individual design criteria and policies.

188 a performance-Based highway Geometric Design process DOTs. This is a top-down approach to building support within the industry, but the advantage of an executive sponsor is its ability to make clear the importance of the transition. Adequately Fund the Transition. As seen with the adoption of the LRFD process in bridge design, the training and software needs of the user community is crucial to minimizing indi- vidual resistance. Following the update of the AASHTO policies, lead agencies, such as FHWA and AASHTO, must develop and fund a program to help train highway designers. This will include developing the training, training instructors, and then making courses available to com- munity. In providing the training, it is important to consider the needs beyond the state DOTs. Local agencies, federal employees, and consultants will require the same training to make sure the revised process is fully implemented. Local technical assistance provider (LTAP) centers and professional societies can be used to help train these groups. A final aspect of the training is considering the universities and colleges. Schools that teach design principles to students should understand the revised process and begin teaching it to the next generation of designers. In addition to funding training, a revised geometric design process may require funding to develop supporting design software; the initial lack of supporting software was one reason noted for the resistance to LRFD. Therefore, any software envisioned to accompany a revised process should be available as quickly as possible. In developing the software, it will be important that it be affordable to the many local agencies, companies, and schools otherwise unable to afford expensive software, such as Safety Analyst. It also needs to be fully tested to make sure it is easy to use and compatible with the other software used by designers. Update Design Guides and Policies. AASHTO publishes several policies that most agencies use or base their own manuals, guides, and policies on. As a result of AASHTO’s policies becom- ing industry standards, these documents will need to be updated to make them consistent with the revised geometric design process. This could include a supplemental recommended process guide as a companion to the design policies, minor changes to the policies, or a completely new structure to the policies. Regardless, it is important that the new guides be in place if it is expected that agencies will be able to properly adopt and apply a revised geometric design process. Develop a Lead State Program. FHWA successfully used lead state programs in several areas, including the integration of the HSM. A lead state program with support from FHWA and AASHTO provides real-world guidance and results in positive peer pressure on agencies and individuals not embracing the change. A lead state program also provides an opportunity to develop documentation that supports a transition. Example documentation may include a data needs report, a manager’s guide, and an example implementation plan that other states might use to efficiently implement changes. Market Successes. Case studies are one approach to explaining and marketing benefits to the general audience. The message of each case study must clearly define and identify the benefits gained from the revised process. For example, a case study may demonstrate how the revised design process resulted in a different design decision that avoided a costly alternative. Equally important, a case study should be a project that after construction, the location has proven to operate efficiently and perform without major incident to demonstrate that there were minimal trade-offs in the design decision. Case studies can also highlight states that have been using performance-based or practical design processes, and the benefits these agencies have experienced. 8.5 Green Book Reorganization The AASHTO Green Book is the core reference document from which state DOT design manuals and policies are derived. Road and highway designers must be thoroughly versant in both the contents and basic philosophy of the Green Book.

research and Knowledge Needs to Fully Implement the revised process 189 The Green Book is updated about every 7 to 10 years. It evolved from a single policy docu- ment, to two documents encompassing separately rural and urban road design. Since the 1984 edition, when the two policies were combined into the one current policy, the basic structure of the Green Book has been largely unchanged. Of course, changes and additions to content have occurred (e.g., inclusion of geometric design of roundabouts) over the years. Yet, the basic for- mat, structure, and indeed, design philosophy and approach have been constant. A change in the approach to geometric design envisioned by this study would necessarily require a rethinking and restructuring of the Green Book. A primary driver is the separation of approaches for new vs. reconstruction projects. Other drivers include the more robust defini- tions of context discussed here, and the concept of differing models or assumptions for geomet- ric elements based on context. Appendix E contains a detailed outline of a future edition of the Green Book that would address the substantive advances presented by this research project. The outline, carried to a second head- ing level, is intended to facilitate a better understanding of the depth and breadth of necessary change to lay out a vision for a future Green Book and to demonstrate the practical implementa- tion of such changes. Presenting such an outline also confirms the need for AASHTO to remain at the center of design policy, and for the Green Book to continue its dominant leadership role. The following is a high level summary of the 13-page outline included in Appendix E. Note that the Green Book could be organized by four major parts. Note also the use of individual chapters for roads defined within the two-dimensional framework of functional classification and context zone. This structure allows for context design model or approach that is unique to such roads to be covered fully just in the one chapter. Although the outline appears lengthy, note that many designers would gravitate only to the specific chapters covering the project types they perform. 8.5.1 A Future Generation AASHTO Policy on Geometric Design Part I—Fundamentals of Geometric Design Chapter 1—Geometric Design and Project Development Chapter 2—The Geometric Design Framework Chapter 3—Road User Performance Characteristics Chapter 4—Elements of Geometric Design—Alignment and Cross Section Chapter 5—Elements of Geometric Design—Intersections and Roundabouts Chapter 6—Elements of Geometric Design—Interchanges and Interchange Ramps Chapter 7—Integration of Technology with Geometric Design Chapter 8—Overview of the Roadway Geometric Design Process Part II—Geometric Design Process for New Roads Chapter 9—New Construction Design Process Overview Chapter 10—New Local and Collector Roads in Rural Context Zones Chapter 11—New Arterial Roads in Rural Context Zones Chapter 12—New Freeways and Fully Controlled Access Highways in Rural Context Zones Chapter 13—New Local and Collector Roads in Suburban Context Zones Chapter 14—New Arterial Roads in Suburban Context Zones Chapter 15—New Freeways and Fully Controlled Access Highways in Suburban Context Zones Chapter 16—New Local and Collector Roads in Urban Context Zones Chapter 17—New Arterial Roads in Urban Context Zones Chapter 18—New Freeways and Fully Controlled Access Highways in Urban Context Zones

190 a performance-Based highway Geometric Design process Part III—Geometric Design Process for Roads to Be Reconstructed Chapter 19—Reconstruction Design Process Overview Chapter 20—Reconstructed Local and Collector Roads in Rural Context Zones Chapter 21—Reconstructed Arterial Roads in Rural Context Zones Chapter 22—Reconstructed Freeways and Controlled Access Facilities in Rural Context Zones Chapter 23—Reconstructed Local and Collector Roads in Suburban Context Zones Chapter 24—Reconstructed Arterial Roads in Suburban Context Zones Chapter 25— Reconstructed Freeways and Controlled Access Facilities in Suburban Context Zones Chapter 26—Reconstructed Local and Collector Roads in Urban Context Zones Chapter 27—Reconstructed Arterial Roads in Urban Context Zones Chapter 28—Reconstructed Freeways and Controlled Access Facilities in Urban Context Zones Part IV—Roads Requiring Resurfacing, Restoration, or Rehabilitation (3R) Chapter 29—3R Design Process for All Road Types and Contexts 8.6 Implications with Driverless/Connected/ Autonomous Technology Advances in vehicle technology are accelerating. Specifically, there is much research and development work on automated driver technology (also referred to as autonomous vehicle technology). The vision of many is that, at some point in the future, vehicles will have sufficient means to interact with the roadway environment and navigate in real time in response to that environment, thereby eliminating the human driver from the task of driving. The benefits of this vision are primarily the presumed elimination of the human element from crash risk. In theory, the existence of a network operated on by 100% autonomous vehicles would result in no crashes. Additional benefits may also accrue to improve the overall efficiency of the road system. Given the historic role of the human driver in the formulation and application of geometric design, such a future vision would seem to have radical implications in the field of roadway design. The entire concept of roadside design to mitigate run-off-road crash frequency and severity would be rendered unnecessary. Cross-section dimensions for lane and shoulder could be lessened considerably. Road capacity could be increased significantly through use of very small headways at high speeds. SSD design would be based on locations of sensors, and assume much more responsive and aggressive avoidance maneuvers, thus greatly shortening necessary distances. Indeed, one might postulate that the presence of a 100% autonomous vehicle fleet essentially negates much of what currently constitutes geometric design controls. It is beyond the scope of this project to delve further into the subject of autonomous vehicles, other than to offer the following observations: • Full implementation of the autonomous vehicle technology is at best decades away. Research continues. The complexity of the urban driving environment will take some time for the tech- nology to fully address. More importantly, it is not clear what the responsibilities of agencies will be to provide a roadway environment that is necessary for the technology to properly work. The potential liability and cost associated with a level of maintenance above that currently provided on all public roads is incalculable. • Even if/when such technology is sufficiently perfected, the existing vehicle fleet will remain on the system for 15 years or more. A very long transition period to full autonomous operation must occur, if indeed such operation is ever fully implemented.

research and Knowledge Needs to Fully Implement the revised process 191 • Societal and cultural issues need resolution. Specifically, does the presence of fully autonomous technology absolve a driver from any and all responsibility? If so, what are the limitations or requirements to obtain a driver’s license and what training is necessary? What should AASHTO’s approach be in considering the onset of autonomous vehicle tech- nology to the subject of geometric road design criteria? The research team’s suggestion is that, given the above significant issues, AASHTO must continue to adopt a basic design model that includes consideration of the human element and active human drivers as fundamental to design. However, to the extent that autonomous driver technology becomes more imbedded in the vehicle fleet, over time the potential benefits and impacts should be observable. This fact makes the performance-based approach to design all the more compelling. Over time, if fewer crashes occur of a given type (say, roadside) or headways reduce and capacity increases, these effects will be measured on the system. The tools used to determine design solutions (e.g., SPFs and CMFs) will evolve to include such effects. Cost effectiveness based on lower crash frequen- cies will produce different outcomes, such outcomes reflecting the evolution of the vehicle fleet. And those design elements insensitive to human driving limitations (e.g., vehicle offtracking) will be unchanged.