Systemic Pedestrian Safety Analysis (2018)

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.

80 Twelve pedestrian countermeasures are presented that could be considered during a systemic pedestrian safety analysis process. Each countermeasure summary describes the purpose, cur- rent use of the treatment, available safety evidence/benefits, and considerations for application within a systemic framework. Each countermeasure summary also includes a photo example (see Figures A1–A12). State or local agencies may choose to expand this list of treatments or limit the list based on local conditions and concerns (such as prior agency studies of effectiveness or political considerations) as well as the degree to which measures are implemented systematically or as a matter of policy. Countermeasures for Signalized or Unsignalized Crossing Locations High Visibility Crosswalks A P P E N D I X Potential Countermeasures Figure A1. Midblock crosswalk with high visibility pavement markings in Bellevue, Washington (www.pedbikeimages.org/Dan Burden). Purpose and Use High visibility crosswalks use continental (ladder) or bar–pair pavement markings to increase the conspicuity of pedestrian crossings (relative to traditional transverse pavement markings) for both pedestrians and motorists. While the use of high visibility crosswalks has become perva- sive throughout the United States, the implementation strategies across states and local agencies may differ. Some agencies have installed them systematically at every marked crossing location, while other agencies have installed them only at priority risk locations (e.g., unsignalized cross- walks, roadways with high pedestrian volumes, high speeds, or school zones).

Potential Countermeasures 81 Effectiveness High visibility crosswalks can provide safety benefits, including crash reductions, as motorists detect them sooner than standard parallel line crosswalk markings during daytime and night- time (Fitzpatrick et al. 2011). They also may help increase motorist yielding, especially when paired with other countermeasures such as highly visible “pedestrians present when active” types of devices such as PHBs or rectangular rapid flashing beacons. Other potential benefits have been noted, including increased pedestrian scanning behavior and fewer pedestrians stranded in the middle of the street while crossing (Pulugurtha et al. 2012). Potential effects on motorist speeds are unclear. Systemic Application Systemic application focuses on the use of high visibility markings of any type to replace standard crosswalk markings or to implement a crosswalk once a thorough assessment has been made of the suitability of a location for a marked crosswalk. It is recommended that a full engineering and safety study be conducted to assess pedestrian need, sight distance, safety, and operational suitability of an uncontrolled location for a new crosswalk and/or a crosswalk combined with other treatments. For a systemic approach, agencies may wish to consider application of high visibility markings to help address identified risks, including poor conspicuity of pedestrian crossings or unexpected locations (such as midblock), failure of drivers to slow or yield, lack of pedestrian compliance or use of crosswalks, and conflicts with turning vehicles at signalized intersections. High vis- ibility crosswalks alone are not sufficient to treat high volume, high speed, multi-lane roads or inadequate lighting. Traffic Calming: Raised Devices Figure A2. Raised midblock crosswalk (www.pedbikeimages.org/Dan Burden). Purpose and Use Raised roadway devices (i.e., speed tables, raised crosswalks, and speed humps) aim to calm traffic and improve the pedestrian environment by reducing vehicle speeds and the need for pedestrian ramps. These treatments may be applied at either intersection or midblock crossings to slow vehicle speeds on local and collector roadways in urban/suburban environments; they also may be applied in other specialized locations experiencing a high volume of pedestrians (e.g., airport pick-up/drop-off zones or college campuses) (Mead et al. 2014, Thomas et al. 2016). State DOTs have been slow to implement raised devices on their networks (only 17% reported they

82 Systemic Pedestrian Safety Analysis used raised crosswalks or speed tables often), but local agencies use them with greater frequency (Thomas et al. 2016). Effectiveness The extant literature has found traffic-calming treatments, specifically raised devices, provide an effective means of reducing vehicular speeds (Engineering Speed Management Counter- measures 2014, available at https://safety.fhwa.dot.gov/speedmgt/ref_mats/eng_count/2014/ reducing_crashes.cfm; Strong and Kumar 2006). In turn, several studies have shown these reduced vehicle speeds can reduce the frequency and severity of pedestrian crashes (Elvik and Vaa 2004; Toolbox of Countermeasures and Their Potential Effectiveness for Pedestrian Crashes 2013, citing Bahar et al. 2007). While the low frequency of pedestrian crashes where these devices are typically employed (low-speed environments) have made these findings somewhat difficult to replicate, a reduction in vehicle speeds may still serve as a surrogate measure of pedestrian safety and comfort. Systemic Application Traffic calming may be implemented through a variety of means, including special programs involving neighborhood requests to reduce speeding on local streets. These types of programs typically establish specific criteria and procedures for implementation and often require neighbor- hood support of the measure. If accepted for implementation on a jurisdictionwide basis, raised traffic calming treatments can be implemented through a risk-based systemic process at appropri- ate locations. However, they are most appropriate on relatively low-speed roads and have special considerations for application to ensure that problems do not migrate elsewhere. Median Crossing Islands Figure A3. Median crossing island in Santa Cruz, California (www.pedbikeimages.org/Dan Burden/ 2006). Purpose and Use Median crossing islands are raised areas intended to protect pedestrians crossing at inter- sections or midblock locations by providing a refuge area to wait for a gap in traffic before begin- ning the second leg of the crossing (Thomas et al. 2016). Also referred to as pedestrian median islands, center islands, and refuge islands, these treatments can be used at both uncontrolled and signalized crossings and may be used with curb extensions to further reduce crossing distance (Mead et al. 2014, Thomas et al. 2016). Median crossing islands have been used extensively by both state and local agencies on roadways with high volumes, high speeds, and multiple lanes (Thomas et al. 2016).

Potential Countermeasures 83 Effectiveness Median crossing islands have been shown to reduce pedestrian–motor vehicle crashes (Zegeer et al. 2017a, b). Median crossing islands appear to help increase the frequency of motorist yielding and reduce motorist speeds (Pulugurtha et al. 2012, Kamyab et al. 2003). This treatment may also reduce multiple-threat crash risk, dash and dart-out crashes, crashes caused by motorists or pedes- trians failing to yield the right-of-way, and crashes at unique midblock crossings. Several studies suggest that median crossing islands encourage safer crossing behaviors by pedestrians, including scanning and compliance in crossing at the marked crosswalk (Pulugurtha et al. 2012, Strong and Kumar 2006), but effects on motorist yielding and speed are less certain (Thomas et al. 2016). Systemic Application Crossing islands may be applied at signalized and unsignalized intersections as well as at mid- block locations on both high- and low-speed roadways. However, FHWA recommends their use for areas with less than 12,000 AADT for vehicles and high pedestrian volumes, as well as roadways with moderate to high travel speeds. Additional treatments may be needed to increase motorist yielding or address speed concerns. Lane Reduction: Road Diet Road Before Road After Figure A4. Road diet in Orlando, Florida (https://safety.fhwa.dot.gov/road_diets). Purpose and Use A road diet reduces the number of motorized traffic lanes along a street in conjunction with a reallocation of that space to other uses such as sidewalks, bicycle lanes, parking, or transit. Also known as “road conversions,” these projects commonly involve converting a four-lane undi- vided road into a three-lane road with two through lanes and a center turn lane (i.e., TWLTL) and bike lanes, but medians may also be used. Road diets benefit pedestrians by reducing the number of lanes to cross, allowing for medians or median islands and other treatments to be added to the road to assist in crossing, and by creating space for walkways, bike lanes, or addi- tional buffers from traffic (Knapp et al. 2014). Approximately three-quarters of surveyed states and local agencies reported using road diets in locations that have excess capacity or to provide additional space for bicyclists and pedestrians (Thomas et al. 2016). Effectiveness Road diets have been found to reduce total crashes 19% to 47% and reduce motor vehicle speeds, which is expected to have pedestrian safety benefits. The larger range of effect was found for main roads through small towns and suburban areas, with the lower range of effect

84 Systemic Pedestrian Safety Analysis in large urban areas (Harkey et al. 2008, Toolbox of Countermeasures and Their Potential Effectiveness for Pedestrian Crashes 2013). Road diets can also reduce travel speeds. New York City observed a downward trend in pedestrian crashes after evaluating effects of 460 road diets (Thomas et al. 2016). Systemic Application As with all types of applications, diagnosis and additional screening are needed to ensure that locations are appropriate for the measure. FHWA’s Road Diet Informational Guide (available at https://safety.fhwa.dot.gov/road_diets/guidance/info_guide/) provides information on the safety benefits and potential trade-offs of road diets, performing a feasibility assessment (including traffic volume considerations, number of driveways and junctions, and intersection function), design- ing a road diet conversion, and evaluating a road diet (Knapp et al. 2014). Road diets are con- sidered feasible for roads with lower-to-moderate traffic volumes (less than 25,000 ADT). The Iowa Department of Transportation developed a statewide screening system for candidate sites for four-to-three lane road conversion that is useful for understanding how existing roadway databases and GIS data can be used in a streamlined screening approach (Statewide Screening for Potential Lane Reconfiguration 2017, available at https://iowadot.gov/systems_planning/ pr_guide/Safety/StatewideScreeningforPotentialLaneReconfiguration.pdf). Curb Extensions and Parking Restrictions Purpose and Use Curb extensions (also known as bulb-outs or neckdowns) and parking restrictions near an intersection or other crossing are two countermeasures that frequently appear together to reduce crossing distance and improve sight lines between motorists and pedestrians. Curb extensions extend the sidewalk and the curb line into the roadway on streets that have on-street parking. The removal of parking adjacent to crossing locations is sometimes referred to as “daylighting.” Guidance suggests that, at a minimum, vehicles should not be parked within 20 feet of an inter- section or other crosswalk, with longer restrictions if speeds are greater than 25 mph (Blackburn et al. 2017). Effectiveness A curb extension and parking restriction combination should reduce pedestrian exposure by shortening crossing distance or width of the street, enhance conspicuity of pedestrians to drivers, and allow pedestrians to better view oncoming traffic. On their own, curb extensions may not improve motorist yielding but may help to improve yielding in combination with other treat- ments such as median islands and advance stop/yield markings and signs (Thomas et al. 2016). Figure A5. Curb extension and parking restriction at midblock crosswalk approach (www.pedbikeimages. org/Andy Hamilton).

Potential Countermeasures 85 Curb extensions may also slow turning traffic at intersections since vehicles must navigate a narrower turning radius. Traffic may be slowed on segments if the roadway is sufficiently nar- rowed. One study (King 1999) found that curb extensions lowered pedestrian crash severity at four out of six locations, possibly indicating a speed-lowering effect. Another benefit is that installing a curb extension is an opportunity to install an ADA-accessible curb ramp that may not have previously existed. The effects on crashes of curb extensions are not known, but at least one study has found pedestrian crash reductions about 30% for removal of on-street parking to off-street parking (Toolbox of Countermeasures and Their Potential Effectiveness for Pedestrian Crashes 2013, citing Gan et al. 2005). Systemic Application Curb extensions and parking restrictions—along with high-visibility crosswalks, improved nighttime lighting, advance yield signs and lines, and in-street pedestrian crossing signs—are all part of FHWA’s STEP countermeasure “crosswalk visibility enhancements” at uncontrolled locations. Curb extensions are recommended for use on two-to-four lane roads with and with- out medians and for any speed (Blackburn et al. 2017). Curb extensions and parking removal may help to address almost all crash types that occur at crossing locations, including midblock or unsignalized intersection crossing crashes involving through vehicles, turning vehicle crashes (at intersections), pedestrian dashes and dart-outs, and multiple threat types, especially if sight lines are obscured by parking. However, curb extensions on their own may not induce drivers to yield to pedestrians more often, and speed effects are not guaranteed, so multiple lane crossings on higher volume, higher speed roads may need addi- tional treatments to help create safe gaps for pedestrians at unsignalized locations. Curb exten- sions should only be used where there is an existing parking lane and where bicycles and transit vehicles can be accommodated in the through lanes. Curb extensions also provide an opportunity for additional pedestrian landing/waiting space, curb ramps, and the potential for landscaping and street furniture. Parking restrictions can also be considered at any crossing location. Oregon DOT, for example, uses parking restrictions in conjunction with curb extensions, lighting, medi- ans, and leading pedestrian intervals (at signalized intersections) (Thomas et al. 2016). Site-Specific Lighting Improvements Figure A6. Recommended lighting configurations for midblock pedestrian crossings (Gibbons et al. 2008, p. 13, Figure 12).

86 Systemic Pedestrian Safety Analysis Purpose and Use Crosswalk and roadway illumination is intended to enhance the visibility of pedestrians crossing or walking along roads at night. Illumination may take the form of overhead lighting, automated detection systems with enhanced lighting, and bollard luminaries. Pedestrian cross- ing areas should have additional, brighter lighting on approaches to and directly at crosswalks (Zegeer et al. 2013, Thomas et al. 2016). Approximately three-quarters of surveyed states and local agencies sometimes provide enhanced crossing illumination (Thomas et al. 2016). Nation- ally, 75% of pedestrians killed in 2016 were struck under various conditions of darkness, includ- ing lighted and unlighted roadways (National Center for Statistics and Analysis 2018). Effectiveness Illuminating crossing locations has been found to increase both motorist yielding and the distance at which motorists yield. Illumination improves pedestrian visibility and decreases the time needed for motorists to detect and react to a pedestrian. Enhanced lighting at crosswalks may also increase the percentage of pedestrians using crosswalks instead of crossing at other unmarked locations (Nambisan et al. 2009). Providing enhanced illumination is expected to reduce the severity and number of all types of crashes at night and has been shown to have a reduction on nighttime pedestrian crashes (Chu 2006, Elvik and Vaa 2004, Kump et al. 2016, Mead et al. 2014, Wei et al. 2017). Crash reduction for nighttime pedestrian crashes has been estimated at 42% for the provision of intersection illu- mination (Elvik and Vaa 2004). Bollard luminaries are a promising lighting treatment that have been found to improve visibility of pedestrians at midblock crosswalks and intersections and reduce the identification time needed to detect pedestrians when compared with other lighting treatments (Bullough et al. 2012, Gibbons et al. 2008). Systemic Application Lighting is appropriate at both intersection and midblock locations and roadways of low-to- high speeds. It is also appropriate for all land use types, although light spillover and energy costs may be issues to be considered. Crash-based effectiveness evidence for pedestrians comes mostly from studies of intersections. Implementation of lighting to improve pedestrian safety should ensure that pedestrian walk- ways and crosswalks are well lit. Lighting should be installed on both sides of wide streets and streets in commercial districts. Uniform lighting levels should be used along corridors with enhancements at risk locations. Lighting to enhance pedestrian environments generally could also be considered as a system- atic countermeasure and may be addressed through land use and safety planning and policy development. Some states may leave lighting initiatives up to municipalities due to maintenance, energy costs, and dark skies initiatives at the local level (Thomas et al. 2016). Countermeasures for Unsignalized Locations: Midblock or Intersection In-Roadway “Yield-to-Pedestrian” Signs (R1-6) Purpose and Use In-roadway “Yield to Pedestrian” (R1-6) signs are installed at uncontrolled crosswalks to encourage motorist yielding. They are often installed at unsignalized crosswalks on lower-speed, two-lane roads. Sometimes these signs are combined with crossing islands and other treatments

Potential Countermeasures 87 on larger or higher speed roads. Almost 70% of state and local agencies reported using this treat- ment on occasion (Thomas et al. 2016). Effectiveness In multiple behavioral studies, in-roadway “Yield to Pedestrians” signs have been shown to increase motorist yielding substantially, particularly on lower-speed, lower-volume roads. The rates of motorist yielding have increased for both rural and urban locations, and regard- less of whether the sign or multiple signs were placed in advance of or directly at the crosswalk (Ellis et al. 2007). The highest yielding rates of the standard treatment (one sign in the middle of the road) seem to peak around 75% to nearly 80% (Thomas et al. 2016). Studies of con- figurations, referred to in reports as a “gateway” configuration, show that applying the signs in between lanes on both the outside (curb edge) and inside (yellow) lane line can improve yielding and, in some cases, contribute to driver speed reductions (Van Houten 2017). Motorist yielding has been higher on roads with posted speeds of 30 mph or lower, including on some multi-lane installations. Sign placement may also be used to encourage yielding further in advance of the crosswalk (Van Houten and Hochmuth 2017). Systemic Application Before selecting this treatment for specific locations, agencies should consider the typical motorist-yielding rates within the jurisdiction and that speed limits, number of lances, and the extent of narrowing achieved by the signs’ placements may affect speeds and motorist-yielding rates. Other researchers have also documented that higher travel speeds and platooning effects may reduce motorist yielding (Bertulis and Dulaski 2014, Gårder 2004, Strong and Kumar 2006). Maintenance of the signs may also be a challenge; placement of signs on a curb mount and use of flexible attachments have helped reduce damage and keep signs in service longer (Van Houten 2017, Van Houten and Hochmuth 2017). Advance Stop/Yield Bars and R1-5/5a Signs Purpose and Use Advanced stop/yield pavement markings with “Stop/Yield Here” (R1-5/5a) signs are a low- cost treatment that has been widely implemented at multi-lane, uncontrolled crosswalks. State and local agencies often use the treatment in combination with other countermeasures to help Figure A7. In-roadway pedestrian signs and center line bollard installed at a midblock crosswalk in the gateway configuration (Ron Van Houten).

88 Systemic Pedestrian Safety Analysis reduce the crash risk associated with motorists failing to yield the right-of-way to pedestrians. In particular, advanced stop/yield bars and R1-5/5a signs are implemented in locations where the motorist’s view of a pedestrian in the crosswalk is obscured by a stopped or slowing motor vehicle in an adjacent lane (Thomas et al. 2016, Zegeer et al. 2017a). Effectiveness Crash reductions in the range of 14% to 36% and behavioral improvements have been asso- ciated with this treatment (Zegeer et al. 2017a, b). Multiple behavioral studies have found this treatment to be associated with increases in the distance motorists stop from an uncontrolled crosswalk. This increased stopping distance provides improved sight lines between vehicles approaching in an adjacent lane to a stopped or slowing vehicle and crossing pedestrians and substantially decreases conflicts (Thomas et al. 2016, Zegeer et al. 2017a). Systemic Application Advance stop/yield markings and signs can be considered a baseline treatment for unsignal- ized intersections or uncontrolled, midblock locations on multilane roads to help address risks associated with multiple lanes such as “Motorist-Failed-to-Yield” and “Multiple Threat” crash types. Bus stops placed in advance of uncontrolled crosswalks are another location that can be screened for potential application if the bus stops cannot be relocated. Other baseline treatments should be considered on an as-needed basis to address potential risks associated with speed, darkness, and motorists’ failures to yield. Pedestrian Hybrid Beacons Purpose and Use PHBs are devices that combine features of warning beacons with stop control for uncon- trolled crosswalks. Also known as high-intensity activated crosswalk or HAWK beacons, these treatments are located on mast arms over the major approaches, with two red lenses above a single yellow lens that are dark when the beacon is not activated. Overhead signs labeled “Cross- walk: Stop on Red” accompany the beacon. Once activated, the beacon displays flashing yellow, followed by steady yellow, and then steady red with a “Walk” signal for pedestrians. Because the steady red changes back to an alternating flashing red pattern during the pedestrian clearance interval, drivers can proceed if pedestrians have cleared their lanes (and no others have begun crossing). Consequently, there may be less vehicle delay than with a standard stop-and-go signal. Figure A8. Advance Stop/Yield bars and R1-5/5a signs help improve sight lines between motor vehicles and pedestrians at multi-lane crossings (www.pedbikeimages.com/Toole Design Group).

Potential Countermeasures 89 However, since driver judgment is allowed during the flashing red phase, there is still a potential for conflicts and multiple-threat type events. Pedestrian countdown signals installed as part of the PHB package seem to enhance intuitiveness of the signals and provide added benefits to pedestrians (Road User Behaviors at Pedestrian Hybrid Beacons 2016). Marked crosswalks should also be present. The original PHB design, which places the beacon over the crosswalk, is preferred to modified designs with an offset beacon. Effectiveness PHBs have been associated with safety improvements at uncontrolled pedestrian crossings, with increased motorist yielding and pedestrian crash reductions as the main safety benefits. Pedestrian crash reductions have been estimated in the range of 47% to 69% (Zegeer et al. 2017a, b; Fitzpatrick and Park 2010b). Compared with sites with no treatment and/or other warning type devices, motorist-yielding studies report higher rates of yielding behavior, ranging from 61% to 98% at sites with PHBs. The effect of PHBs on increasing motorist yielding seems to hold true for wide crossings, roads with higher speed limits (40 to 45 mph), and high volume roads (Road User Behaviors at Pedes- trian Hybrid Beacons 2016, Fitzpatrick et al. 2006). In cities where PHBs are used often, driver- yielding rates also appear to be higher, possibly due to familiarity with the treatment (Fitzpatrick et al. 2014), although other causes such as differences in traffic enforcement and driver popula- tions cannot be ruled out. Systemic Application PEDSAFE recommends PHBs for unsignalized intersections and midblock crosswalks on both low speed and high speed (defined as greater than 45 mph) roadways to address pedestrian needs for a gap to cross and/or lack of motorist yielding. Part of the screening process for systemic application might include whether the pedestrian volume or other warrant conditions are met (such as older or younger populations in the area served or expected demand); however, engineers may want to consider some of these factors during field investigations if not all relevant data types are available during initial screening. PHBs may not be for all areas; urban cores and other locations where pedestrian volumes are always high, or crossing locations where pedestrians appear routinely enough to meet MUTCD traffic signal warrants, may be better served by regular traffic signals. It may also be important to consider driving culture, enforcement, and motorist-yielding compliance rates when consider- ing installing them, especially on higher speed roadways. Figure A9. Installation of PHB signal in Tucson, Arizona (www.pedbikeimages.org/Sree Gajula/ 2009).

90 Systemic Pedestrian Safety Analysis Countermeasures for Signalized Intersections Leading Pedestrian Interval Figure A10. LPIs to reduce conflicts between turning traffic and pedestrians at a signalized intersection in La Mesa, California (www.pedbikeimages.org/ Dan Burden). Purpose and Use Leading pedestrian intervals or LPIs, also known as pedestrian head start or a delayed vehi- cle green, gives pedestrians at signalized intersections a few seconds head start (typically 3 to 7 seconds per PEDSAFE) to cross before parallel-path traffic receives the green signal. This head start allows pedestrians to establish presence in the crosswalk, make pedestrians more visible to turning vehicles, and thus increase motorist yielding (Mead et al. 2014, Thomas et al. 2016). LPIs are used by about two-thirds of states and local agencies, primarily on a case-by-case basis but typically for intersections with heavy turning volumes and high pedestrian volumes, or other identified pedestrian safety improvement locations (Thomas et al. 2016). However, at least one state uses LPIs at locations where motorists are turning “aggressively” and pedestrian volumes are not high enough to command the right-of-way. An audible pedestrian signal alerts visually impaired pedestrians as to when they have the walk indication. Effectiveness LPIs have been found to reduce pedestrian conflicts with turning vehicles. Several crash-based studies have also found pedestrian crash reductions to be associated with the treatment. LPIs are considered to be an effective countermeasure for reducing left- and right-turn-related conflicts and pedestrians ceding right-of-way to turning vehicles, particularly in urban areas (Hua et al. 2009, Van Houten et al. 2000). However, at intersections where right-turn volumes are very high or drivers are turning aggressively, especially outside urban cores or areas where pedestrians may be less expected, restrictions on right turn on red might be needed to supple- ment LPIs in order to enable pedestrians to leave the curb and capture the crosswalk during the leading interval (Hubbard et al. 2008). At urban, high pedestrian volume sites where LPIs have been studied, LPIs have been deemed to reduce pedestrian crash rates and crash severity (King 1999, Brunson et al. 2017, Fayish and Gross 2010, Institute of Transportation Engineers 2004). Systemic Application For systemic applications, if crash/crash type data are lacking, observations of pedestrian crossings and percentages of crossings compromised by motorists failing to yield right-of-way

Potential Countermeasures 91 to pedestrians could serve as a potential surrogate measure of risk. As previously mentioned, right turn on red restrictions may be needed as a companion measure in some traffic conditions. Lighting enhancements and other measures such as high visibility crosswalk markings could also be considered. Extending/Longer Pedestrian Phase Figure A11. Pedestrians crossing on the Walk signal in Walnut Creek, California (www.pedbikeimages.com/ Dan Burden/2006). Purpose and Use Extending the pedestrian walk phase is a method of adjusting signal timing so that pedes- trians have more time to cross at a signalized crossing. However, walk time must be balanced with pedestrian and motorist (including transit and trucks) wait times with respect to overall cycle length (Van Houten et al. 2007). In general, pedestrian level of service is improved through shorter overall cycle lengths (wait time) but longer pedestrian crossing times. Some agencies also use Puffin pedestrian detection technologies to extend pedestrian walk phases when pedestrians have not cleared the intersection. Effectiveness Implementing a longer phase or extending the walk phase by using pedestrian detection has not been studied extensively, but the research suggests that reducing the amount of time that pedestrians must wait and giving them additional time to cross may reduce crashes and improve pedestrian compliance. In one study, a longer pedestrian phase was estimated to reduce the pedestrian crash rate in a study of 244 New York City intersections where pedestrian crossing time was increased by increasing the overall cycle length. Both large main roads and side streets were given more green time, which included more time for pedestrians to cross both streets (Chen et al. 2014). The average pedestrian crash rate fell by 50% at the 244 intersections where pedestrian crossing time was increased and decreased by only 29% at the control intersections. Systemic Application Signals should allow adequate crossing time for pedestrians to complete their crossing. Cur- rent guidance recommends a crossing and clearance interval based upon a maximum walking speed of around 3.5 feet per second, or a slower speed (typically 3.0 feet per second) in areas with high elderly or younger pedestrians (Zegeer et al. 2013, Gates et al. 2016). If pedestrians are

92 Systemic Pedestrian Safety Analysis frequently stranded in the crosswalk, there may be insufficient crossing time. Observations may be needed to determine if clearance intervals are insufficient or if pedestrian crossings are being compromised for reasons such as high volumes of pedestrians, slower walking speeds, or turning vehicles not yielding right-of-way to crossing pedestrians, thereby resulting in pedestrian delays. When implementing a longer pedestrian phase, take care to balance the wait times for pedes- trians and motorists so as not to inadvertently reduce compliance by either mode. For example, pedestrian signal compliance could be high in the types of New York City locations where the treatment was implemented, even with long overall wait times, because the large main roads with high traffic volumes may reduce pedestrians’ temptations to violate the signal. These behav- ioral factors (pedestrian compliance), and others should be considered in screening, and further diagnosis of locations where risks related to pedestrian volumes, crossing distance, and crossing time needs may suggest consideration of this treatment. Additional diagnosis of issues and traffic characteristics is essential. Restricted Left Turns: Fully Protected Pedestrian Walk Phase Figure A12. Pedestrians crossing at an intersection with restricted left-turn phasing in New York City, New York (Chen et al. 2014). Purpose and Use Fully protected/restricted left-turn phases only permit left-turning vehicles to turn while a green arrow is displayed. Typically, an exclusive left-turn lane is provided with this phasing. The FHWA’s Traffic Signal Timing Manual recognizes this protected left-turn phasing as the safest left-turn operation, although notes that it may increase overall intersection delay (Koonce 2017). Studies of permissive phasing and flashing yellow arrows recommend protected left phasing when pedestrians are present (Hurwitz and Monsere 2013, Steyn et al. 2013). Effectiveness The implementation of a protected left-turn phase decreases motor-vehicle angle crashes at signalized intersections and reduces pedestrian crashes. Decreases in crashes come primarily from a reduction in crashes with left-turning vehicles (Toolbox of Countermeasures and Their Potential Effectiveness for Pedestrian Crashes 2013, Harkey et al. 2008, De Pauw et al. 2015). Restricted left turn and fully protected pedestrian phasing also appear to reduce pedestrian crashes at intersections (Chen et al. 2014, Strauss et al. 2014).

Potential Countermeasures 93 Systemic Application Protected left-turn phasing addresses the risk of conflict between pedestrians and motor vehicles, specifically parallel-path left-turning vehicles at signalized locations. Implementation of a protected left-turn phase would be appropriate at signalized intersections with exclusive left-turn lanes and high-turn volumes and conflicts with pedestrians. If these conditions are not present, the implementation would include installing a left-turn arrow signal and exclusive left- turn lanes (which also helps to decrease rear-end motor vehicle crash potential). This treatment may require longer cycle lengths and impact signal system coordination, which can be addressed through performing an intersection capacity analysis. Less restrictive measures such as time-of-day left-turn restrictions or leading pedestrian intervals may be an option for locations lacking separate turn lanes or lower but still impactful turning volumes.