Chapter 5. Assessment and Solution

This chapter provides assessment of the current technical challenges of implementing traffic laws and regulations databases, identifies constraints and limitations, and recommends solutions from the perspective of design, deployment, and digital technology systems.

Assessment

It is possible to transform the guiding rules from the Uniform Vehicle Code (UVC) to requisites for automated driving system (ADS) behavior development. However, as explained above, many traffic rules vary by State and by the local authority. It will take a substantial effort to create a comprehensive database of State laws, even using existing compilations—for example, the American Automobile Association (AAA) Digest of Motor Laws—as a starting point.

New State Traffic Legislation

Currently, traffic laws are maintained and administered by the presiding jurisdiction. There is no sanctioned central repository for all traffic laws at all jurisdictional levels. The regulations database would need to support inclusion of all new traffic laws that go beyond UVC guidance. There will need to be some version/configuration control, such that it can be determined what version of traffic law is being provided to an ADS. There will need to be procedures to update the database as new legislation is enacted. Many States have adopted new traffic laws that were not necessarily relevant the last time the UVC was updated. For example, in 2019, Florida enacted Act 2019-44, Use of Wireless Communications Devices While Driving, which prohibits texting while driving and prohibits the use of handheld wireless communications devices in school crossings, school zones, or work zones. The law also makes these violations a primary offense. This requirement does not apply to a motor vehicle operator who is “Operating an autonomous vehicle, as defined in s. 316.003, in autonomous mode.”¹⁰⁰

Local Traffic Laws

The greatest limitation in creating and maintaining a comprehensive database of all traffic laws relates to local rules. Currently, there is no database that documents all local agency traffic laws. Most States require approval by only their State legislature or department of transportation (DOT) to change local traffic laws.

Automated Vehicle State Legislation

In addition to the variation of traffic laws by State, there is also variation and evolution in the autonomous vehicle enacted legislation or executive orders, as shown in Figure 8. The National Conference of State Legislatures (NCSL) regularly updates a database of these actions. Similar to the State traffic laws, the State automated vehicle (AV) legislation could be transformed into a database of requisites for ADS behavior development. This would also take a substantial effort to create and would require routine monitoring to update the database as new legislation is enacted.

As of the writing of this report, the original equipment manufacturer (OEM) is required to meet Federal Motor Vehicle Safety Standards (FMVSS), and the human driver must pass the State driver’s license requirements and comply with the rules of the road to maintain a license.

High-Level Needs

This section discusses high-level design and deployment needs of the proposed traffic regulation databases. The needs for building digital technology systems are also discussed in terms of system platforms, testing technologies, cybersecurity, and data privacy.

Design and Deployment

Regulation Data Model

Developing a data model for traffic regulations is essential, but will depend on potential user needs for validation. The model has to express the regulations in terms of settings and behaviors that AVs can perceive and act on for various roadway environments (e.g., work zones and adverse weather conditions). It will need to address both permissible vehicle behaviors and traffic controls and signs—for example, right turns permissible at a red traffic signal, except where signed as not permitted. It will need to address static and dynamic controls, whether by time of day (e.g., a school speed zone) or by local dynamic signal and signage (e.g., ramp metering signals or variable speed limits [VSL]). It will need to address cases where ADS has to depend on perception of physical markings and signage and where controls are provided by infrastructure-to-vehicle/vehicle-to-infrastructure (I2V/V2I) communications, as with a signal phase and timing (SPaT) message. And it will need to address incidents and work zones as well as normal operations.

Vehicle Function Regulations

Some traffic laws imply required vehicle equipment or functions. For example, some States require the use of windshield wipers when it rains, which implies that the vehicle must be equipped with wipers. This study assumes that the OEM responsible for the vehicle using the ADS will comply with FMVSS or exemption specifications for safety systems, equipment, and performance as NHTSA requires. However, FMVSS are being updated in response to ADS needsand OEMs will need to continue to track them as changes are made.

Regulations Data Access and Collection

As previously described, State traffic codes are readily available and generally able to be collected from the regulatory sources. Local government codes are unlikely to be so readily available and will present additional challenges for identification and collection. Data on regulatory traffic controls within a jurisdiction are significantly less available in database form, though generally observable in field deployments.

Data Format Limitations with Automated Driving System Behavior

Traffic codes are written for human interpretation and application and may face challenges in being adapted for ADS. This is an issue of identifying driving tasks and different roadway environments and scenarios in which the regulations apply, and specifying the operational limits associated with those tasks. For example, how would an ADS be expected to interpret “exercising due care” in avoiding pedestrians and other vulnerable users? Traffic controls are currently deployed as marked or signed indications along roadways. These controls provide data as images and words to be interpreted by human drivers as permitting or precluding specific driving behaviors. ADS will need to acquire the same information as controls derived from its own imaging systems, static controls associated with mapped locations, or static and dynamic control data provided in I2V exchanges. Some agencies may maintain sign or signal databases, but the mapped and dynamic control data are generally not currently available.

Coordination across Jurisdictions

Just as human drivers need to know the rules of the road in various jurisdictions through which they travel, ADS will need to be aware of its current location and the applicable regulations. In practice, much of the local variability will take the form of signed controls on the roadway. But there may still be a need for a higher level set of behavior principles—like no right turn on red—to be more generally encoded into the AV control algorithms.

Funding Issues

Obtaining and rendering traffic regulations into a database form are somewhat simplified by the general online availability of State traffic codes. There may be costs associated with the digest of the regulations into actionable ADS input. Costs of providing traffic control data to ADS are largely unknown at this point. Those costs will depend on yet-to-be-determined factors including the agency’s own controls databases, availability of local high-definition maps, and the agency’s intent for implementing dynamic controls to be provided over the air in real time.

Digital Technology Systems

Currently, the private industry has taken different approaches to addressing this issue. For example, in a speech at the Automated Vehicle Symposium 2019,¹⁰² Chris Urmson, cofounder and chief executive officer (CEO) of Aurora, emphasized the criticality of accounting for variations in traffic regulations, which may have an impact on ADS behavior and safety. One solution he mentioned is to include the information in Aurora’s high-definition maps developed for the on-road testing regions, and in some cases modified algorithms need to be implemented. Another solution, using a concept similar to this approach is a cloud-based database solution. The INRIX® Road Rules™ tool¹⁰³ is one such example that uses the SharedStreets open data format.¹⁰⁴ These kinds of tools, however, focus on points-of-interest data entry for the placement of control devices and other traffic related information such as signage, parking restrictions, and turn lanes. While it is a reductive task to create a Google Maps™ web mashup interface to manually enter municipal transportation features, the goal of this task is to explore ways to maximize automatic data sharing with regard to ADS mobility governance.

Which digital technology systems infrastructure owner-operators (IOO) may use to implement traffic regulation databases would depend on agency capability to address changing traffic environments, particularly considering that more traffic regulations can be dynamic in the future, and the proposed traffic regulation database will be closely tied to other components of dynamic traffic management (e.g., VSL systems). Emerging active demand and traffic management technologies and new data sources (e.g., those from connected and automated vehicles [CAV] and advanced perception and monitoring systems) are impacting agencies . They have realized the necessity of capacity building for fully leveraging the benefits of the new technologies and data sources. Agency capability in terms of technology systems is discussed from the following three perspectives: system platforms, technology deployment and testing, and cybersecurity and data privacy.

System Platforms

The technology supporting big data and cloud scale systems continues to evolve, offering many choices to agencies. Historically, the primary technology choices for agencies have been hosting data and systems on-premise, in the cloud, across multiple clouds, or using a hybrid approach; other possible choices include computational platforms:

• Cloud environments: Cloud environments may provide subscription-based services, including infrastructure services such as computing power, networking, and security and storage, and provide software services, such as databases, business applications, Internet of Things, and machine learning.
• Premise environments: Premise environments (data centers) are owned and operated by the agency or through contract services. Premise environments typically operate in one or more managed data centers utilizing networking, hardware, and software services to run business applications.
• Computational platforms: Computational platforms are stacks of integrated software and hardware useful for high-volume calculation and analytics. These are typically deployed as high-performance computing systems integrated with advanced models providing analytic and research capabilities.

The ability to deploy and migrate systems and data across a variety of platforms allows agencies to build solutions based on business needs without the constraint of utilizing a single cloud vendor or technology stack.

Technology Testing

Technology testing is an integral part of the deployment of new technologies. Technology testing is the process of analyzing a system or a component by providing defined inputs and comparing them with the desired outputs. Testing can be divided into two categories: manual testing or automated testing.

Manual testing, as the name suggests, is done manually and requires human input, analysis, and evaluation. Automated testing is the automated version of manual testing; using automation in testing helps avoid human errors, which can occur due to human fatigue of performing repeated processes. Automated testing programs will not miss a test by mistake. The automated test program will also provide the means of storing the test results accurately. The results can be automatically fed into a database, which can be used to provide necessary statistics on how the new data system is performing. Automated testing can detect errors in the database, which may have a major impact on ADS traffic law compliance and affect ADS behavior and safety.

Objectives of automated testing are as follows:

• Perform repetitive/tedious tasks to accurately reproduce tests.
• Validate requirements and functionality at various levels.
• Simulate multiple users exercising system functionality.
• Execute more tests in a short amount of time.
• Reduce test team head count.

Cybersecurity and Data Privacy

For any advanced data system, many ownership and maintenance (O&M) considerations involve management and security of data. The Federal Highway Administration (FHWA) Reliability Data Guide’s Data Ownership and Maintenance section¹⁰⁵ presents a sample list of fundamental considerations likely to govern O&M levels of effort and expense:

• Who will pay to collect, store, and share the data?
• Who (if anyone) can sell the data and to whom?
• What privacy issues in the data must be addressed?
• Who is allowed to access the data and what data are they permitted to access (e.g., all of the data, only a subset of the data)?
• For what purposes are the data allowed to be used (e.g., if collected for analysis purposes only, could they also be used for enforcement purposes)?

The Real-Time Data Capture and Management State of the Practice Assessment and Innovations Scan¹⁰⁶ addressed issues related to data capture, management, archiving, and sharing to encourage collaboration, research, and operational development and improvement. The scan documented the following best practices for access, security, and privacy:

• Generally, the holder of the data controls access to the data. Within the transportation and logistics community, this access is carefully controlled.
• Systems are in place that ensure data are accessed only by the intended people, and only to the degree those people need the data. Transportation and logistics industry data can be extremely sensitive, with disastrous consequences for a business if accessed by persons with malicious intentions.
• Usually, data access is password-protected. Because data generated within the logistics systems are often financial, they are strongly encrypted when sent. However, several applications can retrieve aircraft and vessel tracking data, often with other identifying information. The security clearance or password protection to access data through these applications is often minimal.
• The protection of data sources is extremely important. In the search engine industry, data sources are so heavily protected that there is not even disclosure of how, exactly, they are protected.

The scan documented the following best practices for data storage and backup:

• Frequent backups and off-site storage are typical.
• Preventative maintenance should be performed regularly.
• Careful consideration should be devoted to determining how much and for how long data should be stored. In aviation, for instance, data are kept for a relatively short time frame because the need is for real-time, rather than historical, information. At the same time, data can be available for revision if there is an incident to investigate.

The scan documented the following best practices for operations and maintenance:

• Deployment should begin on a reasonable scale, such as implementing in a small geographical area or using easily manageable data.
• Multiple servers should be used to distribute real-time loads. Several technologies enable this load distribution.
• It is important to consider determining the needed resolution or granularity of the data. This may vary depending on the context and use of the data.
• It is necessary to determine what is critical to communicate and what is not. For instance, railroad and airline alert systems only collect the necessary data that can alert an operator to a particular problem.

The scan documented the following best practice for critical failures:

• A common issue is that correcting a problem often depends on a single person, meaning that it depends on the person’s availability. Therefore, it is important to have staff available around the clock to solve potentially catastrophic failures. The higher labor cost is a necessary expense if the system needs to be highly available at all times.