Signal Timing Manual - Second Edition (2015)

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.

Chapter 3. Signal Timing Concepts CHAPTER 3 SIGNAL TIMING CONCEPTS CONTENTS 3.1 TRAFFIC SIGNAL BASICS ..................................................................................................... 3-2 3.1.1 Common Signal Components and Interactions ............................................................ 3-2 3.1.2 Basic Signal Controller Concepts ........................................................................................ 3-4 3.2 INITIAL SIGNAL TIMING CONSIDERATIONS ................................................................ 3-5 3.2.1 Deine the Operating Environment ................................................................................... 3-6 3.2.2 Identify Users .............................................................................................................................. 3-9 3.2.3 Establish User and Movement Priorities ....................................................................... 3-10 3.3 DATA COLLECTION ............................................................................................................. 3-10 3.3.1 Field Visits .................................................................................................................................. 3-11 3.3.2 Trafic Counts ........................................................................................................................... 3-12 3.3.3 Other Trafic Characteristics .............................................................................................. 3-16 3.3.4 Intersection Geometry and Trafic Control Devices ................................................. 3-16 3.3.5 Existing Signal Timing ........................................................................................................... 3-18 3.3.6 Crash History ............................................................................................................................ 3-18 3.4 OPERATIONAL OBJECTIVES AND PERFORMANCE MEASURES ............................ 3-19 3.4.1 Select Operational Objectives ............................................................................................ 3-20 3.4.2 Establish Performance Measures ..................................................................................... 3-21 3.4.3 Non-Vehicle Operational Objectives and Performance Measures ...................... 3-28 3.5 REFERENCES ......................................................................................................................... 3-29 Signal Timing Manual, Second Edion

Chapter 3. Signal Timing Concepts LIST OF EXHIBITS Exhibit 3-1 Chapter Outline Using the Outcome Based Process ....................................... 3-1 Exhibit 3-2 Common Signalized Intersection Components ................................................ 3-3 Exhibit 3-3 Signalized Intersection Interactions ..................................................................... 3-3 Exhibit 3-4 Types of Signal Control ............................................................................................... 3-4 Exhibit 3-5 Initial Signal Timing Considerations ..................................................................... 3-5 Exhibit 3-6 Roadway Classiƒications ............................................................................................. 3-7 Exhibit 3-7 Individual Intersection ............................................................................................... 3-7 Exhibit 3-8 Arterial Intersections .................................................................................................. 3-8 Exhibit 3-9 Grid Network Intersections ...................................................................................... 3-8 Exhibit 3-10 Multimodal Environment .......................................................................................... 3-9 Exhibit 3-11 Data Collection ............................................................................................................ 3-11 Exhibit 3-12 Example Vehicle Trafƒic Volume Proƒiles and Time-of-Day Cycle Lengths ........................................................................................................................... 3-14 Exhibit 3-13 Trafƒic Count Data ...................................................................................................... 3-15 Exhibit 3-14 Example Trafƒic Volume Data ............................................................................... 3-15 Exhibit 3-15 Example Intersection Condition Diagram ....................................................... 3-17 Exhibit 3-16 Example Signalized Intersection Crash Diagram .......................................... 3-19 Exhibit 3-17 Performance Measures, Timing Parameters, and Collection Methods ......................................................................................................................... 3-22 Exhibit 3-18 Example Split Termination Logging ................................................................... 3-25 Exhibit 3-19 Example Split Termination Logging Visualization ....................................... 3-25 Exhibit 3-20 Arterial Efƒiciency Decreased by Signal Timing ............................................ 3-25 Exhibit 3-21 Storage Bay Blocking ................................................................................................ 3-26 Exhibit 3-22 Storage Bay Spillback ............................................................................................... 3-26 Exhibit 3-23 Conƒlicts between Movements at an Intersection ........................................ 3-27 Exhibit 3-24 Red-Light Violation ................................................................................................... 3-28 Exhibit 3-25 Non-Vehicle Operational Objectives and Performance Measures ......... 3-29 Signal Timing Manual, Second Edion

Chapter 3. Signal Timing Concepts LIST OF EXHIBITS Exhibit 3-1 Chapter Outline Using the Outcome Based Process ....................................... 3-1 Exhibit 3-2 Common Signalized Intersection Components ................................................ 3-3 Exhibit 3-3 Signalized Intersection Interactions ..................................................................... 3-3 Exhibit 3-4 Types of Signal Control ............................................................................................... 3-4 Exhibit 3-5 Initial Signal Timing Considerations ..................................................................... 3-5 Exhibit 3-6 Roadway Classiƒications ............................................................................................. 3-7 Exhibit 3-7 Individual Intersection ............................................................................................... 3-7 Exhibit 3-8 Arterial Intersections .................................................................................................. 3-8 Exhibit 3-9 Grid Network Intersections ...................................................................................... 3-8 Exhibit 3-10 Multimodal Environment .......................................................................................... 3-9 Exhibit 3-11 Data Collection ............................................................................................................ 3-11 Exhibit 3-12 Example Vehicle Trafƒic Volume Proƒiles and Time-of-Day Cycle Lengths ........................................................................................................................... 3-14 Exhibit 3-13 Trafƒic Count Data ...................................................................................................... 3-15 Exhibit 3-14 Example Trafƒic Volume Data ............................................................................... 3-15 Exhibit 3-15 Example Intersection Condition Diagram ....................................................... 3-17 Exhibit 3-16 Example Signalized Intersection Crash Diagram .......................................... 3-19 Exhibit 3-17 Performance Measures, Timing Parameters, and Collection Methods ......................................................................................................................... 3-22 Exhibit 3-18 Example Split Termination Logging ................................................................... 3-25 Exhibit 3-19 Example Split Termination Logging Visualization ....................................... 3-25 Exhibit 3-20 Arterial Efƒiciency Decreased by Signal Timing ............................................ 3-25 Exhibit 3-21 Storage Bay Blocking ................................................................................................ 3-26 Exhibit 3-22 Storage Bay Spillback ............................................................................................... 3-26 Exhibit 3-23 Conƒlicts between Movements at an Intersection ........................................ 3-27 Exhibit 3-24 Red-Light Violation ................................................................................................... 3-28 Exhibit 3-25 Non-Vehicle Operational Objectives and Performance Measures ......... 3-29 Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-1 CHAPTER 3. SIGNAL TIMING CONCEPTS This chapter provides an overview of signal timing basics, organized using the outcome based process introduced in Chapter 1 (and shown again in Exhibit 3-1). The outcome based process is a modern approach to signal timing and relects the complex nature of many signalized intersections and operating environments. It encourages practitioners to consider all system users (e.g., pedestrians, bicycles, motor vehicles, emergency vehicles, transit, and rail) and to establish user priorities by movement for each signalized intersection location. The focus is applying those elements to the selection of operational objectives that clearly deine desired signal timing outcomes. In order to assess the effectiveness of a signal timing plan, performance measures must also be identiied for each objective. Performance measures can be used to determine initial success, as well as to monitor and sustain desired outcomes throughout the life of a signal. Trafic signal timing adjustments should be part of a continual process of adapting to changing conditions, not the result of retiming once in a long while. Before discussing the outcome based process, this chapter introduces the basic components of a trafic signal system. Modern trafic signal controllers are essentially computers that accept requests for service and control user displays based on the rules established by the practitioner. The speciic timing parameters have a direct relationship to service priority. Therefore, it is necessary to understand the basic Operang environment, users, and movements are the building blocks for developing a signal ming plan. Exhibit 3-1 Chapter Outline Using the Outcome Based Process Signal Timing Manual, Second Edion

Chapter3-2 3. Signal Timing Concepts operation of a modern trafic signal controller before assessing various alternative timing strategies and values. The second part of this chapter explains the signal timing outcome based process. As shown in Exhibit 3-1, detailed information about the eight steps is divided among several chapters. The irst ive steps essentially deine desired outcomes and are discussed in this chapter. Chapters 5–7 provide the details on how to develop timing strategies, intersection timing parameters, and system timing parameters. Chapter 8 discusses the critical steps of implementation, initial observation, long-term monitoring, and maintenance. The last step is part of continual improvement, which feeds back to earlier steps as needed. Not only can operational objectives be maintained through such efforts, but the necessity of a signal timing program (discussed in Chapter 2) may become more apparent. For example, if a previous reduction in maintenance resulted in increased stops, delay, and fuel consumption, then a stronger case may be made for increased maintenance moving forward. 3.1 TRAFFIC SIGNAL BASICS A few terms warrant introduction because of the potential for multiple interpretations. Terms are generally deined when irst used, but a glossary is available at the end of the manual. • User: A speciic category of persons receiving service at a trafic signal. Users include pedestrians, bicycles, passenger cars, trucks, emergency vehicles, and transit. Users (which in aggregate constitute the “trafic” at a signalized intersection) are the primary focus of the outcome based process. • Vehicle: A general term that encompasses all devices, without regard for the user, including passenger cars, bicycles, trucks, and buses. Traditionally, vehicles have been the primary unit of measure for most trafic signal analysis (e.g., vehicle volume or vehicle delay). Vehicles are sometimes categorized as light (passenger cars and light trucks) and heavy (heavy trucks and buses), and in order to account for their operational differences, vehicle equivalents are commonly applied (i.e., passenger car equivalents for heavy vehicles). In this manual, the term vehicle is used when it is necessary to include multiple user types in the description of trafic signal operations (e.g., a green indication serving passenger cars, trucks, bicycles, and buses) without regard for their speciic needs. • Movement: A term that describes user actions at an intersection (e.g., northbound vehicular left turn or pedestrian using the west crosswalk). Movements can be permitted, requiring users to yield to others when given a green indication, or protected, which gives users the right-of-way without any conlicts. • Practitioner: The person determining signal timing. 3.1.1 Common Signal Components and Interacons Each signalized intersection has common components that provide the basis for signal operations (illustrated in Exhibit 3-2): a controller, a cabinet, displays (or indications), and, typically, detection. Exhibit 3-3 depicts the basic interaction among these signalized elements and the system users. It is important for a practitioner to This chapter explains the steps leading up to the selecon of signal ming values. There are many inial pieces of informaon that must be considered if a signal ming plan is to be effecve. Signal Timing Manual, Second Edion

Chapter 3. Signal Timing Concepts 3-3 recognize the relationship among users, detection, and the signal controller, as well as how that relationship inluences the resultant displays (i.e., vehicle displays, pedestrian displays, and occasionally special displays for speciic users) to ultimately inform users of their right-of-way. Only a basic understanding of the components is required for this chapter. However, it should be noted that in Exhibit 3-2, the vehicle displays (typically for passenger cars, trucks, and buses) also control the bicycles, and the intersection may or may not have bicycle detection and timing to control the shared vehicle displays. As shown in Exhibit 3-3, the user is part of a continuous process. The detectors (e.g., vehicle, pedestrian, or bicycle) send messages to the controller as users approach the intersection. The controller then uses the detector inputs to change the user displays (typically vehicle and pedestrian) based on the signal timing parameters deined by the practitioner. If many users approach the intersection at once, the process becomes more complicated, and the controller must make decisions (based on the timing parameters that control user priorities) about which movements will receive the right-of- way. The way that the controller assigns time to users is highly dependent on the detection and signal timing parameters (controller settings) that have been programmed. Chapter 4 discusses the critical design considerations related to signal hardware (i.e., cabinets, controllers, and detection) that affect the type of operation that is feasible. Exhibit 3-2 Common Signalized Intersecon Components Exhibit 3-3 Signalized Intersecon Interacons Signal Timing Manual, Second Edion

Chapter3-4 3. Signal Timing Concepts 3.1.2 Basic Signal Controller Concepts Signal controllers (speciically the internal program called irmware) are an essential component of trafic signal control. While the remainder of this manual summarizes various functionalities, parameters, and values that can be programmed into signal controllers, this section only introduces basic controller concepts. There are two deining characteristics of signal controllers: (1) how the signal controller interprets demand at the local intersection and (2) how the signal controller relates to other controllers. 3.1.2.1 Interpretaon of Local Demand Signal controllers can adjust operations using external information about user demand (i.e., requests for service). It is the type of external information that the controller utilizes that deines the type of signal control (as outlined in Exhibit 3-4). Fully-actuated signal control is fully adaptive to local trafic conditions because it utilizes demand information from detectors (at least vehicle and pedestrian) located on all approaches. Fully-actuated intersections offer the most lexibility. Trafic signal controllers that utilize detection only in some lanes are considered semi-actuated because they can only adapt operations based on partial demand information. A controller that does not use any detection to adapt operations is considered pretimed. Modern pretimed trafic signal systems are used in some special cases. Pretimed systems are the least costly to build because they do not have the expense associated with detection. They use signal timing values that were calculated and programmed into the controller based on historical data. These types of systems are typically used in central business districts (CBDs) with closely spaced intersections (often in grid networks), where detection has little potential to improve operations given the need to maintain well-deined relationships between intersections. 3.1.2.2 Relaonship to Other Controllers Regardless of the type of detection being used, signal controllers can operate either alone or as part of a system. Uncoordinated signal timing allows the intersection to run independently (or “free”) of any other intersection, while coordinated timing operates several signals as a system. It is often believed, incorrectly, that detection cannot be used to inluence the coordinated phases when an intersection is coordinated. Modern controllers can actuate a portion of the coordinated phases, providing additional lexibility. Detailed information about coordinated timing is available in Chapter 7. Even if operated as part of a system, the local intersection is still an active part of operations. Traditional trafic signal systems (that use coordination to minimize stops Exhibit 3-4 Types of Signal Control Signal Timing Manual, Second Edion

Chapter 3. Signal Timing Concepts 3-5 for through trafic) impose a common cycle length to maintain a consistent relationship between adjacent trafic signals. Coordination essentially constrains the local trafic signal controller to a timing plan that will achieve the operational objective of corridor progression (resulting in fewer stops along the corridor). Adaptive systems are an alternative to traditional coordination and use different algorithms to adjust signal timing parameters. Adaptive systems still rely on local controllers for many timing parameters and share most features of traditional systems, but they have more lexibility in how they adjust timing parameters. Adaptive systems that have advanced system capabilities are discussed in Chapter 9. Adaptive systems should only be considered after the capabilities of traditional systems have been fully considered. Chapters 5–7 focus on traditional trafic signal control, which is currently used for the vast majority of systems and is more than adequate for most locations. 3.2 INITIAL SIGNAL TIMING CONSIDERATIONS This section will give the practitioner a general idea of the type of information that should be gathered for making informed signal timing decisions. Exhibit 3-5 summarizes the basic information that a practitioner might investigate as part of the irst three steps of the outcome based process. Detailed information about each of these topics is provided in the following sections. Outcome Based Process Inial Signal Timing Consideraons STEP 1 Define the Operang Environment Mul- Jurisdiconal Impacts □ Is the system of signals located in a single or mulple jurisdicons? □ If mulple, is signal ming performance consistent across jurisdiconal boundaries? □ Is there an exisng agreement in place that defines certain signal ming parameters? □ Would an agreement between jurisdicons to establish consistent signal ming strategies be beneficial? Locaon and Associated Environment □ Is the intersecon or system of intersecons located in a rural, suburban, or urban area? □ Is the area transional or undergoing changing land uses? Roadway Classificaon □ How are the roadways classified (major street and minor street) where the signal system is located (e.g., freeway interchange, major arterial, minor arterial, major collector, minor collector, or local street)? □ Are there specific pedestrian, bicycle, freight, or transit route needs? □ How does the classificaon affect user expectaons of the facility? Transportaon Network □ How closely spaced are the signalized intersecons? □ Is there a reason to consider mulple intersecons as a system when developing the signal ming? □ Is the minor street coordinated? □ Is there nearby rail (freight or passenger) requiring preempon? STEP 2 Idenfy Users □ What is the exisng mix of users (e.g., pedestrians, bicycles, light vehicles, heavy vehicles including trucks and transit vehicles, priority vehicles, and rail including freight and passenger)? □ Does the mix change by me of day? □ Are there unique travel pa”erns? STEP 3 Establish User and Movement Priories □ Who are the crical users at the intersecon(s)? □ Does the crical user change by me of day? □ Does the jurisdicon have any policies related to user priories? □ What are the crical movements? Exhibit 3-5 Inial Signal Timing Consideraons Signal Timing Manual, Second Edion

Chapter3-6 3. Signal Timing Concepts 3.2.1 Define the Operang Environment The irst step in the outcome based process is deining the operating environment. Signals timed together as a system should have a common operating environment so that the practitioner can work from consistent priorities and objectives when developing signal timing plans. 3.2.1.1 Mul-Jurisdiconal Impacts In some instances, a system of signals may operate across multiple jurisdictions (e.g., a system located along a corridor that runs from within a city’s jurisdiction to another city’s, county’s, or state’s jurisdiction). If a system is close to or across a jurisdictional boundary, the operating agency should coordinate with all affected jurisdictions to discuss and develop an agreement on objectives and various signal timing parameters, as appropriate. System users will not expect signal timing performance to vary across jurisdictional boundaries, so every effort should be made to time the signals in a way that allows a seamless transition. Maintaining consistent timing values (e.g., cycle length and clearance times) across jurisdictional boundaries will help users know what to expect. For example, a consistent cycle length will allow cross-jurisdictional coordination, thereby reducing stops. The practitioner should verify which signal timing values he or she can change and which should only be adjusted within the jurisdiction agreement. 3.2.1.2 Locaon and Associated Environment One of the primary factors affecting signal timing is the environment at the intersection (or system of intersections). In particular, the practitioner will want to identify whether the intersections are located in a rural, suburban, or urban environment because each location will require the practitioner to consider different priorities for timing objectives. Signal timing objectives for intersections located in rural environments generally focus on the existence and use of setback detection, as well as minimizing the number of drivers experiencing decision zones. Decision zones (and associated detection) are described in detail in Chapter 4, but are locations where different drivers may make different decisions, resulting in potential conlicts. Signal timing in suburban areas often focuses on achieving smooth low (minimizing stops) along arterials. This can be achieved through strategies that include coordinating intersections and actuating the pedestrian movements (if pedestrian crossing demand is low during certain or all times of day). Timing objectives may change by time of day to relect changing trafic patterns, and having good detection allows the most lexibility. The primary focus for signal timing in downtown urban environments is typically accommodating all user groups (e.g., pedestrians, bicycles, passenger cars, and transit). In order to manage vehicle queues, a shorter cycle length may be chosen that relects the shorter distance between intersections. Other multimodal strategies include setting the progression speed based on bicycle travel speeds, setting the recall mode based on pedestrian priority, and deining offsets to prioritize transi t vehicle operations. Conducng a site visit and/or gathering informaon (i.e., local knowledge) is important for correctly characterizing the operang environment and validang desired outcomes. Signal ming plans should be transparent to users across jurisdiconal boundaries. The objecves for signal ming will be different depending on if a signal system is located in a rural, suburban, or urban environment. Objecves may also change by me of day or by traffic volume. Signal Timing Manual, Second Edion

Chapter 3. Signal Timing Concepts 3-7 3.2.1.3 Roadway Classificaon The roadway facility within the larger transportation system can inluence user expectations, which in turn can inluence how the roadway should be timed. There is the traditional tradeoff between mobility and access (depicted in Exhibit 3-6), but other factors can impact signal timing settings as well, including • Freight route designations, • Key pedestrian crossings, • Signalized multi-use trail crossings, • Bike boulevards, and • Rapid or high-capacity transit routes. A practitioner should understand the roadway classiications of the transportation network and, equally important, user expectations. 3.2.1.4 Transportaon Network Characteriscs The coniguration of the transportation network can have a signiicant impact on the way its trafic signals are timed. The practitioner should consider the effect that nearby intersections will have on one another based on their spatial relationship as well as the desired speed of trafic. Individual signalized intersection operations (depicted in Exhibit 3-7 and covered in detail in Chapter 6) typically occur with long signalized intersection spacing (a half mile or more) or when other factors (such as late-night operation with low trafic volumes) make independent operations preferable. Individual signalized intersections can be timed without the explicit consideration of other trafic signals, allowing the lexibility to set signal timing parameters that are optimal for the individual intersection. This type of operation is often called “free” (or “isolated”) because it operates without need for or beneit from coordination. In these cases, good detection on all approaches is necessary for high operational and safety performance. It should be noted that a signalized intersection may operate both in a free and a coordinated manner depending Exhibit 3-6 Roadway Classificaons Exhibit 3-7 Individual Intersecon Individual intersecons are somemes referred to as “isolated,” which refers to a mode of operaon, rather than a spaal relaonship. Individual intersecons may also be described as “free” operaon, which indicates they are not currently being coordinated. Source: Adapted from A Policy on Geometric Design of Highways and Streets (1) Signal Timing Manual, Second Edion

Chapter3-8 3. Signal Timing Concepts on trafic conditions. It is not necessary for a signalized intersection to operate in the same manner at all times. For signalized intersections located along arterial streets with high volumes of through trafic (depicted in Exhibit 3-8), individual signalized intersection operations can usually be improved by considering coordination of the major street movements. For most arterial streets with signals that are spaced a half mile apart or less, coordinated operations can yield beneits by improving progression between signals. On arterials with higher speeds and/or limited access, it can be beneicial to coordinate signals spaced a mile apart or even further. Additional guidance on coordination is available in Chapter 7. Signalized intersections can also be located in grid networks (depicted in Exhibit 3-9). Grid networks are more dificult to time because tradeoffs have to be made in order to coordinate the intersections. In these cases, the entire network is often timed together to ensure consistent behavior among intersections. Grid networks, particularly in downtown environments with short block spacing, are frequently timed using pretimed plans (i.e., all phases on vehicle and pedestrian recall) because the same amount of time is generally desired each cycle to maintain consistent relationships between the closely spaced intersections. Signals that are located very close together (less than 500 feet or 7 to 10 seconds apart) often require signal timing settings and strategies that manage queues (typically by keeping cycle lengths low); however, pedestrians may be a driving force in the timing. In some small, closed networks with closely spaced intersections (e.g., grid of four closely spaced, one-way streets, triangle created by three intersections, or closely spaced offset-T intersections), it may be beneicial to operate intersections with a single controller to tightly control the relationship among all movements. Geometric constraints associated with any type of transportation network (i.e., a lack of turn lanes, short turn-lane storage bays, or driveway locations) can have important impacts on operations. Signal timing strategies should relect these location- speciic considerations. Potential strategies are discussed in later chapters. Exhibit 3-8 Arterial Intersecons Exhibit 3-9 Grid Network Intersecons Signal Timing Manual, Second Edion

Chapter3-8 3. Signal Timing Concepts on trafic conditions. It is not necessary for a signalized intersection to operate in the same manner at all times. For signalized intersections located along arterial streets with high volumes of through trafic (depicted in Exhibit 3-8), individual signalized intersection operations can usually be improved by considering coordination of the major street movements. For most arterial streets with signals that are spaced a half mile apart or less, coordinated operations can yield beneits by improving progression between signals. On arterials with higher speeds and/or limited access, it can be beneicial to coordinate signals spaced a mile apart or even further. Additional guidance on coordination is available in Chapter 7. Signalized intersections can also be located in grid networks (depicted in Exhibit 3-9). Grid networks are more dificult to time because tradeoffs have to be made in order to coordinate the intersections. In these cases, the entire network is often timed together to ensure consistent behavior among intersections. Grid networks, particularly in downtown environments with short block spacing, are frequently timed using pretimed plans (i.e., all phases on vehicle and pedestrian recall) because the same amount of time is generally desired each cycle to maintain consistent relationships between the closely spaced intersections. Signals that are located very close together (less than 500 feet or 7 to 10 seconds apart) often require signal timing settings and strategies that manage queues (typically by keeping cycle lengths low); however, pedestrians may be a driving force in the timing. In some small, closed networks with closely spaced intersections (e.g., grid of four closely spaced, one-way streets, triangle created by three intersections, or closely spaced offset-T intersections), it may be beneicial to operate intersections with a single controller to tightly control the relationship among all movements. Geometric constraints associated with any type of transportation network (i.e., a lack of turn lanes, short turn-lane storage bays, or driveway locations) can have important impacts on operations. Signal timing strategies should relect these location- speciic considerations. Potential strategies are discussed in later chapters. Exhibit 3-8 Arterial Intersecons Exhibit 3-9 Grid Network Intersecons Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-9 3.2.2 Idenfy Users User characteristics clearly influence the effectiveness of signal timing. In particular, the mix of users and their trafic patterns will affect how an intersection operates (as depicted in the multimodal environment in Exhibit 3-10). The following users should be considered during the timing process: • Pedestrians. Pedestrians with slower walking speeds (e.g., children and elderly), persons with mobility limitations, and pedestrians with visual impairments need more time to cross the street. Pedestrian walk times and clearance intervals (discussed in Chapter 6) may need to be adjusted to relect local conditions and/or local policies. A leading pedestrian interval may also be warranted when high pedestrian volumes are competing with a high number of turning vehicles. In all cases, timing should, at a minimum, meet requirements in the Manual on Uniform Trafic Control Devices (MUTCD, 2). • Bicycles. High bicycle use at an intersection may warrant special bicycle detection and associated bicycle minimum green times or extension times. Timing parameters related to phase initiation and phase termination (i.e., minimum green, yellow change, and red clearance) are critical to bicycles (see Chapter 6 for more details) and may need to be adjusted to account for their lower speeds and acceleration characteristics. • Light Vehicles. The number of light vehicles (i.e., passenger cars and light trucks) using an intersection will impact many signal timing parameters (discussed throughout Chapters 6 and 7), including cycle length, the time allocated to each phase, and the order of the phases. • Heavy Vehicles. Truck trafic requires accounting for slower acceleration and longer deceleration times and the larger size of vehicles, which can inluence queue storage and detection setting needs. Truck priority (discussed in Chapter 10) should be considered at high-speed or downhill approaches to intersections with high volumes of truck trafic. • Emergency Vehicles. Emergency vehicles may justify preferential treatment through the use of preemption and/or priority (discussed in Chapter 10). • Transit Vehicles. Transit vehicles may justify special phasing and preferential treatment through the use of preemption and/or priority (see Chapter 10). The mix of users and traffic paerns at an intersecon will influence operaons. Exhibit 3-10 Mulmodal Environment Signal Timing Manual, Second Edion

Chapter3-10 3. Signal Timing Concepts 3.2.3 Establish User and Movement Priori es Operating environment and intersection users play a key role in how an intersection (or system of intersections) functions. The practitioner should identify user priorities for the signal timing plan(s), in conjunction with any relevant jurisdictional standards and policies. Priorities may be relative (e.g., pedestrians over vehicles) or absolute (e.g., railroad preemption). For example, if pedestrian volumes are high, pedestrians may be prioritized over vehicles during certain times of the day. In order to establish pedestrian priority during times with high right-turning vehicle trafic, the signal timing plan could provide a leading pedestrian interval (i.e., pedestrians get a walk indication before vehicles get a green indication). In addition to user priorities, the practitioner should also consider movement priorities. Critical movements can be determined using very simple analyses (e.g., critical movement analysis) or through the use of a variety of trafic operations software. It is important for the practitioner to understand that a software package may not automatically generate the desired solution. The practitioner may need to make adjustments to match the desired outcomes. By observing the critical movements at each intersection, the practitioner can understand how the intersection is currently operating and any potential changes that should be included in the timing plan development (e.g., more time allocated to the minor street to reduce phase failures). Giving priority to any one user group or movement requires tradeoffs for other users and movements. If the priority is arterial vehicular through movements, the priority for smooth low (or minimization of arterial stops) for those vehicles might mean more delay for minor street vehicles and pedestrians crossing the arterial. Therefore, before selecting user and movement priorities, it is important to collect data at the intersection(s) and identify patterns, including • Proportion of turning vehicles, • Corridor directionality, • Peaking characteristics, • Lane usage, • Weekday and weekend characteristics, • Typical and atypical trends (e.g., incident, crash, construction, game, concert, shift change, or convention), and • Origin-destinations. 3.3 DATA COLLECTION An effective assessment of the operating environment, users, and priorities includes data collection at the intersection(s). Required data can typically be separated into the following categories: trafic characteristics, intersection geometry, trafic control devices, existing signal timing, and crash history. If an intersection does not yet exist, then estimates will need to be developed, but, if possible, the practitioner should conduct site visits at the existing intersections during multiple trafic demand scenarios. The following sections provide detailed explanations of the data that can be collected to aid in signal timing development, speciically for the initial steps in the outcome based process. Exhibit 3-11 summarizes the types of information that may be User and movement priories should be established prior to development of signal ming plans. User and movement priories will influence many signal ming parameters, including the cycle length, me allocated to each phase, and phase order. Signal Timing Manual, Second Edi on

Chapter3-10 3. Signal Timing Concepts 3.2.3 Establish User and Movement Priori es Operating environment and intersection users play a key role in how an intersection (or system of intersections) functions. The practitioner should identify user priorities for the signal timing plan(s), in conjunction with any relevant jurisdictional standards and policies. Priorities may be relative (e.g., pedestrians over vehicles) or absolute (e.g., railroad preemption). For example, if pedestrian volumes are high, pedestrians may be prioritized over vehicles during certain times of the day. In order to establish pedestrian priority during times with high right-turning vehicle trafic, the signal timing plan could provide a leading pedestrian interval (i.e., pedestrians get a walk indication before vehicles get a green indication). In addition to user priorities, the practitioner should also consider movement priorities. Critical movements can be determined using very simple analyses (e.g., critical movement analysis) or through the use of a variety of trafic operations software. It is important for the practitioner to understand that a software package may not automatically generate the desired solution. The practitioner may need to make adjustments to match the desired outcomes. By observing the critical movements at each intersection, the practitioner can understand how the intersection is currently operating and any potential changes that should be included in the timing plan development (e.g., more time allocated to the minor street to reduce phase failures). Giving priority to any one user group or movement requires tradeoffs for other users and movements. If the priority is arterial vehicular through movements, the priority for smooth low (or minimization of arterial stops) for those vehicles might mean more delay for minor street vehicles and pedestrians crossing the arterial. Therefore, before selecting user and movement priorities, it is important to collect data at the intersection(s) and identify patterns, including • Proportion of turning vehicles, • Corridor directionality, • Peaking characteristics, • Lane usage, • Weekday and weekend characteristics, • Typical and atypical trends (e.g., incident, crash, construction, game, concert, shift change, or convention), and • Origin-destinations. 3.3 DATA COLLECTION An effective assessment of the operating environment, users, and priorities includes data collection at the intersection(s). Required data can typically be separated into the following categories: trafic characteristics, intersection geometry, trafic control devices, existing signal timing, and crash history. If an intersection does not yet exist, then estimates will need to be developed, but, if possible, the practitioner should conduct site visits at the existing intersections during multiple trafic demand scenarios. The following sections provide detailed explanations of the data that can be collected to aid in signal timing development, speciically for the initial steps in the outcome based process. Exhibit 3-11 summarizes the types of information that may be User and movement priories should be established prior to development of signal ming plans. User and movement priories will influence many signal ming parameters, including the cycle length, me allocated to each phase, and phase order. Signal Timing Manual, Second Edi on Chapter 3. Signal Timing Concepts 3-11 helpful to a practitioner when determining objectives and performance measures. In many instances, the data described herein may have been collected previously as part of a signal warrant analysis or trafic operations study. Care should be taken when reusing older data to understand and account for its potential inaccuracy. Some trafic data should be regularly collected at the same location along a corridor (typically at a mid- block location free of peak-period queuing) in order to track changes in trafic demand. Data Collecon Category Potenal Data to Collect Field Data □ Photos and/or video of the intersec
on(s) □ Observa
ons from adop
ng the role of different types of users □ Upstream and downstream bo‚leneck considera
ons □ Observa
ons about opera
ons watched from a sta
onary posi
on □ Poten
al impacts from
ming changes Traffic Counts 24-Hour Weekly Counts □ Tube counts at cri
cal loca
ons □ Vehicle classifica
ons Movement Counts □ Pedestrian volumes by crosswalk □ Bicycle volumes by intersec
on approach and movement □ Light vehicle volumes by intersec
on approach and movement □ Heavy truck volumes by intersecon approach and movement □ Transit volumes by intersecon approach and movement Other Traffic Characteriscs □ Vehicular speed □ Travel me □ Queues □ Vehicular delay by movement Intersecon Geometry and Traffic Control Devices □ Travel lanes □ Storage bays □ Pedestrian crosswalks □ Intersecon widths □ Prohibited turning movements □ One-way streets □ Approach grades □ Sight-line restricons □ On-street parking □ Loading zones □ Transit stops □ Profile grade □ Intersecon skew angle □ Adjacent land uses □ Nearby access points □ Exisng vehicular and pedestrian signal displays □ Control equipment □ Exisng detectors and associated detector channels Exisng Signal Timing □ Phase sequence □ Limitaons in the phase sequence due to geometric issues □ Use of overlaps □ Yellow change □ Red clearance □ Minimum green □ Maximum green □ Walk interval □ Flashing don’t walk interval □ Passage me □ Detector se“ngs □ Condional service □ Time-of-day plans □ Coordinated phase(s) □ Cycle length □ Splits and/or force-offs □ Offsets □ Offset reference point Crash History □ Crash type □ Crash severity □ Environmental collisions □ Causal factors 3.3.1 Field Visits Observing the study area during the expected timing plan periods (e.g., weekday a.m. peak, weekday p.m. peak, and weekend midday peak) will be informative for the development of signal timing. As part of the
ield visit, the practitioner should • Take photos and/or video of the intersection operations. Exhibit 3-11 Data Collecon Field visits are an essenal part of developing quality signal ming plans. Signal Timing Manual, Second Edion

3-12 Chapter 3. Signal Timing Concepts • Adopt the role of the different types of intersection users: o Motorized Vehicle Driver: Drive the corridor making a sample of movements (e.g., through, left turn, and minor street). o Pedestrian: Walk the intersection. o Bicyclist: Bike the intersection. o Transit User: Access nearby stations and ride transit. • Observe the critical intersection movements, progression interactions, and user characteristics and interactions/conlicts in a stationary position over multiple cycle lengths. • Consider upstream and downstream bottlenecks that may be inluencing trafic demand. • Think ahead to the impact of timing changes. Particularly at intersections near capacity, minor changes in signal timing can have major impacts (both positive and negative) and should be based upon an informed decision. Multi-day observations are desirable to validate signal timing plans and their effectiveness at accomplishing planned outcomes/operational objectives. 3.3.2 Traffic Counts Often referred to as “counts,” the quantiication of trafic volumes by user type is an important fundamental measure that guides signal timing development for both intersections and networks. There are generally two types of counts conducted as part of a timing project: 24-hour weekly counts and peak-period movement counts by user type. Preferably, these count data are available on an ongoing basis or, at a minimum, represent multiple time periods and days of the week suficient to relect system variability. However, these counts only represent a snapshot in time, so it is advantageous to continuously monitor trafic volumes at a congestion-free, mid-block location. This type of monitoring allows a practitioner to track volume changes over time, as well as determine whether 24-hour and peak-period counts are representative of seasonal and yearly trends. The practitioner should be aware that counts may be inaccurate due to data collection errors, and it is possible for trafic to vary signiicantly from day to day, week to week, and month to month. Too often, data are entered into a model and the results used without the practitioner verifying how slight volume changes will affect operations. For example, coordination is often run well into the evening under the assumption that late-night trafic patterns are the same as those during the peak period, when that might not be the case depending on the operating environment. It cannot be overemphasized that trafic counts are, at best, a snapshot in time, and need to be applied with judgment and understanding of the effects on signal operations. An important limitation of counts taken at trafic signals is that they may not represent actual demand. This is a major mistake that is very common in practice. Trafic demand is the total number of users desiring to reach a certain point in a network. Trafic volume is the total number of users that can reach that certain point in the network. One way to measure demand is to set an intersection in free operation (with adequate maximum green times to clear standing queues) and count the volume of trafic. This technique can only be used to measure demand when there is adequate Twenty-four-hour weekly counts and peak-period turning movement counts should be collected as part of signal ming development. Traffic volumes do not represent traffic demand if queuing exists. Signal Timing Manual, Second Edion

3-12 Chapter 3. Signal Timing Concepts • Adopt the role of the different types of intersection users: o Motorized Vehicle Driver: Drive the corridor making a sample of movements (e.g., through, left turn, and minor street). o Pedestrian: Walk the intersection. o Bicyclist: Bike the intersection. o Transit User: Access nearby stations and ride transit. • Observe the critical intersection movements, progression interactions, and user characteristics and interactions/conlicts in a stationary position over multiple cycle lengths. • Consider upstream and downstream bottlenecks that may be inluencing trafic demand. • Think ahead to the impact of timing changes. Particularly at intersections near capacity, minor changes in signal timing can have major impacts (both positive and negative) and should be based upon an informed decision. Multi-day observations are desirable to validate signal timing plans and their effectiveness at accomplishing planned outcomes/operational objectives. 3.3.2 Traffic Counts Often referred to as “counts,” the quantiication of trafic volumes by user type is an important fundamental measure that guides signal timing development for both intersections and networks. There are generally two types of counts conducted as part of a timing project: 24-hour weekly counts and peak-period movement counts by user type. Preferably, these count data are available on an ongoing basis or, at a minimum, represent multiple time periods and days of the week suficient to relect system variability. However, these counts only represent a snapshot in time, so it is advantageous to continuously monitor trafic volumes at a congestion-free, mid-block location. This type of monitoring allows a practitioner to track volume changes over time, as well as determine whether 24-hour and peak-period counts are representative of seasonal and yearly trends. The practitioner should be aware that counts may be inaccurate due to data collection errors, and it is possible for trafic to vary signiicantly from day to day, week to week, and month to month. Too often, data are entered into a model and the results used without the practitioner verifying how slight volume changes will affect operations. For example, coordination is often run well into the evening under the assumption that late-night trafic patterns are the same as those during the peak period, when that might not be the case depending on the operating environment. It cannot be overemphasized that trafic counts are, at best, a snapshot in time, and need to be applied with judgment and understanding of the effects on signal operations. An important limitation of counts taken at trafic signals is that they may not represent actual demand. This is a major mistake that is very common in practice. Trafic demand is the total number of users desiring to reach a certain point in a network. Trafic volume is the total number of users that can reach that certain point in the network. One way to measure demand is to set an intersection in free operation (with adequate maximum green times to clear standing queues) and count the volume of trafic. This technique can only be used to measure demand when there is adequate Twenty-four-hour weekly counts and peak-period turning movement counts should be collected as part of signal ming development. Traffic volumes do not represent traffic demand if queuing exists. Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-13 upstream storage for the queue and there is no downstream congestion blocking departing traf ic. For under-capacity conditions, demand is equivalent to the measured traf ic volume. However, if more vehicles arrive for a movement than can be served, the movement is considered to be operating over capacity (also called oversaturation). When an intersection movement (or movements) approaches or arrives at capacity, then additional delays occur in the network. Additional information on oversaturation is presented in Chapter 12. Unless the practitioner measures demand arriving at the intersection through queue observation (vehicles unserved at the end of green) or measurement of departure rates from an upstream under-capacity (also called under-saturated) intersection, the true demand at an over-capacity intersection may be unknown. This can cause problems when developing signal timing plans. For example, time may be added to a given movement, only to have it used up by the previously unserved demand and possibly transfer the over-capacity problem to another location. For more information on collecting traf ic volumes, see the Federal Highway Administration Trafic Monitoring Guide and Trafic Monitoring Guide Supplement (3, 4). 3.3.2.1 Twenty-Four-Hour Weekly Counts Weekly traf ic volumes should be collected at critical locations along the study corridor using temporary road tubes, system detectors, or detectors at signalized intersections in place along the corridor. Using existing system detectors or intersection detectors will reduce the time and cost of the data collection effort, but the sources should be validated in order to understand the accuracy of the resulting data. Volume pro iles developed from the weekly counts are an important element in the data collection effort and are used to identify • The number of timing plans that should be used during the weekdays and weekend, • When to transition from one timing plan to the next, • Directional distribution of traf ic along the corridor, and • Times to collect peak-period turning-movement counts. Exhibit 3-12 illustrates how traf ic volumes can change over a 24-hour period. In this case, traf ic peaks in the morning and evening. There are two cycle lengths that have been selected at this location—a longer cycle length of 110 seconds during the peak periods (7:00 a.m. to 10:00 a.m. and 3:00 p.m. to 8:00 p.m.) and a shorter cycle length of 85 seconds between the peak periods (10:00 a.m. to 3:00 p.m.). Generally, it is best to start the longer cycle length before it is needed. This allows the transition to inish before the longer cycle length is required and gives the intersection time to recover from any negative impacts of transition (described in detail in Chapter 7) (5). Note that in the example in Exhibit 3-12, the intersection is operating fully-actuated during the late night and early morning. Signal Timing Manual, Second Edion

3-14 Chapter 3. Signal Timing Concepts 3.3.2.2 Movement Counts Volumes from movement counts are used to assess the existing and future operations at an intersection. Movement counts are typically collected at each study intersection during representative trafic periods, which can be identiied based on the daily trafic volume proiles (discussed in the previous section). Depending on the trafic volumes and trafic patterns along the corridor, movement counts are often only conducted during peak periods, commonly the weekday morning, midday, and evening peak time periods. For other time periods (i.e., weekend or off-peak), the daily trafic volume proiles can be used in combination with the peak-period movement counts to develop volume adjustment factors, which can be used to estimate the volumes for any time period. At some locations, the trafic volumes may be higher or more critical during the weekend time periods, which could warrant performing movement counts for both weekday and weekend time periods. Additionally, seasonal trafic patterns may need to be considered and incorporated in the scope of work. Procedures for conducting, recording, and summarizing trafic counts are described in the Manual of Transportation Engineering Studies (6). Exhibit 3-13 presents categories of trafic count data that are typically collected as part of movement counts as well as data that should be collected, if possible, with the understanding that not all information is needed or useful in all scenarios. In general, trafic counts should document volumes, categorized by intersection approach (e.g., northbound or southbound), movement, and mode. If heavy vehicles are a signiicant portion of the vehicular trafic, the trafic count may need to be further categorized by vehicle classiication. The presence of special users at an intersection (e.g., elderly pedestrians, school children, or emergency vehicles) should also be documented. Exhibit 3-14 shows an example of turning-movement count data, including vehicle Exhibit 3-12 Example Vehicle Traffic Volume Profiles and Time-of- Day Cycle Lengths Signal Timing Manual, Second Edion

3-14 Chapter 3. Signal Timing Concepts 3.3.2.2 Movement Counts Volumes from movement counts are used to assess the existing and future operations at an intersection. Movement counts are typically collected at each study intersection during representative trafic periods, which can be identiied based on the daily trafic volume proiles (discussed in the previous section). Depending on the trafic volumes and trafic patterns along the corridor, movement counts are often only conducted during peak periods, commonly the weekday morning, midday, and evening peak time periods. For other time periods (i.e., weekend or off-peak), the daily trafic volume proiles can be used in combination with the peak-period movement counts to develop volume adjustment factors, which can be used to estimate the volumes for any time period. At some locations, the trafic volumes may be higher or more critical during the weekend time periods, which could warrant performing movement counts for both weekday and weekend time periods. Additionally, seasonal trafic patterns may need to be considered and incorporated in the scope of work. Procedures for conducting, recording, and summarizing trafic counts are described in the Manual of Transportation Engineering Studies (6). Exhibit 3-13 presents categories of trafic count data that are typically collected as part of movement counts as well as data that should be collected, if possible, with the understanding that not all information is needed or useful in all scenarios. In general, trafic counts should document volumes, categorized by intersection approach (e.g., northbound or southbound), movement, and mode. If heavy vehicles are a signiicant portion of the vehicular trafic, the trafic count may need to be further categorized by vehicle classiication. The presence of special users at an intersection (e.g., elderly pedestrians, school children, or emergency vehicles) should also be documented. Exhibit 3-14 shows an example of turning-movement count data, including vehicle Exhibit 3-12 Example Vehicle Traffic Volume Profiles and Time-of- Day Cycle Lengths Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts counts by approach and movement, heavy-vehicle percentages, pedestrian counts, and bicycle counts. Typical Traffic Count Data Collected Addional Traffic Count Data Collected If Possible □ Pedestrian volumes by crosswalk □ Bicycle volumes by movement □ Light vehicle volumes (throughput, not necessarily demand) by movement □ Heavy-vehicle (trucks and buses) volumes by movement □ Queuing at beginning of first count period and end of each count period □ Vehicle volumes by lane (i.e., lane u‚liza‚on) □ Transit volumes (e.g., percent buses, light rail, and streetcar) □ Number of pedestrian actua‚ons (i.e., ra‚o of cycles with pedestrian service to total cycles) □ Number of preemp‚ons and/or disrup‚ng occurrences (e.g., trains or emergency vehicles) □ Number of signal priority requests and services (e.g., transit or truck) □ Number of slow accelera‚ng vehicles at/near front of queue □ Satura‚on flow rate or car-following/driver behavior characteris‚cs (qualita‚ve or quan‚ta‚ve) If possible when conducting movement counts (whether done in the ield or using video), the presence of queues should be noted at the start of the counts and at the end of every counting period (typically every 15 minutes). If counts are performed using Exhibit 3-13 Traffic Count Data Exhibit 3-14 Example Traffic Volume Data 3-15 Signal Timing Manual, Second Edion

3-16 Chapter 3. Signal Timing Concepts video, the practitioner should review the associated video to check the quality of the counts; understand the actual trafic demand if greater than the capacity; and evaluate the inluences of various modes, nearby bus stops, driveways, on-street parking, and weather conditions that may affect intersection performance (see Chapter 11 for weather-related issues). Turning-movement counts are often one of the more costly items in a trafic signal timing analysis. FHWA published Signal Timing on a Shoestring (7), which provides some guidance on ways to minimize the data collection effort through a short-count method or by estimating turning-movement counts using link trafic volume data. These guidelines may be useful on corridors or networks where the trafic conditions are predictable. However, when developing timings where trafic patterns are rapidly changing, it may be more appropriate to collect 2- to 3-hour turning-movement counts for the various study time periods. 3.3.3 Other Traffic Characteriscs Other trafic characteristics (e.g., speed, travel time, and queues), in addition to trafic volumes, may be beneicial to the signal timing plan development process. Not only will these data be useful when developing the signal timing plans, but these trafic characteristics can also be used as performance measures to evaluate the effect of signal timing changes. Performance measures are explained in detail in Section 3.4. 3.3.4 Intersecon Geometry and Traffic Control Devices Intersection geometry can inluence the order of movements at an intersection, as well as the amount of time given to those movements. The signal timing practitioner should take an active role in new intersection design or intersection improvements, as the best signal timing plan cannot overcome bad geometrics. A site survey should be conducted to record relevant geometric information (using a condition diagram like the example shown in Exhibit 3-15), including • Number of lanes (regular and special purpose, including transit and bicycle), • Speed limits, • Lane widths, • Lane assignment (e.g., exclusive left turn, through only, or shared through/right turn), • Presence of storage bays, • Length of storage bays, • Length of pedestrian crosswalks, • Intersection widths for all approach legs, • Presence of prohibited turning movements, • Presence of one-way streets, • Approach grades, • Sight-line restrictions (buildings and foliage), • Presence of on-street parking, Some traffic characteriscs can be used to guide the selecon of ming values as well as be used to compare the before-and-aer effect of signal ming plans. Signal Timing Manual, Second Edion

3-16 Chapter 3. Signal Timing Concepts video, the practitioner should review the associated video to check the quality of the counts; understand the actual trafic demand if greater than the capacity; and evaluate the inluences of various modes, nearby bus stops, driveways, on-street parking, and weather conditions that may affect intersection performance (see Chapter 11 for weather-related issues). Turning-movement counts are often one of the more costly items in a trafic signal timing analysis. FHWA published Signal Timing on a Shoestring (7), which provides some guidance on ways to minimize the data collection effort through a short-count method or by estimating turning-movement counts using link trafic volume data. These guidelines may be useful on corridors or networks where the trafic conditions are predictable. However, when developing timings where trafic patterns are rapidly changing, it may be more appropriate to collect 2- to 3-hour turning-movement counts for the various study time periods. 3.3.3 Other Traffic Characteriscs Other trafic characteristics (e.g., speed, travel time, and queues), in addition to trafic volumes, may be beneicial to the signal timing plan development process. Not only will these data be useful when developing the signal timing plans, but these trafic characteristics can also be used as performance measures to evaluate the effect of signal timing changes. Performance measures are explained in detail in Section 3.4. 3.3.4 Intersecon Geometry and Traffic Control Devices Intersection geometry can inluence the order of movements at an intersection, as well as the amount of time given to those movements. The signal timing practitioner should take an active role in new intersection design or intersection improvements, as the best signal timing plan cannot overcome bad geometrics. A site survey should be conducted to record relevant geometric information (using a condition diagram like the example shown in Exhibit 3-15), including • Number of lanes (regular and special purpose, including transit and bicycle), • Speed limits, • Lane widths, • Lane assignment (e.g., exclusive left turn, through only, or shared through/right turn), • Presence of storage bays, • Length of storage bays, • Length of pedestrian crosswalks, • Intersection widths for all approach legs, • Presence of prohibited turning movements, • Presence of one-way streets, • Approach grades, • Sight-line restrictions (buildings and foliage), • Presence of on-street parking, Some traffic characteriscs can be used to guide the selecon of ming values as well as be used to compare the before-and-aer effect of signal ming plans. Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-17 • Presence of loading zones, • Presence of transit stops, • Pro ile grade near the intersection, • Intersection skew angle, • Adjacent land uses, and • Access points near the intersection. It is also important to record information about the existing traf ic control devices, in order to determine the capabilities, limitations, and functionality of the signal equipment. Necessary information includes • Number, location, and type of existing vehicular, pedestrian, bicycle, and transit signal displays; • Type of control equipment (requires agency coordination for access to cabinet); • Number, location, length, type, and condition of existing detectors (vehicular, pedestrian, bicycle, transit, and railroad); and • Number of detector channels associated with the detection for each traf ic movement. Exhibit 3-15 Example Intersecon Condion Diagram Signal Timing Manual, Second Edion

3-18 Chapter 3. Signal Timing Concepts 3.3.5 Exisng Signal Timing The existing signal timing can help the practitioner understand what currently exists in the ield and provide a baseline for improvement. Key information to obtain from the existing signal timing (explained in detail throughout Chapters 5 and 6) includes • Phase sequence (ring-and-barrier diagram), • Limitations in the phase sequence due to geometric issues, • Use of overlaps, • Yellow change and red clearance intervals, • Minimum green and maximum green, • Pedestrian walk and lashing don’t walk (FDW) intervals, • Passage time, • Detector settings, • Time-of-day plans, and • Special features. If the intersection is operating in coordination, then the following information (explained in detail in Chapter 7) will be required in addition to that listed above: • Coordinated phase(s), • Cycle length, • Splits and/or force-offs, • Offsets, and • Offset reference point (it is critical to understand what the controller actually uses as its offset reference, so that it can be correctly related to the optimization method used). If observations are made that reveal the existing signal timing plan is working adequately, then this should be noted. Many locations have traf ic patterns that may be stable and require few signal timing changes, except those related to changes in recommended practices (such as clearance intervals). The existing timing may have evolved over time in response to site-speci ic conditions. This initial review may result in focusing efforts on limited issues or even not undertaking timing changes, except when speci ic problems are identi ied. 3.3.6 Crash History Signal timing is one of the many factors that may contribute to crashes, so practitioners should be aware of the potential impacts that signal timing changes can have on safety. The practitioner should obtain 3 to 5 years of crash data; summarize the crashes by type, severity, and environmental conditions; and prepare a collision diagram of the crashes to help identify trends (as shown in Exhibit 3-16). If the signal ming isn’t broken, don’t try to fix it. Signal Timing Manual, Second Edion

3-18 Chapter 3. Signal Timing Concepts 3.3.5 Exisng Signal Timing The existing signal timing can help the practitioner understand what currently exists in the ield and provide a baseline for improvement. Key information to obtain from the existing signal timing (explained in detail throughout Chapters 5 and 6) includes • Phase sequence (ring-and-barrier diagram), • Limitations in the phase sequence due to geometric issues, • Use of overlaps, • Yellow change and red clearance intervals, • Minimum green and maximum green, • Pedestrian walk and lashing don’t walk (FDW) intervals, • Passage time, • Detector settings, • Time-of-day plans, and • Special features. If the intersection is operating in coordination, then the following information (explained in detail in Chapter 7) will be required in addition to that listed above: • Coordinated phase(s), • Cycle length, • Splits and/or force-offs, • Offsets, and • Offset reference point (it is critical to understand what the controller actually uses as its offset reference, so that it can be correctly related to the optimization method used). If observations are made that reveal the existing signal timing plan is working adequately, then this should be noted. Many locations have traf ic patterns that may be stable and require few signal timing changes, except those related to changes in recommended practices (such as clearance intervals). The existing timing may have evolved over time in response to site-speci ic conditions. This initial review may result in focusing efforts on limited issues or even not undertaking timing changes, except when speci ic problems are identi ied. 3.3.6 Crash History Signal timing is one of the many factors that may contribute to crashes, so practitioners should be aware of the potential impacts that signal timing changes can have on safety. The practitioner should obtain 3 to 5 years of crash data; summarize the crashes by type, severity, and environmental conditions; and prepare a collision diagram of the crashes to help identify trends (as shown in Exhibit 3-16). If the signal ming isn’t broken, don’t try to fix it. Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-19 After reviewing historical crash data, a site visit can help the practitioner identify causal factors. For example, ield observations may indicate that drivers in some movements are more aggressive than others or that certain movements are consistently running red lights. These characteristics may contribute to an increased number of crashes and could be the result of signal timing parameters (e.g., insuficient green time, poor offset, or long cycle length). The practitioner should evaluate the effect of parameters such as • Clearance intervals; • Detector settings, size, and locations; • Cycle length; • Offsets; and • Phase sequence. A variety of tests and other evaluation tools are discussed further in FHWA’s Signalized Intersections: An Informational Guide (8). 3.4 OPERATIONAL OBJECTIVES AND PERFORMANCE MEASURES Operational objectives and performance measures are at the heart of the outcome based process. The objectives and their associated performance measures tangibly deine user priorities by movement for the location being timed. Instead of choosing signal timing values based solely on software outputs (such as vehicle delay), the outcome based process develops signal timing values based on objectives and performance measures that have been selected for a speciic location. Note that some performance measures should be simple and understandable to system users, such as stops per mile and travel time. Others, such as system delay, may be helpful to system operators but of little interest to users. Exhibit 3-16 Example Signalized Intersecon Crash Diagram Signal Timing Manual, Second Ediƒon

3-20 Chapter 3. Signal Timing Concepts 3.4.1 Select Operaonal Objecves By clearly establishing objectives, a practitioner will be able to choose timing values that relect user needs. Objectives may be chosen based on known problems in the study area, as a result of public comments, staff observations, or known discrepancies with established policies. The identiication of problems during timing analysis may be an iterative exercise done in conjunction with determining objectives. In other words, previous problems may help shape the operational objectives and/or, by clearly deining operational objectives, deiciencies may be more apparent and easily addressed. There are many possible objectives for trafic signal operations, and signal timing strategies will change depending on the objective(s) that are chosen. The practitioner should understand that some objectives may be mutually exclusive. For example, objectives that focus on pedestrians will require different signal timing solutions than objectives that focus on vehicles. Below is a list of 15 specific operational objectives that can be used individually or in combination to focus signal timing efforts. Not all are useful in conveying performance in an understandable manner to the traveling public. 3.4.1.1 Vehicle-Specific Objec ves 1. Vehicle Safety: Minimize vehicle collisions, reduce vehicle conlicts, and provide suficient time for vehicles to execute movements. 2. Vehicle Mobility—Capacity Allocation: Serve vehicle movements as eficiently as possible, while also distributing capacity as fairly as possible across movements and modes. Prioritize movements according to need without excessively delaying other movements. 3. Vehicle Mobility—Corridor Progression: Move vehicles along high-priority paths (typically along high-volume movements on corridors) as eficiently as possible without excessively delaying other movements. 4. Vehicle Mobility—Secondary Progression: Move vehicles along low-priority paths (such as major street turning movements or minor street movements) as eficiently as possible without excessively delaying other movements. 5. Environmental Impact Mitigation: Minimize the amount of pollution induced by improving the eficiency of vehicle trajectories, either by reducing delay or stops or using infrastructure-to-vehicle technology to inluence vehicle travel patterns. 6. Queue Length Management: Prevent formation of excessive queues on critical lane groups, such as freeway exit ramps. 7. Vehicle and Driver Costs: Minimize stops and delay in order to reduce vehicle operating costs and driver delay costs. 3.4.1.2 Pedestrian-Specific Objec ves 1. Pedestrian Safety: Minimize pedestrian involvement in collisions, reduce pedestrian conlicts, and provide suficient time for pedestrians to execute movements. 2. Pedestrian Mobility: Serve pedestrian movements as eficiently as possible. Objec ves will help focus the goals for a signal ming project. Signal Timing Manual, Second Edion

3-20 Chapter 3. Signal Timing Concepts 3.4.1 Select Operaonal Objecves By clearly establishing objectives, a practitioner will be able to choose timing values that relect user needs. Objectives may be chosen based on known problems in the study area, as a result of public comments, staff observations, or known discrepancies with established policies. The identiication of problems during timing analysis may be an iterative exercise done in conjunction with determining objectives. In other words, previous problems may help shape the operational objectives and/or, by clearly deining operational objectives, deiciencies may be more apparent and easily addressed. There are many possible objectives for trafic signal operations, and signal timing strategies will change depending on the objective(s) that are chosen. The practitioner should understand that some objectives may be mutually exclusive. For example, objectives that focus on pedestrians will require different signal timing solutions than objectives that focus on vehicles. Below is a list of 15 specific operational objectives that can be used individually or in combination to focus signal timing efforts. Not all are useful in conveying performance in an understandable manner to the traveling public. 3.4.1.1 Vehicle-Specific Objec ves 1. Vehicle Safety: Minimize vehicle collisions, reduce vehicle conlicts, and provide suficient time for vehicles to execute movements. 2. Vehicle Mobility—Capacity Allocation: Serve vehicle movements as eficiently as possible, while also distributing capacity as fairly as possible across movements and modes. Prioritize movements according to need without excessively delaying other movements. 3. Vehicle Mobility—Corridor Progression: Move vehicles along high-priority paths (typically along high-volume movements on corridors) as eficiently as possible without excessively delaying other movements. 4. Vehicle Mobility—Secondary Progression: Move vehicles along low-priority paths (such as major street turning movements or minor street movements) as eficiently as possible without excessively delaying other movements. 5. Environmental Impact Mitigation: Minimize the amount of pollution induced by improving the eficiency of vehicle trajectories, either by reducing delay or stops or using infrastructure-to-vehicle technology to inluence vehicle travel patterns. 6. Queue Length Management: Prevent formation of excessive queues on critical lane groups, such as freeway exit ramps. 7. Vehicle and Driver Costs: Minimize stops and delay in order to reduce vehicle operating costs and driver delay costs. 3.4.1.2 Pedestrian-Specific Objec ves 1. Pedestrian Safety: Minimize pedestrian involvement in collisions, reduce pedestrian conlicts, and provide suficient time for pedestrians to execute movements. 2. Pedestrian Mobility: Serve pedestrian movements as eficiently as possible. Objec ves will help focus the goals for a signal ming project. Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-21 3. Pedestrian Accessibility: Provide the ability for pedestrians, including special-needs groups, to execute movements. 3.4.1.3 Bicycle-Specific Objec ves 1. Bicycle Safety: Minimize bicycle involvement in collisions, reduce bicycle con
licts, and provide suf
icient time for bicycles to execute movements. 2. Bicycle Mobility: Serve bicycle movements as ef
iciently as possible. 3.4.1.4 Transit-Specific Objec ves 1. Transit Safety: Minimize transit vehicle involvement in collisions, reduce transit vehicle con
licts, and provide suf
icient time for transit vehicles to execute movements. 2. Transit Mobility: Serve transit vehicles as ef
iciently as possible. 3. Transit Accessibility: Provide the ability for transit vehicles to execute maneuvers (i.e., loading/unloading activities) and for transit users, including special-needs groups, to access transit. 3.4.2 Establish Performance Measures Performance measures (also known as measures of effectiveness or MOEs) should be selected for each operational objective to evaluate the success of the timing plan(s). Most practitioners use delay and stops as measures because they are readily calculated in optimization software. However, optimization software will not always calculate performance measures that are appropriate for the operational objectives. For example, some models optimize system delay, which is not perceived by users, while others combine performance measures (e.g., performance index that combines stops and delay). Users perceive stops
irst and their delay second. Minimizing stops on arterials has the most impact on reducing user complaints, and this objective might be achieved through strategies that increase system delay slightly. For example, longer cycle lengths may be selected to reduce arterial stops, which could also result in an increase to system delay. In the
ield, arterial travel time studies have been the chief means of assessing performance. Some objectives (such as pedestrian priority) may need to be assessed qualitatively, as the performance measures may be dif
icult to measure in the
ield or through most analysis tools. However, objectives should not be overlooked simply because they are not available through software applications. Some signal timing decisions may require a qualitative assessment. Exhibit 3-17 presents a variety of traditional vehicle performance measures that can be collected to measure signal timing effectiveness. While these are the performance measures that are most often used and easily measured or estimated, they should only be used when they are consistent with the desired outcomes. Information about non-motorized performance measures is available in Section 3.4.3. In addition to summarizing timing parameters and data collection methods for each vehicle performance measure, Exhibit 3-17 also indicates (1) whether each performance measure is generally used to evaluate an individual intersection or system of intersections and (2) whether it is typically measured in the
ield or derived from a Signal Timing Manual, Second Edion

3-22 Chapter 3. Signal Timing Concepts software program or related measure. The performance measures are described in detail below. Objecve Performance Measure In te rs ec o n Sy st em M ea su re d De riv ed Influenal Signal Timing Parameter(s) Typical Method of Collecon Corridor Progression Quality of Progression – Percent Arrival on Green X X X □ Cycle length □ Offset □ Phase order □ Accurate detec on and controller logging Quality of Progression – Ra o of Arrival on Green to Arrival on Red X X □ Cycle length □ Offset □ Phase order □ Probe vehicle (GPS floa ng car, GPS fleet data) Number of Stops per Mile X X □ Cycle length □ Split □ Offset □ Phase order □ Probe vehicle (GPS floang car, GPS fleet data) Travel Time/ Average Speed X X □ Cycle length □ Split □ Offset □ Probe vehicle (GPS floang car, GPS fleet data, BluetoothTM, or other re- idenficaon) Capacity Allocaon Delay (Vehicle or Person) X X X □ Cycle length □ Split □ Offset □ Phase order □ Typically esmated from models □ Manual queue esmaon method □ Call to service controller logging Phase Failures X X X □ Cycle length □ Split □ Phase order □ Queue detecon plus signal indicaon data (manual observaon or automated technology) □ Approximate with phase terminaon logging (max outs/force-offs versus gap outs) Queuing X X X □ Cycle length □ Split □ Offset □ Phase order □ Manual observaon or automated technology (video, radar) □ Approximate with occupancy Vehicle Safety Safety-Related X X □ Detecon locaon and se€ngs □ Clearance intervals □ Rate of deceleraon from vehicles □ Red-light running (cameras, controller logging) □ Number of conflicts □ Crash records Combinaon Composite Index X X X □ Cycle length □ Split □ Offset □ Equaon that combines mulple performance measures Exhibit 3-17 Performance Measures, Timing Parameters, and Collecon Methods Signal Timing Manual, Second Edion

3-22 Chapter 3. Signal Timing Concepts software program or related measure. The performance measures are described in detail below. Objecve Performance Measure In te rs ec o n Sy st em M ea su re d De riv ed Influenal Signal Timing Parameter(s) Typical Method of Collecon Corridor Progression Quality of Progression – Percent Arrival on Green X X X □ Cycle length □ Offset □ Phase order □ Accurate detec on and controller logging Quality of Progression – Ra o of Arrival on Green to Arrival on Red X X □ Cycle length □ Offset □ Phase order □ Probe vehicle (GPS floa ng car, GPS fleet data) Number of Stops per Mile X X □ Cycle length □ Split □ Offset □ Phase order □ Probe vehicle (GPS floang car, GPS fleet data) Travel Time/ Average Speed X X □ Cycle length □ Split □ Offset □ Probe vehicle (GPS floang car, GPS fleet data, BluetoothTM, or other re- idenficaon) Capacity Allocaon Delay (Vehicle or Person) X X X □ Cycle length □ Split □ Offset □ Phase order □ Typically esmated from models □ Manual queue esmaon method □ Call to service controller logging Phase Failures X X X □ Cycle length □ Split □ Phase order □ Queue detecon plus signal indicaon data (manual observaon or automated technology) □ Approximate with phase terminaon logging (max outs/force-offs versus gap outs) Queuing X X X □ Cycle length □ Split □ Offset □ Phase order □ Manual observaon or automated technology (video, radar) □ Approximate with occupancy Vehicle Safety Safety-Related X X □ Detecon locaon and se€ngs □ Clearance intervals □ Rate of deceleraon from vehicles □ Red-light running (cameras, controller logging) □ Number of conflicts □ Crash records Combinaon Composite Index X X X □ Cycle length □ Split □ Offset □ Equaon that combines mulple performance measures Exhibit 3-17 Performance Measures, Timing Parameters, and Collecon Methods Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-23 Practitioners should reference the FHWA Trafic Analysis Toolbox (9) and Performance Measurement Fundamentals (10) for more information on data collection practices. Field data collection may provide the most accurate performance measures, but can often be cost or resource prohibitive. Many performance measures can be approximated through the use of trafic operations/safety software. The use of software can provide a lower-cost approximation of a variety of performance measures. However, some ield data are often necessary or helpful in creating models that relect reality and to ensure a level of validation and reasonableness in the results where possible. Chapter 5 contains more information about software considerations. 3.4.2.1 Quality of Progression—Percent Arrival on Green Percent arrival on green is a measure of the proportion of users (typically vehicles) that arrive during a green (or walk) indication relative to those that arrive during a red (or don’t walk) indication at a particular intersection. This is best measured in the ield, which can be done with some modern controller software. Optimization software programs are often used to approximate the effectiveness of progression between intersections, typically through the use of a time-space diagram (explained in detail in Chapter 5). However, ield observations may also be useful when attempting to determine the quality of progression. In order to maximize the accuracy of percent arrival on green accounted for in the ield, detection should be located beyond queued vehicles, yet as close to the intersection stop bar as practical. 3.4.2.2 Quality of Progression—Rao of Arrival on Green to Arrival on Red A similar performance measure to percent arrival on green is the ratio of intersections that a user (typically a vehicle) arrives at on green (or walk) versus arrives at on red (or don’t walk). This type of performance measure compares the quality of progression between different corridors. Even corridors with the best coordination will have vehicles that arrive on a red indication at some intersections, and this performance measure provides an easy, corridor-level analysis of progression quality. 3.4.2.3 Number of Stops per Mile Although the ratio of arrival on green to arrival on red provides a convenient way to compare multiple corridors, a high value for this ratio may be hard to achieve if a corridor does not have many signalized intersections. The number of stops per mile may provide a better snapshot of the condition of the corridor operations. It is a performance measure that is based on the relationship among user arrivals at an intersection, the phase status, and the queue proile. As noted previously, users are more aware of stops than they are of minor differences in delay. Additionally, if improved air quality or reduced emissions are desired objectives, then minimizing the number of stops may be an important performance measure. Research has documented the negative impacts associated with deceleration and acceleration of vehicles with respect to emissions and air quality. 3.4.2.4 Travel Time/Average Speed Travel time studies evaluate the overall quality of trafic movements along an arterial. Travel time is an easily understood and intuitive performance measure to the general public, operators, planners, and maintenance staff. Practitioners should refer to Signal Timing Manual, Second Edion

3-24 Chapter 3. Signal Timing Concepts the FHWA Travel Time Data Collection Handbook (11) and Travel Time Reliability Measures (12) for additional information on travel time data collection. While travel time provides an easily understood overall performance measure, it does not reveal the underlying traveler experience, which is sensitive to stops
irst and delay (especially small amounts that may be imperceptible) second. Use of travel time should not be done at the expense of ignoring stops and delay (two outcomes of the selection of cycle length, offsets, and splits). Modern controllers can monitor arrival on green and estimate approach delay, allowing practitioners to adjust timing parameters and measure results at the intersection level, while also looking at overall travel time as an aggregate measure of performance. 3.4.2.5 Delay (Vehicle or Person) Delay is the difference in travel time that a user experiences between free-
low (unimpeded) conditions and current conditions. It is a primary measure in optimization models because it is easily quanti
ied. It can also be used in models to estimate users’ operating costs. However, incremental changes in delay at an intersection are less noticeable to roadway users than other mobility-related performance measures, such as number of stops or overall trip travel time. It is also not readily measured in the
ield. Delay at a signalized intersection can be the result of (1) signal control and timing, (2) queues that impede travel, or (3) factors such as bus blockages, parking maneuvers, and distracted drivers. Understanding if and to what degree these forms of delay contribute to user experiences can be important and should be noted during
ield observations. Delay can ultimately be expressed in two ways: 1. Unit delay (seconds/vehicle), which is related to the user’s perception of disutility at an intersection; or 2. Total accumulated delay (vehicle-hours), which is related more to the economic performance of an intersection. One vehicle-hour of delay is accumulated when one vehicle is delayed for a full hour, or 3600 vehicles are delayed for 1 second each, etc. (13). 3.4.2.6 Phase Failures A phase failure (also called a “split failure” when the intersection is coordinated) is de
ined as the occurrence of one or more stopped vehicles that cannot proceed through a signalized intersection on a green indication. Occasionally, a movement may be served twice per cycle, in which case the term “cycle failure” may be used if all vehicles are not served within a cycle. Phase, split, or cycle failures are typically easy to observe in the
ield; the challenge for the practitioner is to determine how or if correction is necessary. A simpli
ied approach for approximating phase failures in an automated fashion is measuring and recording signal phase termination types. This involves comparing the proportion of gap outs (which imply enough time to serve demand) to max outs or force-offs (which imply not enough time to serve demand). Because the coordinated phases are typically not actuated, information may not be available for those phases. Exhibit 3-18 illustrates an example of individual and aggregated gap out versus max out/force-off conditions at a signal. Visualization graphics of gap out versus max out/force-off conditions are also available (depicted in Exhibit 3-19) to aid in identifying Signal Timing Manual, Second Edion

3-24 Chapter 3. Signal Timing Concepts the FHWA Travel Time Data Collection Handbook (11) and Travel Time Reliability Measures (12) for additional information on travel time data collection. While travel time provides an easily understood overall performance measure, it does not reveal the underlying traveler experience, which is sensitive to stops
irst and delay (especially small amounts that may be imperceptible) second. Use of travel time should not be done at the expense of ignoring stops and delay (two outcomes of the selection of cycle length, offsets, and splits). Modern controllers can monitor arrival on green and estimate approach delay, allowing practitioners to adjust timing parameters and measure results at the intersection level, while also looking at overall travel time as an aggregate measure of performance. 3.4.2.5 Delay (Vehicle or Person) Delay is the difference in travel time that a user experiences between free-
low (unimpeded) conditions and current conditions. It is a primary measure in optimization models because it is easily quanti
ied. It can also be used in models to estimate users’ operating costs. However, incremental changes in delay at an intersection are less noticeable to roadway users than other mobility-related performance measures, such as number of stops or overall trip travel time. It is also not readily measured in the
ield. Delay at a signalized intersection can be the result of (1) signal control and timing, (2) queues that impede travel, or (3) factors such as bus blockages, parking maneuvers, and distracted drivers. Understanding if and to what degree these forms of delay contribute to user experiences can be important and should be noted during
ield observations. Delay can ultimately be expressed in two ways: 1. Unit delay (seconds/vehicle), which is related to the user’s perception of disutility at an intersection; or 2. Total accumulated delay (vehicle-hours), which is related more to the economic performance of an intersection. One vehicle-hour of delay is accumulated when one vehicle is delayed for a full hour, or 3600 vehicles are delayed for 1 second each, etc. (13). 3.4.2.6 Phase Failures A phase failure (also called a “split failure” when the intersection is coordinated) is de
ined as the occurrence of one or more stopped vehicles that cannot proceed through a signalized intersection on a green indication. Occasionally, a movement may be served twice per cycle, in which case the term “cycle failure” may be used if all vehicles are not served within a cycle. Phase, split, or cycle failures are typically easy to observe in the
ield; the challenge for the practitioner is to determine how or if correction is necessary. A simpli
ied approach for approximating phase failures in an automated fashion is measuring and recording signal phase termination types. This involves comparing the proportion of gap outs (which imply enough time to serve demand) to max outs or force-offs (which imply not enough time to serve demand). Because the coordinated phases are typically not actuated, information may not be available for those phases. Exhibit 3-18 illustrates an example of individual and aggregated gap out versus max out/force-off conditions at a signal. Visualization graphics of gap out versus max out/force-off conditions are also available (depicted in Exhibit 3-19) to aid in identifying Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-25 problematic movements. Reviewing this type of information is emphasized in NCHRP Project 3-79, “Measuring and Predicting the Performance of Automobile Traf€ic on Urban Streets” (14). 3.4.2.7 Queuing Queuing Queuing is the length or number of users waiting for a green (or walk) indication. It is typically measured at the beginning of the green (or walk) indication when the standing queue length is longest. Queuing is often a direct result of the signal timing and detection parameters selected, but it is important to recognize that queues can also be in€luenced by downstream or upstream traf€ic operations. For example, in Exhibit 3-20, green time is being wasted at the nearest signal due to a queue and red-light display at the downstream signal. Exhibit 3-18 Example Split Terminaon Logging Exhibit 3-19 Example Split Terminaon Logging Visualizaon Exhibit 3-20 Arterial Efficiency Decreased by Signal Timing Signal Timing Manual, Second Edion

3-26 Chapter 3. Signal Timing Concepts Queues can be estimated using optimization software, but conducting a ield visit is the best way to evaluate conditions. A ield visit, both before and after implementation of a new timing plan, provides the opportunity to observe operational issues such as storage bay blocking or spillback (depicted in Exhibit 3-21 and Exhibit 3-22, respectively) and approaches that are not serving the full demand. One of the challenges of ield data collection is visiting the corridor in a way that allows observation of the most critical time periods and important intersections. 3.4.2.8 Safety-Related Performance Measures The following bullets summarize common safety-related performance measures that are in luenced by signal timing. Additional information on safety-related performance measures can be found in the Highway Safety Manual (15). • Number, type, and severity of crashes can be in luenced by signal timing parameters for both multiple-vehicle and single-vehicle collisions. Types of crashes at a signalized intersection can include angle, head-on, rear-end, sideswipe, and collisions with pedestrians and bicycles. Severity ranges from property-damage-only to various levels of injury to fatal. However, it should be Exhibit 3-21 Storage Bay Blocking Exhibit 3-22 Storage Bay Spillback Signal Timing Manual, Second Edion

3-26 Chapter 3. Signal Timing Concepts Queues can be estimated using optimization software, but conducting a ield visit is the best way to evaluate conditions. A ield visit, both before and after implementation of a new timing plan, provides the opportunity to observe operational issues such as storage bay blocking or spillback (depicted in Exhibit 3-21 and Exhibit 3-22, respectively) and approaches that are not serving the full demand. One of the challenges of ield data collection is visiting the corridor in a way that allows observation of the most critical time periods and important intersections. 3.4.2.8 Safety-Related Performance Measures The following bullets summarize common safety-related performance measures that are in luenced by signal timing. Additional information on safety-related performance measures can be found in the Highway Safety Manual (15). • Number, type, and severity of crashes can be in luenced by signal timing parameters for both multiple-vehicle and single-vehicle collisions. Types of crashes at a signalized intersection can include angle, head-on, rear-end, sideswipe, and collisions with pedestrians and bicycles. Severity ranges from property-damage-only to various levels of injury to fatal. However, it should be Exhibit 3-21 Storage Bay Blocking Exhibit 3-22 Storage Bay Spillback Signal Timing Manual, Second Edion Chapter 3. Signal Timing Concepts 3-27 noted that signal timing is only one factor that inluences crashes, even at a signalized intersection. Crash reports are often available through the jurisdiction that operates and maintains the intersection. • Number of potential conlict points, which are locations where two movements conlict at an intersection. Exhibit 3-23 illustrates all of the potential conlict points at a standard signalized intersection with permitted left-turn phasing. The number of conlict points can be inluenced by the type of phasing that is chosen. However, phasing can also inluence conlict points outside of the intersection (i.e., turn-lane conlicts), so should be chosen carefully. • Frequency of pedestrian conlicts with vehicles and bicycles can inluence safety for many modes at an intersection. Pedestrian visibility to turning vehicles and bicycles is an important qualitative performance measure. • Frequency of bicycle conlicts with conlicting vehicle movements can be another qualitative performance measure. Conlicts can be reduced by providing suficient time for bicycles to start moving and proceed through an intersection before conlicting movements. • Frequency of max out or force-off conditions due to vehicular demand can be a safety-related metric as the signal controller will terminate a phase once the max out or force-off is reached, regardless of vehicle proximity to the stop bar or intersection. • Frequency of users violating the red (or solid don’t walk) indication can be an important metric for understanding where and for what reason(s) red-light Exhibit 3-23 Conflicts between Movements at an Intersecon Signal Timing Manual, Second Edion

3-28 Chapter 3. Signal Timing Concepts violations may be occurring (as shown in Exhibit 3-24). This metric can be tracked through manual observation or automated technologies, including video or any detection zone that can identify a vehicle entering the intersection on red. 3.4.2.9 Composite Index One performance measure may not accurately represent the operation at a signalized intersection or along a corridor. Some agencies have considered using an “index” that combines two or more performance measures. This composite allows agencies to include multiple performance measures in their assessment. For example, the Orange County Transportation Authority Regional Trafic Signal Synchronization Master Plan (16) explains how to combine three performance measures—average speed, green signals per red signals, and stops per mile—into a Corridor Synchronization Performance Index (CSPI). This index is an easy way to compare multiple corridors using several performance measures. 3.4.3 Non-Vehicle Opera onal Objec ves and Performance Measures The practitioner should remember that not all operational objectives are easily measured. In some cases, they may be qualitative in nature. Objectives should not be chosen because they are easy to measure. They should be chosen after the desired outcome is understood. Exhibit 3-25 is a matrix that summarizes some non-vehicle performance measures as they relate to operational objectives. While this is not meant to be an exhaustive list, it is meant to demonstrate additional multimodal objectives and potential performance measures that can be chosen to evaluate a desired outcome. Just as the practitioner should carefully choose the operational objectives based on a speciŒic operating environment, he or she must also carefully choose performance measures to serve as true representations of the outcome based objectives. Exhibit 3-24 Red-Light Violaon Signal Timing Manual, Second Edi on

3-28 Chapter 3. Signal Timing Concepts violations may be occurring (as shown in Exhibit 3-24). This metric can be tracked through manual observation or automated technologies, including video or any detection zone that can identify a vehicle entering the intersection on red. 3.4.2.9 Composite Index One performance measure may not accurately represent the operation at a signalized intersection or along a corridor. Some agencies have considered using an “index” that combines two or more performance measures. This composite allows agencies to include multiple performance measures in their assessment. For example, the Orange County Transportation Authority Regional Trafic Signal Synchronization Master Plan (16) explains how to combine three performance measures—average speed, green signals per red signals, and stops per mile—into a Corridor Synchronization Performance Index (CSPI). This index is an easy way to compare multiple corridors using several performance measures. 3.4.3 Non-Vehicle Opera onal Objec ves and Performance Measures The practitioner should remember that not all operational objectives are easily measured. In some cases, they may be qualitative in nature. Objectives should not be chosen because they are easy to measure. They should be chosen after the desired outcome is understood. Exhibit 3-25 is a matrix that summarizes some non-vehicle performance measures as they relate to operational objectives. While this is not meant to be an exhaustive list, it is meant to demonstrate additional multimodal objectives and potential performance measures that can be chosen to evaluate a desired outcome. Just as the practitioner should carefully choose the operational objectives based on a speciŒic operating environment, he or she must also carefully choose performance measures to serve as true representations of the outcome based objectives. Exhibit 3-24 Red-Light Violaon Signal Timing Manual, Second Edi on Chapter 3. Signal Timing Concepts 3-29 Operaonal Objecve Performance Measure Pe rc en t A rr iv al o n Gr ee n (o r W al k) Ra o o f A rr iv al o n Gr ee n (o r W al k) to A rr iv al o n Re d (o r D on ’t W al k) N um be r o f S to ps pe r M ile Tr av el T im e/ Av er ag e Sp ee d De la y Ph as e Fa ilu re s Q ue ui ng Ac ci de nt R at e Co m po sit e Pe rf or m an ce M ea su re Pedestrian Safety X X Mobility X X X X X Bicycle Safety X X Mobility X X X X X X Transit Safety X X Mobility X X X X X X X X 3.5 REFERENCES 1. A Policy on Geometric Design of Highways and Streets, 6th Edition. American Association of State Highway and Transportation Oficials, Washington, D.C., 2011. 2. Manual on Uniform Trafic Control Devices for Streets and Highways, 2009 Edition. United States Department of Transportation, Federal Highway Administration, Washington, D.C., 2009. 3. Trafic Monitoring Guide. Ofice of Highway Policy Information, Federal Highway Administration, United States Department of Transportation, Washington, D.C., 2001. 4. Trafic Monitoring Guide Supplement. Ofice of Highway Policy Information, Federal Highway Administration, United States Department of Transportation, Washington, D.C., 2008. 5. Shelby, S. G., D. M. Bullock, and D. Gettman. Resonant Cycles in Trafic Signal Control. In Transportation Research Record: Journal of the Transportation Research Board, No. 1925, Transportation Research Board of the National Academies, Washington, D.C., 2005, pp. 215–226. 6. Robertson, H. D., J. E. Hummer, and D. C. Nelson. Manual of Transportation Engineering Studies. Institute of Transportation Engineers, Washington, D.C., 2010. 7. Henry, R. D. Signal Timing on a Shoestring. Report FHWA-HOP-07-006, Federal Highway Administration, United States Department of Transportation, 2005. 8. Chandler, B. E., M. C. Myers, J. E. Atkinson, T. E. Bryer, R. Retting, J. Smithline, J. T. P. Wojtkiewicz, G. B. Thomas, S. P. Venglar, S. Sunkari, B. J. Malone, and P. Izadpanah. Signalized Intersections Informational Guide, Second Edition. Report FHWA-SA-13- 027, Federal Highway Administration, United States Department of Transportation, 2013. 9. Federal Highway Administration. Trafic Analysis Toolbox. http://ops.hwa.dot.gov/traficanalysistools/. Accessed Dec. 22, 2013. Exhibit 3-25 Non- Vehicle Operaonal Objecves and Performance Measures Signal Timing Manual, Second Edi
on

3-30 Chapter 3. Signal Timing Concepts 10. Federal Highway Administration. Performance Measurement Fundamentals. http://ops.hwa.dot.gov/perf_measurement/fundamentals/. Accessed Dec. 22, 2013. 11. Turner, S. M., W. L. Eisele, R. J. Benz, and D. J. Holdener. Travel Time Data Collection Handbook. Report FHWA-PL-98-035, Federal Highway Administration, United States Department of Transportation, 1998. 12. Federal Highway Administration. Travel Time Reliability Measures. http://ops.hwa.dot.gov/perf_measurement/reliability_measures/index.htm. Accessed Dec. 22, 2013. 13. Florida Intersection Design Guide. Florida Department of Transportation, 2013. 14. Bonneson, J. A., M. P. Pratt, and M. A. Vandehey. NCHRP Project 3-79, “Measuring and Predicting the Performance of Automobile Trafic on Urban Streets.” Transportation Research Board of the National Academies, Washington, D.C., 2008. 15. Highway Safety Manual. American Association of State Highway and Transportation Oficials, Washington, D.C., 2010. 16. Regional Trafic Signal Synchronization Master Plan. Orange County Transportation Authority, 2009. Signal Timing Manual, Second Edion