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.
TCRP Web-Only Document 53 113 APPENDIX E: EVALUATION PLAN
TCRP Web-Only Document 53 114 This chapter describes how the SAIC team will evaluate the installed safety improvements and the test of higher speed LRT operations. CONDITIONS UNDER EVALUATION The evaluation will take place over three phases or âconditionsâ: ï· Existing or Baseline Conditions.âThe existing or baseline conditions represent the current crash, near-miss, and risky behaviors along the North First St. corridor at the three test intersections prior to any safety improvements or changes in speed. The purpose of collecting data for the existing conditions is to establish a baseline for the existing operations to compare against the improved and test conditions. ï· After Safety Improvement Conditions. âThis phase begins following implementation of all safety improvements (namely blank-out signs, passive signs, pavement markings, flexible post delineators and raised pavement markers) by the VTA at the test intersections. It is expected that the installation of the improvements at Brokaw Rd. and Charcot Ave. will be completed by the end of September 2009. Improvements at the Trimble Rd. crossing are expected to be completed during the summer of 2010. During these conditions, light rail operation is not expected to change; i.e., no change in headway or operating speeds. ï· After Conditions with LRV Speed Increase.âThis condition begins when all agencies have reached a mutual agreement to increase the operating speed for light rail (based on successful implementation of the supplemental safety measures) at the three study intersections, and therefore assumes that the results attained from the âAfter Safety Improvements Conditionsâ are satisfactory to the City and the VTA. During this test period, light rail speed will be increased to 40 mph. The after conditions with LRV speed increase will last for 12 months. MEASURES OF EFFECTIVENESS The research team will use a number of measures of effectiveness (MOEs) to evaluate the impacts of the intersection improvements and the higher speed LRT operations on safety as well as to determine the success of the test. Primary MOEs that will be used include: ï· Driver, pedestrian, and bicyclist behaviors (risky behaviors and compliance with traffic control devices); ï· Train operator behaviors; ï· Crashes (auto-auto and auto-train); ï· Near misses (auto-train); and ï· LRV speeds. Each of these MOEs is described in more detail below. Driver, Pedestrian, Bicyclist, and Train Operator Behaviors Specific driver, pedestrian, bicyclist, and train operator behaviors that will be collected at the test intersections as part of the evaluation include the following:
TCRP Web-Only Document 53 115 ï· Mainline left-turn change and clearance interval violationâLeft-turn motorist from North First St. enters intersection at the end of the yellow change interval or during the all-red clearance interval with and without train presence; ï· Mainline U-turn change and clearance interval violationâU-turn motorist from North First St. enters intersection at the end of the yellow change interval or during the all-red clearance interval with and without train presence; ï· Mainline left-turn red-light violationâLeft-turn motorist from North First St. enters intersection during the red interval with and without train presence; ï· Mainline through red-light violationâThrough motorist from North First St. enters intersection during the red interval without train presence; ï· Lane change violationâMotorist in either the through or right-turn lane makes an illegal lane change to turn left from North First St.; ï· Mainline U-turn red-light violationâU-turn motorist from North First St. enters intersection during the red interval with and without train presence; ï· Cross-street red-light violationâMotorist on cross-street enters intersection during the red interval with and without train presence; ï· Mainline left-turn stop bar intrusionâLeft- or U-turn motorist on North First St. stops on or beyond the stop bar during the red interval; ï· Mainline through lane stop bar intrusionâThrough motorist on North First St. stops on or beyond the stop bar during the red interval; ï· Track intrusion violationâMotorist wrongly enters track right-of-way; ï· Mainline left-turn vehicles queued on tracksâ Left-turn motorist from North First St. stops on tracks due to a queue spillback on cross-street; ï· Cross-street left-turn vehicles queued on tracksâ Left-turn motorist from cross-street stops on tracks due to a queue spill back on North First St.; ï· Cross-street through vehicles queued on tracksâ Through motorist from cross-street stops on tracks due to a queue spill back in the downstream receiving lanes. ï· Vehicle stopped on tracksâMotorist stops on tracks for reasons other than queuing, blocking, or yielding to violating motorists or pedestrians; ï· Pedestrian standing on tracksâPedestrian did not complete crossing during Walk or Flashing Donât Walk signal display and is standing between the tracks or in the track right- of-way; ï· Pedestrian intersection violationâPedestrian crosses North First St. during the Donât Walk signal display; ï· Pedestrian jay-walking violationâPedestrian crosses North First St. outside of a designated crosswalk; ï· Bicyclist mainline left-turn red-light violationâLeft turn bicyclist on North First St. enters intersection during the red interval; ï· Bicyclist cross-street red-light violationâBicyclist on cross-street enters intersection during the red interval; ï· Light rail vehicle violationâLight rail operator enters or clears intersection during the horizontal white bar LRV signal indication; and
TCRP Web-Only Document 53 116 ï· Light rail emergency brakingâLight rail operator applies emergency/maximum brakes. Risky behavior analysis will be performed for each intersection to evaluate the magnitude of each risky behavior action and its relationship to actual crashes. For example, a high frequency of red-light running is more critical to safety than the same frequency of left-turn vehicles making a turn on yellow. The analysis will focus on establishing a baseline risky behavior condition based on the prevailing traffic volumes, history of crashes, near misses, and track intrusion data as provided by the VTA and the City of San Jose. It is expected that following the implementation of the on-going improvements by the VTA, the frequency of risky behavior for left-turn conflicts should be reduced when compared to the frequency and type of conflicts under the base condition. Similar results are anticipated for track intrusion and near-miss incidents. Crashes and Near-Misses In addition to the surrogate safety measures listed above, the research team will gather and analyze crash and near miss data from the City of San Jose and the VTA. Measures of effectiveness include: ï· Number / rate (by type) of vehicle-vehicle crashes; ï· Number / rate (by type) of vehicle-train collisions; and ï· Number / rate (by type) of vehicle-train near-misses. As crashes are rare events, due to the timeframe of the evaluation, there likely will not be enough time to determine if there was a statistically significant change in actual crash rates. Therefore, auto-train near-miss data (maintained by the VTA) will also be examined. While a near-miss does not result in an actual collision, near-misses are excellent surrogate measures of safety. And as near-misses occur more frequently than actual collisions, the evaluation time frame should be sufficient to determine if there is a statistically significant change in the rate of near- misses. LRV Speeds As a final MOE to the evaluation, LRV speeds will be measured and examined to determine the actual train speeds at different points along the test corridor, particularly through the three test crossings. In order for the test to be successful, train operators must achieve speeds of 40 mph through the test crossings. DATA COLLECTION PLAN An overview of the data to be collected, the data collection dates, and the data sources is shown in Table 24. The research team has already collected a variety of baseline data, as indicated in Table 24, which has been analyzed to establish the existing or baseline conditions (these data are presented following discussion of the data collection plan).
TCRP Web-Only Document 53 117 Table 24. Overview of Data to be Collected for Evaluation Evaluation Condition Data to be Collected Estimated Date to Collect Data Data Source Existing or Baseline Conditions Traffic, pedestrian, and bicycle counts Complete City of San Jose Signal plans, signal phase sequence and signal timing Complete City of San Jose Intersection crash data (2005-2008) Complete City of San Jose Collision and near miss data (2005- 2008) Complete VTA Light rail ridership data for year 2008 Complete VTA Observations of driver, pedestrian, and bicyclist behaviors Complete Field video data collected by SAIC team After Safety Improvement Conditions Crash data for intersections of: --North First St. & Brokaw Rd. --North First St. & Charcot Ave. Every 3 months following implementation of safety improvements: Nov 2009, Feb 2010, May 2010 City of San Jose Collision and near miss data for intersections of: --North First St. & Brokaw Rd. --North First St. & Charcot Ave. Every 3 months following implementation of safety improvements: Nov 2009, Feb 2010, May 2010 VTA Observations of driver, pedestrian, bicyclist, and train operator behaviors Nov 2009 Field video data collected by SAIC team Traffic, pedestrian, and bicycle counts Nov 2009 City of San Jose or samples from video data After Conditions with LRV Speed Increase Crash data for intersections of: --North First St. & Brokaw Rd. --North First St. & Charcot Ave. --North First St. & Trimble Rd. Every 3 months following increase in train speeds: Sept 2010, Dec 2010, Mar 2011, Jun 2011 City of San Jose Collision and near miss data for intersections of: --North First St. & Brokaw Rd. --North First St. & Charcot Ave. --North First St. & Trimble Rd. Every 3 months following increase in train speeds: Sept 2010, Dec 2010, Mar 2011, Jun 2011 VTA On-board measurement of LRV speeds and observations of driver, pedestrian, and bicyclist behaviors through crossings Sept 2010 On-board data collected by SAIC team Traffic, pedestrian, and bicycle counts Sept 2010, June 2011 City of San Jose or samples from video data To understand the impact of the safety improvements on driver, pedestrian, bicyclist, and train operator behaviors (risky behaviors and compliance with traffic control devices) at the intersections, the research team will collect a second round of data, which will represent the âAfter Safety Improvement Conditions.â Currently, the research team plans to begin the data collection for these conditions in November 2009. As improvements at the intersections of North First St. and Brokaw and North First St. are expected to be complete by the end of
TCRP Web-Only Document 53 118 September, this will allow time for drivers, pedestrians, and bicyclists to become accustomed to the changes at the intersections before data collection commences. This second phase of data collection is planned for mid-November 2009 at the Brokaw Rd. and Charcot Ave. crossings. The team will follow the same data collection methodology for the video observations performed for the existing conditions. In addition, the team will request updated traffic, pedestrian, and bicyclist counts from the City of San Jose, if such data have been collected. If not, the team will obtain samples from the video data. Traffic, pedestrian, and bicyclist counts will be compared with the baseline conditions to determine if there has been a change in counts between the two conditions. The team will also request crash data from the city, as well as collision and near- miss data from the VTA every 3 months following implementation of the improvements to determine if the improvements had any impact on safety. Finally, to understand the impact of the increase in train speeds to 40 mph, the research team will collect a third round of data, which will represent the âAfter Conditions with LRV Speed Increase.â This third round of data collection will begin approximately 3 months following the speed increase, which is anticipated for June 2010. The primary MOEs for the test conditions will be the crash data (collected every 3 months beginning in September 2010) from the City of San Jose and the collision and near-miss data (collected every 3 months beginning in September 2010) from the VTA. Due to the resources needed to conduct, reduce, and analyze an additional round of field video data collection, the research team is proposing instead to have two data collectors ride LRVs between the Metro / Airport Station northbound to the Bonaventura Station and between the Bonaventura Station southbound to the Metro / Airport Station. Both the research team and the VTA feel that there is little reason to expect any changes in driver behavior during this phase of the evaluation (any behavioral changes that would result from the safety improvements would be captured during the âAfter Safety Improvement Conditionsâ). During the rides, data collectors will carry portable GPS devices, which will record LRV speeds northbound and southbound between the two stations. Data collectors will be stationed in the train operator cab as a âsilentâ observer and will record driver, pedestrian, and bicyclist behaviors at the intersections, as well as near misses. In addition, traffic, pedestrian, and bicyclist counts will be compared with those of the two previous conditions to determine if exposure levels have changed. DATA COLLECTION METHODOLOGY Field Video Data Collection Methodology For the first two evaluation conditions, video data will be collected to identify driver, pedestrian, bicyclist, and train operator behaviors (e.g., risky behaviors and compliance with traffic control devices). Baseline field video data were collected in June 2009, as was described in Chapter 3. Field video data will also be collected in the After Safety Improvement Conditions in the exact same manner that they were collected in the Baseline Conditions (see Chapter 3 for a complete description of the field video data collection methodology). Methodology for On-Board Observations Two data collectors will ride the trains between the Metro / Airport station and the Bonaventura station for 5 consecutive days during one week. The data collectors will ride the trains between the two stations during the designated data collection hours, which will be identical to those used
TCRP Web-Only Document 53 119 in the video data collection methodology (i.e., 7:00 a.m. to 9:00 a.m., 11:00 a.m. to 2:00 p.m., and 4:00 p.m. to 7:00 p.m.). The first data collector will begin at the Metro / Airport Station at 7:00 a.m. and will board the first northbound train. The data collector will ride the train to the Bonaventura station. At the Bonaventura station, the data collector will exit the northbound train and will board the next southbound train. Once back at the Metro / Airport station, the data collector will repeat the processes of riding the trains between the two stations. The second data collector will go through the same process but on the next train following the first data collector. During the rides, data collectors will carry portable GPS devices, which will record LRV speeds northbound and southbound between the two stations. The GPS data will allow an analysis of the speed profiles northbound and southbound between the two stations. The speed profiles will indicate if operators were able to achieve 40 mph and if so at what point and over what distance. As the speed data will be collected passively by the portable GPS devices, the data collectors will be free to observe what is happening both outside the LRV and inside the operatorâs cab. Specifically, data collectors will be trained to watch for unusual / unsafe driver, pedestrian, and bicyclist behaviors when the LRV is approaching the crossings, as well as when the LRV is in the crossing. Data collectors will record any observed behaviors or near miss incidents on a data collection sheet. Assuming a train headway of 7.5 seconds in each direction, there would be approximately 64 northbound and 64 southbound trains during one day (8 hours) of data collection. Two data collectors should be able to ride about one-half of the trains, meaning 32 trains per day northbound and 32 trains per day southbound. As the test of higher speeds will be through three intersections in the northbound direction, this would allow for 96 observations. With a single test intersection in the southbound direction, this would allow for 32 crossing observations per day. Over a 5-day data collection period, there would be a total of approximately 480 crossing observations in the northbound direction and 160 crossing observations in the southbound direction for a total of 640 crossing observations.