This chapter discusses the deployment of the countermeasures. It first presents an overview of the countermeasures and the installation challenges. Also presented is a comparison of the countermeasures in terms of cost and device availability. A description of each individual countermeasure, its purpose, the setting for each San Francisco installation, device availability and approval status, cost information, and key issues are presented in Chapter 3.

2.1. Overview of Installation Challenges

Phase II involved the installation and evaluation of a broad range of pedestrian safety measures, from nearly routine signal timing changes to customized infrared and video detection equipment. However, there were several common challenges that the San Francisco team faced:

• Selection and Confirmation of Countermeasures: The initial process of selecting countermeasures and assigning them to specific locations was described in the Phase I Final Report and the Phase II Implementation Plan and Preliminary Engineering Report. This involved development of a comprehensive list of candidate countermeasures, which were then rated on several criteria, such as cost, presumed effectiveness, ease of implementation, and ability to attract additional funding. These were matched to particular study intersections by careful analysis of the pedestrian safety problem at each intersection and the physical characteristics of the intersection. ( For example, if a high number of pedestrian injuries were attributed to vehicles failing to yield to pedestrians while turning left, the TURNING VEHICLES MUST YIELD TO PEDESTRIANS sign and median refuge islands were considered. The median refuge islands required sufficient median space, so were not useful at most locations.)

  This effort had to balance the desire (for both safety and research reasons) to spread deployment to a large number of sites with the limited project budget. The initial countermeasure plan was also modified during Phase II conceptual engineering. This step included intensive research on other installations nationally and preliminary design. Local conditions could affect the feasibility of countermeasures significantly. (For example, installing video detection cameras at a location with trolley wires could be complicated by potential camera blockages.)

  While SFMTA staff proposed and finally decided on the selection of countermeasures, FHWA and SAIC staff (and other stakeholders) were actively involved. The preliminary countermeasure plan was reviewed by an External Stakeholder Advisory Group, comprised primarily of the existing Pedestrian Safety Advisory Council, and by the Pedestrian Safety Interdepartmental Working Group. The plan was also approved by the San Francisco Municipal Transportation Agency Board (effectively the “transportation commission” for the City and County of San Francisco). This was followed by a two-day site visit in November 2004 with FHWA and SAIC, including extensive field work and follow-up communications.

• Scheduling of Countermeasures: A complex, phased installation schedule was established for each pedestrian safety measure at specific intersections. This was intended to accommodate the evaluation plan, as designed by the University of California at Berkeley Traffic Safety Center (TSC). This phased installation was implemented to facilitate data collection with limited field crew staffing over a wide geographic area of the city, and to conduct video recording for analysis of pedestrian/driver behavior. It was intended that at some locations two or more countermeasures would be installed, but at separate times to allow for separate evaluation. It was also intended that multiple baseline observations would be made to allow for statistical controls. This phased installation schedule had to be coordinated with San Francisco DPT sign, paint and signal shops, which process dozens of work order requests on a monthly basis. Typically the shops prioritize work orders, but do not schedule work tasks for specific days weeks ahead.

• Internal and Inter-Agency Coordination: The signal timing changes for All Red phases, increased Walk phase time, and Pedestrian Head Starts required coordination with the DPT signal shop’s high priority signal upgrade projects and other signal timing changes. Installation of fixed radar speed signs and flashing beacons required prior determination of whether electrical power was available in the field, requiring notification and permission of utility agencies. Installation of median refuge islands involved planning and design coordination with the Department of Public Works for construction of the islands. The Department of Telecommunications and Information Services installed flashing beacons, but needed to work closely with the DPT Signal Shop. The video detection installation required close coordination with Econolite (the manufacturer of the video camera and detection system), the developer of the D4 customized signal controller software, the DPT Signal Shop, and SFgo (citywide integrated traffic management system) engineers. In general, any pedestrian safety measure that touched the street pavement required checking with street repaving and construction project schedules so that installation would not be scheduled before a major repaving or construction project on the street.

• Public Hearing Process: Some pedestrian safety measures such as red visibility curbs, and construction of median refuge islands required a public hearing. Parking changes related to the red visibility curbs also required City and County Board of Supervisor approval.

• Weather: Heavy, continual rains during the winter months caused delays in installation of several of the pedestrian safety measures that involved painting of surface pavement such as red visibility curb zones, advanced limit lines, “LOOK” stencils and painted islands for installation of “Yield to Pedestrians” signs.

2.2. Comparison of Countermeasures: Ease of Implementation

Four countermeasures proved especially challenging to deploy:

• Portable Radar Portable Changeable Message Speed Limit Signs;

• Changeable Message Speed Limit Signs;

• Video Detection Of Pedestrians to Extend Crossing Time; and

• Flashing Beacons With Infrared Bollards.

All involved procurement or deployment of sophisticated electronic equipment that engineering staff and electricians were not highly familiar with. All required customized design layouts.

The portable radar Portable Changeable Message Speed Limit Sign required extensive staff support for several reasons. First, comprehensive testing and adjustment were necessary after delivery to ensure that the device worked properly. Second, it required a formal Memorandum of Understanding (MOU) between the SFMTA and the San Francisco Police Department (SFPD), as the latter stored, maintained, transported, and set up the device. Third, it required continuing coordination between SFMTA and the SFPD, including scheduling deployment to coincide with data collection. Finally, to avoid vandalism and to heighten the novelty impact of the device, it was moved to different sites on a daily basis, requiring extensive labor.

The primary challenge with the Changeable Message Speed Limit signs was to locate them on poles where they could both utilize existing electrical power and also be effective and visible. Additionally, the devices required extensive fine-tuning by SFMTA electricians, primarily to adjust the detection zone.

The video detection of pedestrians to extend crossing time was a completely innovative use of the technology, which Econolite believes is the only U.S. application of this type. A customized detection zone scheme and logic for adjusting the signal timing had to be developed, tested, and refined, and the Econolite Autoscope detection software needed to be coordinated with the D4 traffic signal controller software.

The flashing beacons with infrared bollards required the most substantial construction of any countermeasure, and included installation of conduit and wiring the device across a four-lane arterial street. This required investigation of possible conflicts with high-risk utilities (including contacting numerous utility providers). While individual components (the detection bollards and the beacons themselves) were commercially available, the combination was custom-designed. The detection bollards included an in-surface activation device (ISAD), which was not widely in use. The detection bollards’ effectiveness was highly sensitive to their location and placement.

2.3. Comparison of Countermeasures: Costs

The overall cost of this project was slightly greater than $1.1 million, including $681,000 in federal funding. The federal funding averaged roughly $120,000 per year.

The total costs of the nearly six-year-long project included the following rough estimates:

It should be noted that San Francisco construction and labor costs are substantially higher than the national average. On an annual basis, the project cost was actually a relatively small share of the total San Francisco public expenditures for pedestrian safety planning, design, enforcement, and outreach/education.

The federal funding was limited in comparison to the unit cost of higher-cost pedestrian safety countermeasures. This influenced the selection of countermeasures.

If this project were replicated with a focus strictly on improving pedestrian safety cost-effectively, the data collection/evaluation and other program management costs could be substantially lower than the above rough estimates. These costs were driven partly by an extensive data collection and analysis effort using extremely labor-intensive video observations and a major intercept survey of over 1,000 pedestrians. “Other Program Management Costs” include primarily SFMTA costs for such items as: preparation of major technical reports, progress reports, and contract documents; planning and design work on countermeasures that were never implemented; meetings and tele-conferences with FHWA and SAIC (including several site visits); preparation of invoices and financial reports; internal progress reporting; work planning meetings and memos; coordination with other projects and departments; and background research activities directly related to the project.

Cost estimates provided in the body of this report include: materials/equipment, installation labor, and engineering/administration labor. This includes shop and engineering/planning labor from conceptual design through fine-tuning and initial operations/maintenance. (See Table 2.3-1 for a summary of estimated capital costs. Additional detail is provided in Chapter 3, including a breakdown of estimated costs into materials, installation labor, and engineering/administrative labor. San Francisco construction costs tend to be significantly higher than national averages.)

It was not possible to track costs precisely due to the accounting system limitations. Material/equipment costs are precise figures, whereas labor costs are based on careful estimates by key project staff. Labor costs also include overhead and fringe benefits.

The Portable Changeable Message Speed Limit Sign and the outreach program carried significant operating costs for operating personnel but, once installed, other countermeasures did not require personnel for operation except for small modifications and maintenance for many of the devices. In general, the labor costs far exceeded the equipment and materials costs.

The least expensive countermeasures in total per-unit costs were the “LOOK” pavement stencils ($300 per stencil) and the “TURNING TRAFFIC MUST YIELD TO PEDESTRIANS” signs ($800/sign). The retroreflective materials averaged less than $3 per item, but several thousand were distributed. The most expensive countermeasures were the video detection system to adjust signal timing (out of pocket costs of $17,300) and the flashing beacon with automated (infrared) detection ($62,600). Although several other countermeasures appear to be higher, the video detection costs as listed are significantly understated since the PedSafe project did not pay for the video detection camera equipment or its initial installation, and substantial technical assistance was provided free of charge by Econolite, the manufacturer of the equipment.

Table 2.3-1
Summary of Estimated Capital Costs
(including installation labor, materials/equipment, and engineering/administrative costs)

In general, the engineering/administrative costs were quite substantial, and often exceeded the material/equipment costs and the installation labor. The engineering/administrative costs for the first-time use of a technology are often much higher than continuing costs. Therefore, the listed costs, especially for the most innovative or complicated countermeasures, would not be good predictors of future costs for new installations.

2.4. Comparison of Countermeasures: Availability and Standard Use

All countermeasures were explicitly or implicitly consistent with the Federal Manual on Uniform Traffic Control Devices (MUTCD) and the California MUTCD. It was not necessary to obtain special approval to experiment with any of the countermeasures from FHWA or the California Department of Transportation. Several countermeasures considered experimental when initially proposed by the San Francisco team were added to the MUTCD in the 2003 revision. Formal local approval from the Municipal Transportation Agency Board of Directors and the County Board of Supervisors were required for visibility red curb zones.

The following countermeasures are considered approved or standard devices or treatments and are now widely used in California, although they were regarded as innovative when they were initially proposed:

• In-Street Pedestrian Signs
• “TURNING TRAFFIC MUST YIELD TO PEDESTRIANS” Signs
• Modified Signal Timing
• “Pedestrian Head Start”
• Advanced Stop Lines and Red Visibility Curb Zones
• ADA Curb Ramps
• Median Refuge Islands

The following countermeasures are in limited use as “off the shelf” products, but not considered experimental:

• “LOOK” Pavement Stencils⁶
• Flashing Beacons (Push Button-Actuated)
• Portable Changeable Message Speed Limit Signs
• Changeable Message Speed Limit Signs
• Distribution of Retro-Reflective Materials
• Video Public Service Announcements

The following countermeasures were “custom-made” and involved innovative technologies, although they did not require approval as formal experiments:

• Automated (Video) Detection of Pedestrians to Extend Crossing Time
• Flashing Beacons (Automated Detection with Infrared Bollards)

⁶The LOOK pavement stencils included a Chinese-English stencil that was custom designed by SFMTA staff, working with the vendor.