Framework for Assessing Potential Safety Impacts of Automated Driving Systems (2022)

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.

3   This chapter describes the need for the framework, goals of the framework, target audience, and the technical definitions and background needed to understand subsequent discussions. Need for Framework Automated driving system (ADS) technologies are quickly advancing and are expected to have disruptive impacts on transportation safety in the coming years. ADSs will change planning, design, and operational criteria, which means there is a growing need for roadways that are traditionally planned, designed, and operated with human drivers in mind to begin to adapt to ADS. There will be opportunities and risks in planning and implementing ADS technologies for transporta­ tion agencies, technology firms and service providers, automakers, and research organizations. For a smoother transition to ADS transportation, state and local agencies must understand how and when traditional safety­related processes and procedures may be affected and have the tools to assess safety impacts of competing options. In some cases, ADSs may positively impact safety in ways that allow agencies to reprioritize investments, such as lane­marking maintenance. We need to understand better the safety impacts and give transportation stakeholders the tools they need to achieve safety goals. For example, some ADS technologies could mitigate certain crash types and severities while increasing the risk of others. Consequently, practitioners need a framework to use in current and future safety planning, design, operational decisions, and investments on multimodal infrastructure. This framework should translate the goals and objectives of state and local agencies into a hypothesis on how their policies and actions can influence safety. For instance, the primary goals of state and local agencies deploying a low­speed shuttle pilot may include improving safety and mobility. The hypothesis may be stated as “The rate of vehicle collisions among travelers going to and from transit station X during the pilot is less than before.” While this framework focuses on considering the safety performance of ADS under various scenarios, it provides the flexibility to account for other socio­ economic impact areas such as public health and safety (e.g., access to healthcare), environment (e.g., air quality), accessibility, equity, etc. For example, this report focuses on helping agencies explore questions such as “Will Level 4 low­speed automated shuttles improve safety performance for road users?” and does not focus on questions such as “Will Level 4 low­speed automated shuttles improve access to healthcare for persons with disabilities?” Goals of the Framework This framework addresses the following key questions: What are the key factors influencing ADS safety and how do they relate to planning, design, and operations decisions and tools? C H A P T E R 1 Introduction and Background

4 Framework for Assessing Potential Safety Impacts of Automated Driving Systems With limited ADS safety­related data available, how can the impacts of ADS technologies on safety be estimated and support decision­making? What steps can agencies take to achieve safety goals, meet user needs, and prioritize invest­ ments in ways that consider the impacts of ADS? Target Audience The primary audience of this framework includes transportation infrastructure owners and operators (IOOs), safety industry and advocacy groups, and ADS manufacturers. Federal agencies may find this useful to understand gaps in data collection, management, and analysis tools. The outputs from the framework may support building trust in road users in ADS technology, and therefore the outputs have been framed in a way that is accessible to a broad audience. ADS devel­ opers and manufacturers may benefit from data sources and analysis methods to understand the safety impacts of the technology they seek to commercially deploy. For example, this framework may help industry determine and track target safety levels for commercial deployments without safety drivers. It also may help private and public sector stakeholders build public trust in deploy­ ments by quantifying safety impacts. Technical Background This section provides the technical background needed for subsequent discussions. Dynamic Driving Task (DDT): Includes all real­time operational and tactical functions to operate a vehicle in on­road traffic, excluding the strategic functions (e.g., trip scheduling, selection of destinations and waypoints) and including the following (SAE, 2018): A. Lateral vehicle motion control via steering. B. Longitudinal vehicle motion control via acceleration and deceleration. C. Monitoring the driving environment via object and event detection, recognition, classifi­ cation, and response preparation. D. Object and event response execution. E. Maneuver planning. F. Enhancing conspicuity via lighting, signaling, gesturing, etc. Automated Driving System (ADS): The hardware and software that are collectively capable of performing the entire DDT on a sustained basis. This term is used specifically to describe a Level 3, 4, or 5 driving automation system (SAE, 2018). The different levels of automation are described in Figure 5 in chapter 2. Operational Design Domain (ODD): Operating conditions under which a given driving automation system or feature is specifically designed to function, including environmental, geographical, and time­of­day restrictions, and/or the requisite presence or absence of cer­ tain traffic or roadway characteristics (SAE, 2018). ODD is typically defined by the ADS technology developer and original equipment manufacturer (OEM). More details on the ODD are provided in Chapter 3, Overview of the Framework Elements. DDT Fallback: This occurs when the ADS is unable to continue to perform the entire DDT (i.e., under normal operating conditions). For Level 3 ADS features, the human fallback­ ready user is expected to respond to a request to intervene by either resuming manual driving if the vehicle remains drivable or achieving a minimal risk condition if the vehicle is not drivable. For a Level 4 or 5 ADS, the feature or system performs the fallback by auto­ matically achieving a minimal risk condition (SAE, 2018). Radio Detection and Ranging (Radar): Radar is a range­finding technology that supports per­ ception. Radars operate by transmitting a radio signal toward a region of interest and detecting

Introduction and Background 5   the signals reflected back from objects within the field of view. Radar is a popular choice for automated vehicles (AVs) because it is relatively inexpensive and robust (Patole et al., 2017). Light Detection and Ranging (Lidar): Lidar is a subset of radar and has been growing as a key enabling technology for AVs. Lidar allows generation of high­definition (HD), three­ dimensional (3D) maps by sending and receiving high­frequency radar. Lidar works similar to radar: it transmits a wave (in this case, light) and detects the reflected light pulse from an object within the detectable region. Lidar has a much higher resolution and frequency (900–1,500­nm wavelengths) (Yole Développement, 2015). Communications Vehicle-to-Everything (V2X) Communication: It is a compendium of V2X communications occurring over the dedicated short­range communication (DSRC) or cellular spectrum to provide vehicle­to­vehicle (V2V), vehicle­to­infrastructure (V2I), vehicle­to­pedestrian (V2P), and vehicle­to­network (V2N) communications. An onboard unit (OBU) enables the vehicles to communicate with other vehicles, infrastructure, pedestrians, and cellular network around them to enhance safety, mobility, and environmental aspects of driving. In V2I communica­ tion, the OBU communicates with a roadside unit (RSU) to dispatch important information, such as hazardous road conditions (Joseph, 2018). Dedicated Short-Range Communication (DSRC): The term “dedicated” refers to the fact that the Federal Communications Commission (2002) dedicated 75 MHz of licensed spectrum in the 5.9 GHz band for DSRC, though part of the spectrum is now shared with unlicensed Wi­Fi users and the rest is shared simultaneously with cellular technologies (discussed next). DSRC takes place over hundreds of meters, a shorter distance than other common wireless communications. While the main purpose for deploying DSRC was a collision­prevention application, DSRC has characteristics (e.g., low latency, high reliability, security, and interoperability) that make it ideal for many other applications beyond colli­ sion avoidance (Kenney, 2011). Additionally, DSRC experiences little interference, even in extreme weather conditions, due to its short range, making it ideal for handling communications to and from cars moving at high speeds. Cellular-V2X (C-V2X): C­V2X is a wireless broadcast interface that permits a single platform for V2V, V2N, and V2I communication. C­V2X can operate within a dedicated frequency band for low­latency use cases (5.9 GHz) or use more traditional connectivity channels (Qualcomm, 2019). Object and Event Detection and Response (OEDR): The subtasks of the DDT that include monitoring the driving environment (detecting, recognizing, and classifying objects and events and preparing to respond as needed) and executing an appropriate response to such objects and events (i.e., as needed to complete the DDT and/or DDT fallback) (SAE, 2018). HD Maps: These types of maps are designed and made for self­driving cars and AV features operating at Levels 3 and 4. These maps have extremely high precision (to the centimeter level) because the cars need precise instructions on how to maneuver within a particular lane along the route (Vardhan, 2017).