Oregon Regional Intelligent Transportation
Systems (ITS) Integration Program
Final Phase II Report
Submitted to:
United States
Department of Transportation
Science
Applications International Corporation (SAIC)
Cambridge Systematics,
Inc.
July 12, 2002
TABLE OF CONTENTS
1.3.2 I-5/Barbur Boulevard
Parallel Corridor Traffic Management Demonstration Project
1.3.3 COATS Bi-State Rural
Integration
1.3.4 Transit Buses as
Traffic Probes
2.1.1 Use of Trip Planning
Information
2.1.3 Perceived Personal
Security
2.1.4 Overall Customer
Satisfaction
2.2.1 Tri-Met Customer
Satisfaction Survey
2.2.2 Baseline Transit
Tracker Intercept Survey
2.2.3 Comparison of
Baseline Results from Tri-Met Survey and
SAIC Survey
3 I-5/Barbur
Boulevard Parallel Corridor Traffic Management Demonstration Project
3.1.1 Data Collection
Approach
3.2 Customer
Satisfaction Study
3.2.1 Data Collection
Approach
4 COATS Bi-State
Rural Integration
5.1 Deployment
Plans and Schedules
5.1.1 Transit Tracker
Information Displays
5.1.2 I-5/Barbur Boulevard
Parallel Corridor Traffic Management Demonstration Project
5.1.3 COATS Bi-State Rural
Integration Project
5.1.4 Transit Buses as
Traffic Probes Project
Figure 1‑1. Transit Tracker Display
Figure 1‑2. On-Line Transit Tracker Information
Figure 1‑3. CCTV View of I-5 at Barbur Blvd.
Figure 1‑4. Map of I-5/Barbur Boulevard Study Area
Figure 1‑5. Tri-Met Bus Dispatch Station
Figure 2‑1. Age Distribution
of Survey Respondents Compared to 2001 Tri-Met Survey
Figure 2‑2. Age Distribution
of Survey Respondents Compared to 2000 and 2001 Tri-Met Surveys
Figure 2‑3. Frequency with
Which Respondents Ride the Bus
Figure 2‑4. Distribution of
Most Frequent Trip Purposes
Figure 2‑5. Frequency With
Which Respondents Use Bus Schedule Information
Figure 2‑6. Alternative Trip
Planning Behaviors
Figure 2‑7. Respondents’
Perceptions of Accuracy of Schedule Information
Figure 2‑8. Comparison of Auto
Ownership and Use of Schedule Information
Figure 2‑9. Distribution of
Responses for Those Reporting an Integer for
Wait Time
Figure 2‑10. Distribution of
Responses for Those Giving a Range for Wait Time
Figure 2‑11. Riders’ Level of
Satisfaction with the Bus’ Adherence to the Posted Schedules
Figure 2‑12. Riders’
Perceptions of Security
Figure 2‑13. Riders’
Perceptions of Service Quality
Figure 3‑1. I-5 Northbound/Barbur Boulevard Study Area
Figure 3‑2. I-5 Northbound AM Peak Period Average Hourly
Volumes
Figure 3‑3. I-5 Northbound AM Peak Period Average Speed
Figure 3‑4. I-5 Northbound AM Peak Period Average Travel
Time
Figure 3‑5. I-5 Northbound AM Peak 30-min (7:30-8:00 AM)
Volumes
Figure 3‑6. I-5 Northbound AM Peak 30-min (7:30-8:00 AM)
Speed
Figure 3‑7. Barbur Boulevard Northbound AM Peak Hourly
Volume
Figure 3‑9. Barbur Boulevard Northbound AM Peak Average
Speed
(Baseline Speed from Hose Counters)
Figure 3‑10. I-5 and Barbur Boulevard Northbound AM
Speed by
Day of Week
Figure 3‑11. I-5 and Barbur Boulevard Northbound AM
Speed by
Hour of Day
Figure 3‑12. I-5 Northbound AM Peak Monthly Crash Rate
Figure 3‑13. I-5 Northbound AM Peak Changes in Speed due
to Incidents
Figure 3‑14. I-5 Northbound AM Peak Travel Time During Incident
Conditions
Figure 3‑15. Age Distribution
of Panel Members
Figure 3‑16. Years Panel Members Have Been Commuting in
Portland
Figure 3‑17. Respondents’
Commute Times and Distances
Figure 3‑19. Frequency With
Which Panel Members Use Barbur Boulevard to Avoid Delays on I-5
Figure 3‑20. Panel Members’
Satisfaction Ratings for Northbound I-5
Figure 4‑1. Historic Crash
Rates at Selected Rural Oregon Highways
Table 2‑1. Ridership Data for
Four Bus Stops in Portland
Table 2‑2. Number of Completed
Baseline Customer Satisfaction Surveys.
Table 2‑3. Automobile Ownership Versus Frequency of Bus
Use
Table 2‑4. Survey Questions
About Riders’ Trip Planning Behavior
Table 2‑6. Comparison of Use of Trip-Planning
Information
Table 2‑7. Χ2 Results
for Auto Ownership Vs. Use of Schedule Information
Table 2‑8. Χ2 Results for Age Vs. Use of
Schedule Information
Table 2‑9. Nighttime Boardings
for Spring 2001
Table 3‑1. Summary of Northbound AM Freeway and Arterial Traffic Performance
Table 3‑2. Summary of Freeway Incidents During
Evaluation Period
Table 3‑3. Summary of Northbound AM Baseline Traffic
Performance During Incident Conditions
Table 3‑4. Customer
Satisfaction Survey—Qualifying Questions
Table 3‑5. Companies Contacted
for Survey Participation
Table 3‑6. Panel Member Comments Regarding Morning
Commute Time
Table 3‑7. Panel Member Comments Regarding VMS
Table 3‑8. Panel Member Comments Regarding Barbur
Boulevard as a Primary Route Choice
Table 3‑9. Comments About Traffic Signals and Incident
Operations on Barbur Boulevard
Table 3‑10. Summary of Panel
Members’ Ratings of Traffic Information
Table 3‑11. Comments About Accuracy and Timeliness of
Traffic Information
Table 4‑1. COATS Sites
Identified for Crash Data Analysis
Table 4‑2. Summary of Incident
Conditions
For further information
please contact:
Kelley K. Pécheux
SAIC
P.O. Box 50132
8301 Greensboro Dr.
McLean, VA 22101
703.676.2385
Mark R. Carter
SAIC
P.O. Box 50132
8301 Greensboro Dr.
McLean, VA 22101
202.366.2196
Andre Chandra
CSI
1300 Clay Street
Oakland, CA 94612
510.873.8700
The Portland metropolitan region has a long history of investing in multi-modal transportation solutions to enhance mobility and maintain the region’s livability standards and reputation. As a complementary means to enhance the efficiency and safety of travel, the Portland region has been actively involved in the planning and deployment of Intelligent Transportation Systems (ITS) since the early 1990s. Portland’s commitment to ITS as a solution to transportation problems is evidenced by TransPort 2000, a regional ITS plan consisting of 16 integrated and interoperable projects. Transport 2000 includes significant bi-state, urban-rural, and multi-modal components and represents the joint planning efforts of a regional committee consistent with statewide and regional planning processes in Oregon and Washington. The Transport 2000 projects build on the region’s significant existing ITS investment and fills current gaps in planning, emergency management, traveler information, and parking subsystems.
Under the direction and partial funding of the United States Department of Transportation (USDOT), National ITS evaluations are being conducted to accelerate the integration and interoperability of ITS in metropolitan and rural areas. To investigate the success of ITS across the country and to provide insights into the potential strengths and weaknesses of the overall national integration program, the Oregon Regional ITS Integration Program was selected for independent national evaluation. Specifically, four projects were selected for evaluation:
· Regional Intermodal Transit Traveler Information and Security System—Transit Tracker Information Displays,
· I-5/Barbur Boulevard Parallel Corridor Traffic Management Demonstration Project,
· COATS Bi-State Rural Integration Project, and
· Transit Buses as Traffic Probes project.
This document outlines the evaluation strategies, data collection plans, and baseline results for the Oregon Regional ITS Integration Program.
Regional Intermodal Transit Traveler Information and
Security System—Transit Tracker Information Displays
Transit Tracker information displays, a component of the Regional Intermodal Transit Traveler Information and Security System, use global positioning satellite technology and algorithms to calculate real-time bus and train arrivals, which are displayed at transit stops. The signs display a minute-by-minute countdown of the arrival time of the buses once they are within some pre-selected time of arriving at the stop (generally 15 to 25 minutes). If a bus is outside of this time threshold, the signs simply show the scheduled arrival time. The focus of this project is to make the bus and light rail arrival information available to riders through the information displays.
Evaluation Approach
The purpose of the Transit Tracker evaluation is to collect and analyze data related to a change in bus riders’ behaviors and perceptions of service and security as a result of the Transit Tracker information displays. Four measures of effectiveness (MOEs) were selected to test the impact of the information on riders’ behaviors and perceptions:
· Riders’ use of trip planning information,
· Riders’ perceptions of system efficiency,
· Riders’ perceptions of personal security, and
· Riders’ overall satisfaction with the system.
In addition, Bus Dispatch System (BDS) data will be examined in Phase III to determine accuracy of the system information, and website use statistics will be used to determine if the real-time information increases customer use of the Tri-Met website. These results, along with qualitative lessons learned during implementation and operation of the system, will be documented in the Phase III report.
Findings
The SAIC team conducted baseline Transit Tracker intercept surveys in Portland for four days in January 2002. Two teams of two surveyors went to different bus stops during the morning peak, over the lunch hour, and during the evening peak for three days. In all, 240 completed surveys were obtained.
Riders’ use of trip
planning information
Between 41 and 76 percent of respondents indicated that they rarely or almost never use four types of fixed-schedule information, depending on the type of information (i.e., printed brochure schedules, schedules posted at bus stops, on-line Internet schedule, and 238-RIDE phone number). A couple of reasons why riders may not use or need fixed-schedule information were identified. Forty-five percent of respondents indicated that they frequently or almost always just go to the bus stop and wait for the next bus to arrive (not knowing the scheduled arrival time). Over half of respondents indicated that they do not use schedule information because they frequently or almost always have their routes and times memorized. Chi-squared tests showed that the use of schedule information is dependent on automobile ownership. Specifically, non-auto owners tend to use guides posted at stops and the 238-RIDE phone number more often than auto owners; and auto owners tend to use the Internet schedule more often than non-auto owners.
When asked about the accuracy of the fixed-schedule information, 70 percent indicated that the information that they do use is frequently or almost always accurate.
Riders’ perceptions of
system efficiency
Rider perception of on-time performance was used as a measure of system efficiency. About 73 percent of respondents reported that the bus they catch at the stop is usually on time. Only 10 percent reported that the bus is not usually on time, and about 16 percent reported that they did not know if the bus was usually on time (either because they had never been to the stop before, or because they did not know the scheduled arrival time). When asked how long they usually wait for the bus at the stop, 26 percent of the respondents gave a range, while the remaining 74 percent reported an integer value. The most common response was 5 minutes (42 percent of respondents), followed by 10 minutes (32 percent of respondents), and 15 minutes (13 percent of respondents). The average number of minutes of the respondents who gave an integer value for wait time was 8.6 minutes. The most common response for those reporting a range was 5 – 10 minutes (46 percent) with the second most common response being 10 – 15 minutes (17 percent). Taking an average of each range, the average wait time reported for those giving a range was 9.2 minutes (only 0.6 minutes higher than those reporting an integer value).
When asked how satisfied they were with the bus’ adherence
to the posted schedules, 91 percent of respondents indicated that they were
either satisfied or extremely satisfied. In other words, respondents seem to be very
satisfied with bus service in terms of on-time performance.
Riders’ perceptions of
personal security
An overwhelming 97 percent of respondents reported that they agree or completely agree that they feel safe waiting for the bus at the stops during the day. Only about 63 percent of the same riders reported that they agree or completely agree that they feel safe waiting for the bus at the stops at night, while 20 percent reported that they disagree or completely disagree that they feel safe at night. Therefore, while it may be difficult, if not impossible, for Transit Tracker to have much impact on bus riders’ perceptions of personal security during the day, there does exist some room for improvement at night.
Average nighttime (9:00 p.m. to close of service) boardings from the Spring of 2001 were also examined to see if Transit Tracker might increase perceived personal security at night. (It was hypothesized that if there is an increased sense of security related to the presence of Transit Tracker, riders may feel safer riding at night, when they normally may not be comfortable riding the bus.) The nighttime ridership numbers for the four stops ranged from only 8 to 16 riders. While the numbers are low, they are not uncommon, and they will be compared with nighttime ridership numbers after Transit Tracker has been installed at the same locations. An increase in ridership after Transit Tracker installation may be indicative of an increased sense of security at night; however, overall ridership numbers will also be examined to determine if there is any corresponding upwards trend in overall ridership that may account for the increase at night.
The baseline survey conducted by SAIC was compared to a baseline survey of 830 transit riders that was conducted by Tri-Met in the spring of 2000. The results of the two surveys, in terms of riders’ perceptions of efficiency, personal security, and overall customer satisfaction, were very similar.
Riders’ overall
satisfaction with the system
Respondents were asked to rate how satisfied they are with the bus service at the stops surveyed. An overwhelming 91 percent of respondents indicated that they are either satisfied or completely satisfied with the bus service at the stops, while only 4.5 percent reported being dissatisfied or completely dissatisfied.
I-5/Barbur Boulevard Parallel
Corridor Traffic Management Demonstration Project
The City of Portland and the Oregon Department of Transportation (ODOT) are implementing cooperative strategies to test the deployment of ITS on a parallel freeway/arterial corridor. As part of the regional advanced traffic management system (ATMS) program, traffic and incident management along the regional freeway and arterial systems are planned with freeway/arterial integration. Specifically, this project will accelerate the current deployment and integration of traffic surveillance and control devices in a high volume freeway/arterial corridor.
Evaluation Approach
The baseline data were obtained through field data collection along Barbur Boulevard and I-5 (e.g., volumes, speeds, crash rates) and a web-based driver survey (e.g., drivers’ perceptions and behaviors). The results from the mobility/safety study and the customer satisfaction study will be compared to similar data collected after installation of the system during Phase III data collection. These results, along with qualitative lessons learned during implementation and operation of the system, will be documented in the Phase III report.
The mobility/safety study on I-5 and Barbur Boulevard aims to discover impacts of the freeway-arterial integration on traveler mobility, using the following measures of effectiveness:
· Speed/travel time in the primary direction during incident conditions,
· Speed/travel time reliability in the primary direction during incident conditions,
· Vehicle throughput in the primary direction during incident conditions,
· Incident detection, response and clearance times, and
· Freeway crash rates.
In order to test the impacts of the corridor traffic management project on customer satisfaction, an understanding of baseline perceptions is required. Thus, the initial questionnaire was designed to obtain baseline information on a variety of issues including: commuter demographics, commute time and distance, access to and use of traveler information, commute patterns, frequency of incident-related delays, and perceptions of traffic conditions on I-5 and Barbur Boulevard.
Findings
Mobility/safety study
On average over the peak period, traffic on northbound I-5 during the morning peak period has remained the same for the last two years. I-5 is the major corridor connecting downtown Portland with the suburban areas to the north and south. The I-5 northbound corridor within the study area has three lanes, with a posted speed limit of 55 mph. It carries about 3,900 vehicles per hour during the morning peak period at average speeds of 45-55 mph, depending on the segment. Speed and travel time reliabilities have also been consistent the last two years, averaging about +/- 9 mph and +/- 2.5 minutes, respectively.
A closer inspection of the volume and speed data during the morning peak 30-minute period revealed a big difference between historic and baseline traffic. Between 7:30 and 8:00 a.m., which is one of the corridor’s busiest 30-minute periods, northbound I-5 experienced a volume increase of 42.9 percent since 2000/2001. This finding was also confirmed with a more noticeable drop in speeds found during the same 30-minute period between historic and baseline evaluation periods.
Similar analysis was done on Barbur Boulevard traffic volumes; however, there are not enough historic volume data to warrant definitive conclusions, as data were obtained only from three dates in 2000. Nevertheless, the comparison between historic and baseline Barbur Boulevard volumes shows that northbound traffic volumes have been consistent since 2000, with slight increases between 7 and 9 a.m.
Barbur Boulevard, although having a lower average speed than I-5, had excellent speed reliability compared to I-5, both throughout the peak period and on average throughout the week. For example, while I-5 showed slightly lower average speeds on Wednesday and Friday, average speeds on Barbur Boulevard remained very consistent.
The results of the freeway crash analysis show that the average number of crashes per month for the past two years has been declining. Injury crash rates have remained roughly the same for the last two years, at a rate of about two to three injury crashes per month. Crashes that only resulted in property damages, however, have steadily decreased, from 58 crashes per month in 2000 to only 48 crashes in January 2002. The evaluation team found no obvious reason for this consistent decline in PDO crash rates, but it is perhaps worth noting that the 2002 statistics only include the month of January. For a more representative 2002 crash rates, a few more months of data would be necessary.
Incident reports coinciding with the evaluation time periods were also analyzed. This analysis revealed that there were seven incidents occurring in the northbound direction of I-5 during the morning peak, with no secondary incidents. The duration of the incidents ranged from 46 minutes to two hours, but averaged just over an hour.
On average, vehicle speed declined by four to six miles per hour after an incident. Individual incident characteristics show that incidents that occurred early in the peak period suffered the greatest, as the recoveries were slowed by the increasing peak period traffic. On the other hand, incidents that occurred later in the peak experienced less impact, as peak period traffic was most likely beginning to clear. The average decrease in speeds translates to about one minute of travel time increase. At 30 minutes prior to the incidents’ confirmation, the average travel time on I-5 northbound was 8.4 minutes, which increased to over 9 minutes after the incident occurred.
By comparing standard deviations of speed with and without the incidents, speed and travel time on I-5 northbound became less reliable (increased standard deviation) during incident conditions, with speed standard deviation increasing by 23 percent and travel time standard deviation increasing by 35 percent. Comparing the speeds on I-5 and Barbur Boulevard, there is evidence that incidents caused traffic to deviate from the freeway mainlines to the arterial, as hourly Barbur Boulevard traffic volumes increased by an average of 18 percent during incident conditions, and Barbur Boulevard speeds decreased by about 8 percent. Also, speeds on Barbur Boulevard were 28 percent less reliable during incident conditions.
Customer satisfaction
study
Four hundred sixty downtown Portland commuters qualified to participate in the I-5/Barbur Boulevard customer satisfaction panel survey. The following is a summary of the key findings of the baseline customer satisfaction panel survey:
· 61 percent of the panel members generally do not find out about delays on their route until after they see the congestion on the roadway. In other words, they do not receive traffic information soon enough to make important decisions about their commute time and route.
· Of the respondents who reported that they generally find out about incident-related delays before they leave home, about 65 percent reported that they generally use an alternate route that does not include a freeway, and about 15 percent indicated that they make no change.
· The respondents who reported that they generally do not find out about incident-related delays on their route until after they leave home were also asked to indicate what they usually do to avoid the delays when they become aware of them. Nearly 56 percent reported that they generally use an alternate route that does not include a freeway, while about 38 percent indicated that they make no change at all.
· The most common answer as to how often they experience incident-related delays on northbound I-5 in the morning was less than once per month (given by about 22 percent of the panel members). About 21 percent of the panel members reported that they experience incident-related delays in the study area during their morning commute two times per month. Nearly 25 percent, however, reported delays four or more times per month, or nearly once per week on average.
· About 73 percent of panel members indicated that they do use Barbur Boulevard to avoid delays on I-5.
· While nearly 36 percent of panel members reported being satisfied with traffic operations on northbound I-5 in the morning during a typical commute, 33 percent reported that they were either dissatisfied or extremely dissatisfied with normal traffic operations. On the other hand, only about 5 percent of panel members reported that they were satisfied with traffic operations during incident conditions, and nearly 78 percent reported being dissatisfied or extremely dissatisfied. These ratings indicate that there is much room for improvement of drivers’ satisfaction with traffic operations along I-5 during normal operations and during incidents.
· Radio reports in the car are perceived by more panel members as being timely and useful compared to the other information sources and were rated by the fewest panel members as being rarely or almost never timely or useful. Freeway VMS were rated by more panel members as being frequently or almost always accurate than the other information sources considered.
· Overall, the panel members reported the traffic information to be more accurate than either timely or useful.
While a sample of 460 drivers is a large enough sample to be representative of the population of commuters on northbound I-5 in the morning, the method of survey administration (i.e., Internet) limits the applicability of the results. In other words, the opinions of the sample are representative of I-5 commuters who work in an office setting and have access to a computer/Internet and can provide valuable information about their behaviors and perceptions; however, the results cannot be generalized to the population as a whole.
COATS Bi-State Rural Integration Project
The California Oregon Advanced Transportation System (COATS) is a project that seeks to encourage regional, public, and private sector cooperation between California and Oregon to better facilitate the planning and implementation of ITS in the bi-state area. The COATS study area includes 13 counties in northern California and the southern half of Oregon and is not defined by county lines but rather by roadway segments. The intent of this project is to facilitate the use of ITS to enhance safety, improve the movement of people, goods, and services, and subsequently promote the economic development of the region.
Evaluation Approach
Measures of effectiveness originally selected for this
evaluation included efficiency, safety, and customer satisfaction. However, the original evaluation plan was
adjusted to as a result of a concurrent
evaluation of COATS being conducted by the Western Transportation Institute at
Montana State University, Bozeman. The
scope of this national evaluation is now concentrated solely on the safety
analysis. Therefore, pre-deployment
crash rates were the focus of the Phase II data collection, and the results
will be compared to similar data collected after installation of the COATS
projects. These results will be
documented in the Phase III report.
Findings
Crash statistics were obtained from ODOT’s 2000 Statewide Crash Rate Tables, to
serve as the baseline for this study.
Four segments that are currently in the pre-deployment stage were
selected for analysis:
·
OR 242 between MP 55 and
Sisters (advanced warning system for narrow lane widths),
·
OR 42S between US 101
and OR 242 (automated flood warning system),
·
US 97 between MP 143 and
the Klamath County Line (animal detection system), and
·
US 101 between Coos Bay
and OR 42 (automated flood warning system).
In general, year 2000 crash rates in rural Oregon, based on data from the studied segments, averaged 1.12 crashes per million vehicle miles of travel (VMT), ranging as low as 0.73 at OR 42S, to as high as 1.39 at OR 242. Crash rates at all of the study segments, except at US 97 dropped in 2000, with an average reduction of 30 percent. On the other hand, US 97 experienced an increase of 0.5 crashes per million VMT when compared to the average rates from 1996 to 1999. The reason for the drops in incident rates is unknown. On the other hand, US 97 experienced an increase of 0.5 crashes per million VMT when compared to the average rates from 1996 to 1999.
Transit Buses as Traffic Probes Project
This regional transit-traffic management integration project addresses the technically challenging integration and utilization of real-time transit data for the purpose of establishing arterial (and freeway) network status. This project will support the use of travel time data for real-time management of traffic signals as well as analysis of corridor performance. Tri-Met buses traveling along their normal routes (in regular revenue service) will be collecting data as they normally do. Appropriate bus routes will be chosen in order to capture data from desired arterials. The real-time collection of travel times on several river crossings are of particular interest, as these tend to be congestion points in the system, and buses are traveling with the mixed-flow traffic on these bridges.
The information collected from the buses will be color-coded and added to the ODOT network status maps that are available on ODOT’s website. These data will allow the agencies to better monitor and manage the transportation system and will fill gaps in network management.
Evaluation Approach
For this project, data gathered will not focus on “before” and “after” system deployment, as with the other three projects. Instead, data will be collected after system deployment, and information will be presented in a case study format. The case study will reference quantitative data (such as the additional amount of roadway from which real-time traffic data are generated as a result of using buses as probes). The case study will address the reliability and accuracy of the bus probe data, the utility of the information gathered to traffic managers, and the institutional issues associated with this type of project. For the institutional issues study, interviews will be held with ODOT, City of Portland, and Tri-Met personnel.
Evaluation Risk Assessment
The continuation of the evaluation of the Regional Oregon ITS Integration Program offers significant opportunities, with little to no risk. Based on these opportunities and the evaluation team’s experience in developing the evaluation plan, working with the project partners, collecting baseline data, and analyzing baseline conditions, the evaluation team recommends that the FHWA COTR consider continuing with Phase III evaluation efforts.
This document outlines the evaluation strategies, data collection approach, and baseline results for the Oregon Regional Intelligent Transportation System (ITS) Integration Program. Under the direction and partial funding of the United States Department of Transportation (USDOT), National ITS evaluations are being conducted to accelerate the integration and interoperability of ITS in metropolitan and rural areas.
The Portland metropolitan region has a long history of
investing in multi-modal transportation solutions to enhance mobility and
maintain the region’s livability standards and reputation.(1) As a complementary
means to enhance the efficiency and safety of travel, the Portland region has
been actively involved in the planning and deployment of ITS infrastructure
since the early 1990s. Considering a
forecast regional population increase of nearly 500,000 residents by the year
2040 and a related increase of 55 percent in vehicle miles traveled (VMT) on
regional transportation facilities in the next 20 years, integrated ITS
deployment is considered critical in meeting future transportation demands.(2)
To investigate the success of ITS across the country and to provide insights into the potential strengths and weaknesses of the overall national integration program, the Oregon Regional ITS Integration Program was selected for independent national evaluation. Specifically, four projects were selected for evaluation:
· Regional Intermodal Transit Traveler Information and Security System—Transit Tracker Information Displays,
· I-5/Barbur Boulevard Parallel Corridor Traffic Management Demonstration Project,
· COATS Bi-State Rural Integration Project, and
· Transit Buses as Traffic Probes project.
This Phase II report represents the second major deliverable of the evaluation effort. The Evaluation Plan, which presents the detailed objectives, hypotheses, and data needs for each evaluation goal, was the first deliverable.(3) The next major deliverable will be the Phase III report, which will include before-and-after analyses of the projects’ impacts on system performance, safety, and customer satisfaction by comparing the data collected in Phase II and Phase III of the evaluation. This document presents the plan for conducting the independent evaluation and is structured in the following format:
· Section 1 – Introduction – Provides background information on the projects, including project participants, system components, and system objectives.
· Section 2 – Regional Intermodal Transit Traveler Information and Security System—Transit Tracker Information Displays – Details the data collection plan, data collection process, and baseline results related to riders’ perceptions of service efficiency, trip planning information, and safety.
· Section 3 – I-5/Barbur Boulevard Parallel Corridor Traffic Management Demonstration Project – Details the data collection plans, data collection processes, and baseline results related to system efficiency, safety, and customer satisfaction.
· Section 4 – COATS Bi-State Rural Integration Project – Details the data collection plan, data collection process, and baseline results related to traveler safety in rural Oregon.
The Transit Bus as Traffic Probes project is not addressed in this Phase II report, as there are no “baseline” data to be collected. Due to the nature of the project, the evaluation will be written as a case study (with quantitative data) and will be presented in the Phase III report.
Portland’s commitment to ITS as a solution to transportation problems is evidenced by TransPort 2000, a regional ITS plan consisting of 16 integrated and interoperable projects. Transport 2000 includes significant bi-state, urban-rural, and multi-modal components and represents the joint planning efforts of a regional committee consistent with statewide and regional planning processes in Oregon and Washington. The Transport 2000 projects build on the region’s significant existing ITS infrastructure investment and fills current gaps in planning, emergency management, traveler information, and parking subsystems. These projects integrate:
· Transit with ATIS,
· Transit with freeway and arterial management,
· Freeway and arterial management,
· Arterial and incident management, and
· Freeway and incident management.
When deployed, the projects will complete the region’s core intelligent transportation infrastructure consistent with the USDOT’s National ITS goals for safety, efficiency, productivity, mobility, and environmental improvements.(2)
The project partners have been working to continue progress on the incremental conceptualization, planning, design, and deployment of operational ITS. The Transport 2000 partners include the following agencies:
· Washington and Oregon Departments of Transportation (WSDOT and ODOT);
· City of Portland Department of Transportation;
· Metro—The directly-elected regional government that offers a wide range of services, including transportation and land use planning, to more than 1.3 million residents in three counties and 24 cities. By working with residents and local and state partners, Metro’s goal is to provide effective transportation options to move people and goods throughout the region;
· Southwest Washington Regional Transportation Council (RTC)—The metropolitan planning organization for Southwest Washington. RTC’s mission is to minimize transportation-related air pollution and to encourage and promote the development of a balanced, efficient, and affordable regional transportation system that meets the mobility needs of people and goods, within and through the Southwest Washington region;
· The Port of Portland—Owns and maintains five marine terminals, four airports (including Portland International) and seven business parks. Its mission is to provide competitive cargo and passenger access to regional, national, and international markets while enhancing the region's quality of life;
· Tri-County Metropolitan Transportation District of Oregon (Tri-MET)—Provides public transit service for the Portland metropolitan area. Based on Tri-Met’s goals and objectives of increased mobility, increased system performance, reduced costs, and improved customer satisfaction, the agency has proposed ten ITS projects, three of which are part of Transport 2000;(1)
· C-TRAN—Clark County's public transit provider operates 29 buses, C-VAN curb-to-curb service for people who cannot access regular route service, carpool and vanpool services, and a Bike & Bus program; and
· Academic and private partners.
The Oregon Regional ITS Integration Program includes 16 different projects. The following four projects were selected for national evaluation:
· Regional Intermodal Transit Traveler Information and Security System—Transit Tracker Information Displays,
· I-5/Barbur Boulevard Parallel Corridor Traffic Management Demonstration Project,
· COATS Bi-State Rural Integration, and
· Transit Buses as Traffic Probes.
These projects cover a range of transportation modes and a host of intelligent transportation technologies. Three of the projects are targeted at improving transportation in the Portland metropolitan region, while the COATS project is focused on a primarily rural area of Southwest Oregon and Northern California. This section describes each of these projects in more detail and lists the expected benefits of the projects, as defined by the project partners.
Transit Tracker information displays are part of the larger Regional Intermodal Transit Traveler Information and Security System. The goal of this overall system is to provide a seamless and complete regional multi-modal traveler information system that will result in more complete information service and enhanced public transportation security. This system will serve transit riders with a variety of services and information including interactive access to schedule and fare information, trip planning, and en-route information on Tri-Met bus and light rail, C-TRAN bus, and transit service to and from Portland International Airport. The system will also provide access to integrated regional traffic and transit information from fixed sites or through personal access services (e.g., Internet, dial-in services). For public transportation security, this project will provide for integrated emergency notification and response throughout the four county service region, exclusive of transit provider or current vehicle location.
Tri-Met has an extensive
operational traveler information distribution infrastructure and is in the
process of completing the installation of bus mall kiosks, smart bus signs,
interactive pylons, and automated ticket vending machines. For seamless regional transit operations, it
is expected that C-TRAN will also provide an equivalent level of traveler
information service and security to its customers.
Figure 1‑1. Transit Tracker Display
Tri-Met has installed its prototype Transit Tracker information displays. Transit Trackers use global positioning system (GPS) technology to calculate real-time bus and train arrival information, which is displayed on the signs at transit stops. The signs display a minute-by-minute countdown of the arrival time of the transit vehicles once they are within a pre-selected time of arriving at the stop (generally 15 to 25 minutes, a parameter that is selected by Tri-Met). If the bus or train is outside of this time threshold, the signs simply show the scheduled arrival time (Figure 1‑1). The focus of this project is to make the arrival information available to riders through the information displays.
The information is also available on Tri-Met’s website (Figure 1‑2). Using the on-line Transit Tracker information, riders can check that their bus or train is on time before leaving home or work.
Figure 1‑2. On-Line Transit Tracker Information
While Tri-Met is skeptical about the ability of the Transit Tracker displays, per se, to increase ridership, they believe that the information is a good tool to reduce the anxiety and frustration sometimes associated with riding transit, especially when using an unfamiliar transit system for the first time. The primary goals of the Transit Tracker displays are to provide useful information to customers that was not previously available, thereby improving customer satisfaction with Tri-Met’s transit system and possibly increasing rider security.
The City of Portland and ODOT are implementing cooperative strategies to test the deployment of ITS on a parallel freeway/arterial corridor. As part of the regional advanced traffic management system (ATMS) program, traffic and incident management along the regional freeway and arterial systems are planned with freeway/arterial integration. Specifically, this project will accelerate the current deployment and integration of traffic surveillance and control devices in a high volume freeway/arterial corridor.
This project will facilitate cooperative use of specific agency devices and control software (arterial signal timing, freeway ramp meter signal operation, electronic message signs, CCTV) to integrate these capabilities and manage shared resources in the high-density, parallel I-5/Barbur Boulevard freeway/arterial corridor (Figure 1‑3). This integration will allow for multi-agency traffic-responsive corridor management that includes emergency and transit priority treatment.
A map of the study area
is shown in Figure 1‑4
(designated by the box). The study area
lies south of downtown Portland and begins generally from the point where I-5
crosses into Multnomah County (near the S.W. Barbur Boulevard Exit #294, which
is where Barbur Boulevard begins to parallel I-5) and extends approximately 3.5
miles north toward the downtown Portland area.
Figure 1‑3. CCTV View of I-5 at Barbur Blvd.
Project partners expect the benefits of the I-5/Barbur Boulevard parallel corridor traffic management demonstration project to include:
· Sustained or increased corridor capacity/throughput during incident conditions,
· Reduction of on-ramp incidents using ramp metering,
· Improved safety and efficiency of urban corridors,
· Improved integration of regional freeway systems with local signal systems,
· Improved incident detection and notification to reduce incident response time, and
· Provision of freeway and arterial corridor status to system operators.
Figure 1‑4. Map of I-5/Barbur Boulevard Study Area
The California Oregon Advanced Transportation System (COATS) is a project that seeks to encourage regional, public- and private-sector cooperation between California and Oregon to better facilitate the planning and implementation of ITS in the bi-state area. The COATS study area includes 13 counties in northern California and the southern half of Oregon and is not defined by county lines but rather by roadway segments. The intent of this project is to facilitate the use of ITS to enhance safety, improve the movement of people, goods, and services, and subsequently promote the economic development of the region.(4)
The COATS bi-state area contains transportation links vital to the region’s economy and commercial industry. Numerous primary and secondary routes serve commercial vehicles destined for urban centers throughout the West. Unpredictable weather patterns and mountainous topography add to the transportation challenges. Travelers throughout the corridor must contend with diverse and rapidly-changing weather conditions including snow, high winds, fog, and heavy rain. The combination of varied driving conditions and abundant off-road, commercial, and recreational traffic produces an immediate and expanding need for increased traffic safety measures and information dissemination techniques.(4)
In the short-term, there are three main strategies of the COATS project: (4)
· Address operational efficiency and public safety (monitor road-weather conditions with road-weather information, wind monitoring stations, automated flood warning systems, automated visibility systems, etc. and monitor roadway rights-of-way for potential animal-vehicle conflicts or for detecting landslides);
· Advise unfamiliar travelers of unsafe driving conditions through advance warning systems, variable message signs (VMS), and highway advisory radio (speed/travel conditions, wide loads on narrow lanes, etc); and
· Provide for the development of centers to coordinate, communicate, and cooperate with each other, nearby communities, local organizations, State agencies, and other regions (Redding and Eureka, CA and Salem, Bend, and Medford, OR).
Project partners expect the benefits of the COATS bi-state rural project to include:
· Improved traffic and roadway status information dissemination and access to avoid stranding drivers in remote locations due to unexpected road closures, restrictions, and adverse weather conditions;
· Safer rural travel that is also more efficient and convenient; and
· Improved coordination among the agencies involved in managing bi-state corridors through the provision of more complete real-time information to operations and maintenance personnel.
Portland’s regional traffic management centers require a complete status for both freeway and arterial roadways to effectively perform traffic control and incident management and to provide traffic information to the public. The regional freeway system is suitably instrumented for this purpose, but extended arterial network surveillance is cost prohibitive. Thus, transit probe data may afford a viable opportunity to provide arterial status information in this region. (1)
Tri-Met and C-TRAN operate the two transit systems serving
the four-county area in Oregon and Washington (Figure 1‑5).
These transit properties operate an extensive fixed-route schedule in
the region, and Tri-Met’s buses are suitably equipped to collect schedule
adherence data and serve as traffic probes.
Tri-Met currently collects and post-processes bus operation data,
including schedule adherence, for consideration in fixed-route scheduling. Additionally, Tri-Met is beginning to share
these data with local 
jurisdictions for
operational use. (1)
This regional transit-traffic management integration project addresses the technically challenging integration and utilization of real-time transit data for the purpose of establishing arterial (and freeway) network status. This project will support the use of travel time data for real-time management of traffic signals as well as analysis of corridor performance. (1)
Figure 1‑5. Tri-Met Bus Dispatch Station
Tri-Met buses traveling along a number of key corridors will report speeds or travel times on selected street segments determined to be of interest to the City of Portland and ODOT Region 1. The real-time collection of speeds on several river crossings are of particular interest, as these tend to be congestion points in the system, and buses are traveling with the mixed-flow traffic on these bridges (i.e., there are no transit stops).
The information collected from the buses will be color-coded and added to the ODOT network status maps that are available to the public on ODOT’s website. These data will allow the agencies to better monitor and manage the transportation system and will fill the gaps in network management. Although the long-term plan includes giving speed and travel time information collected by the AVL technology on the buses directly to the motorists, more variable message signs are needed before this can be fully realized.
Project partners expect the benefits of the Transit Buses as Traffic Probes project to include:
· Better information for which to make traffic and incident management decisions,
· More accurate information disseminated to public,
· More efficient operation of the freeway and arterial transportation network, and
· Increased traveler mobility.
The purpose of the Transit Tracker evaluation is to collect and analyze data related to a change in bus riders’ behaviors and perceptions as a result of the Transit Tracker information displays. Four measures of effectiveness were selected to test the impact of the information on riders’ behaviors and perceptions:
· Riders’ use of pre-trip and en-route planning information,
· Riders’ perceptions of system efficiency,
· Riders’ perceptions of personal security, and
· Riders’ overall satisfaction with the system.
In order to test the information’s impact on riders’ behaviors and perceptions, an understanding of baseline behaviors and perceptions is required. This baseline information will be analyzed to provide a basis for comparison with any data collected during the post-deployment period.
In addition to the baseline surveys conducted for the national evaluation, Tri-Met conducted surveys of over 800 transit riders prior to the installation of any Transit Tracker displays. Results of both surveys are presented in this section of the report.
The Transit Tracker evaluation concentrates on the behaviors and perceptions of Tri-Met bus riders in Portland. These behaviors and perceptions are being evaluated primarily through intercept surveys. These intercept surveys were conducted at stops where Transit Tracker had not yet been installed, but where plans existed for near-term installation. These data will be compared to data collected at the same locations after Transit Tracker information displays have been installed to determine if there are differences in riders’ behaviors and perceptions.
The approach was to use one survey instrument to obtain information for each of the four goals of the Transit Tracker evaluation: (1) assess riders’ use of pre-trip and en-route planning information, (2) assess riders’ perceptions of system efficiency, (3) assess riders’ perceptions of personal security, and (4) assess riders’ overall satisfaction with they system. Surveys were administered at four Tri-Met bus stops in Portland, the locations of which were suggested by Tri-Met. Riders were approached and told that surveyors were doing a customer satisfaction survey of the Tri-Met bus service. They were asked if they would mind answering a few questions while they waited for their bus to arrive. Riders who agreed to participate were given a copy of the survey to follow along as the surveyor read each question aloud. Surveyors recorded each rider’s responses on a separate survey form. A copy of the baseline Transit Tracker survey can be found in Appendix A.
One of the goals of the Transit Tracker evaluation is to assess riders’ use of pre-trip and en-route planning information. The hypothesis is that Transit Tracker will provide riders with useful information with which they can make informed decisions about their trips in real time. For example, currently riders have access to only fixed-schedule information (from paper brochures, schedules posted at bus stops, on-line Internet schedules, etc.). While frequent users may have many of their schedules and routes memorized, newer riders may not be aware of when or how often the buses run. Transit Tracker information displays will allow riders to see not only the number of minutes until the next bus arrives, but also the number of minutes until the next two or three buses arrive. With this type of information, riders may decide to take a different route or may decide to run an errand while waiting, instead of waiting at the stop. This information could be especially useful in inclement weather conditions (which are common in Portland) when riders may not want to wait outside too long for the bus to arrive. This type of information can afford the transit rider with more opportunities to make alternative route or travel decisions, as well as opportunities to do other things while they wait for their bus. Finally, Transit Tracker also offers the intangible benefit of reducing anxiety and stress associated with waiting for a bus that is late, therefore improving customer satisfaction.
Several questions on the survey were formulated to determine if and how often riders make use of the current fixed-schedule information that is available to them, as well as their perceptions of the accuracy of such information. In the Phase III post-deployment surveys, questions will be geared toward users’ perceptions of how Transit Tracker has changed their strategy for catching the bus as well as their perceptions of the accuracy and usefulness of the information provided by the Transit Tracker displays. In addition, riders will be directly asked if they used the information to make a decision related to their trip.
The real-time bus arrival information displayed on the Transit Trackers is also available on the Tri-Met website. The information became available for select routes and locations in March 2002, and Tri-Met is planning to expand it to all bus stops. Thus, a website usage survey will also be conducted to determine how many hits the website receives before and after transit tracker, how long users are on the website, and what type of information they obtain. Riders will also be asked in the follow-up intercept interviews if and how often they access the website to obtain real-time bus schedule information for pre-trip planning information.
Another goal of the Transit Tracker evaluation is to assess riders’ perceptions of system efficiency. It is hypothesized that riders will perceive an increase in efficiency, in terms of on-time performance, even if there is no change in the actual system performance. This is due to the fact that the Transit Tracker signs will provide riders with real-time bus arrival information. In other words, even though the bus may be operating behind schedule, the Transit Tracker displays will show the actual (versus scheduled) time of arrival. If the system is accurate, the bus will arrive when the counter on the display nears zero and the sign displays the word “Due.” In the minds of riders, arrival of the bus when the displays say it will arrive, whether or not it is at the scheduled arrival time, may indicate and may constitute an improvement in system efficiency.
Several questions on the survey were formulated to measure users’ perceptions of the efficiency of the transit system in terms of how long they typically wait for the bus and if it is usually on time. The responses to these questions will be compared to similar questions asked in post-deployment surveys at the same bus stops.
Another goal of the Transit Tracker evaluation is to assess transit riders’ perceptions of personal security. It is hypothesized that riders will perceive an increase in personal security, even if there are no other measures taken to increase security (such as increased police presence). While Transit Tracker will likely have little impact on the actual security of transit riders, having access to more accurate arrival time information may afford them the opportunity to wait elsewhere for the bus, such as a coffee shop, if they have a long wait, if it is after dark, or in areas where they are not familiar or comfortable. In addition, having access to Transit Tracker information on the Internet may allow customers to wait longer before leaving home, shortening their wait times at stops. Having information available to make these types of decisions could have an impact on users’ perception of personal security.
Several questions on the survey were formulated to determine users’ current perceptions of personal security. The responses to these questions will be compared to similar questions asked in post-deployment surveys at the same bus stops.
In addition to surveys, ridership data is also being examined, with attention focused on ridership statistics after dark. Tri-Met has expressed their skepticism about relating increased ridership to the Transit Tracker information displays. Transit ridership in Portland has been increasing steadily over the past several years, and it would therefore be difficult to conclude that the Transit Tracker information contributed directly to an increase in ridership. However, if there is an increased sense of security related to the presence of the Transit Tracker information, riders may feel safer riding at night, when they normally may not be comfortable riding the bus. Therefore, nighttime ridership before and after Transit Tracker installation will be compared for the four stops.
Finally, several survey questions were formulated to determine users’ overall satisfaction with Tri-Met bus service. While much of a riders’ satisfaction, or lack thereof, with the bus system may have to do with his or her perceptions of accuracy, efficiency, and safety, there may be other factors that influence bus riders’ perception of service quality.
Findings are presented in two parts. First, the findings of the Tri-Met surveys administered in the Spring of 2000, prior to installation of the Transit Tracker displays, are summarized. Subsequently, the findings of the baseline surveys conducted by SAIC in January 2002 are presented.
Between April 24 and May 14, 2000, Tri-Met conducted intercept surveys with bus riders at two bus stops and one light rail stop in Portland.(5) A copy of the Tri-Met baseline intercept survey can be found in Appendix B. The purpose of the survey was to obtain the following baseline information prior to installing Transit Tracker displays:
· Overall satisfaction ratings with the transit service at each location;
· Perceived waiting time;
· Perceptions of on-time performance;
· Perceptions of safety when waiting for the bus or light rail;
· Transit use characteristics including riding frequency, time of day, day of the week, and level of transit dependence; and
· Passenger age and gender.
In all, 830 surveys were administered at three locations. The following is a summary of the key findings:
· 76 percent indicated that they were somewhat or very satisfied with the transit service.
· 72 percent were “heavy” transit uses (making more than 46 trips per month on average.
· 57 percent were transit dependent.
· Fewer than 22 percent knew the scheduled arrival time of the next bus or train before coming to the stop. Those who did know got their information from a printed schedule or a Tri-Met Guide. Of those who did not know, 45 percent (at one location) said they looked at the schedule posted at the stop, while only 8 percent at another location looked at the posted schedule.
· Riders reported waiting for the bus or light rail an average of 8.5, 11.6, and 13.5 minutes at the three stops where surveys were administered.
· 70 percent at two of the locations and 63 percent at the other reported the bus or train that they wait for is usually on time.
· Personal safety was generally not a concern for passengers at these locations. Overall, 97 percent of respondents indicated that they felt safe waiting for the bus or train.
The survey results indicate that the majority of the respondents were generally satisfied with Tri-Met service, perceive the service as usually being on time, and voice few concerns for their personal safety.
While these results indicate an overall satisfaction with the Tri-Met service, Tri-Met believes that the Transit Tracker project offers the opportunity to improve the perceptions of on-time performance and increase the proportion of riders who report being very satisfied by improving their waiting experience.
SAIC conducted additional baseline Transit Tracker intercept surveys in Portland on Tuesday afternoon, January 22, 2002 through Friday morning, January 25, 2002. The survey instrument used by the SAIC team was similar to that of the survey used by Tri-Met (discussed in the previous section); however, the surveys differed in the following three ways: some of Tri-Met’s questions were slightly re-worded, not all Tri-Met’s questions were included, and new questions were added to the SAIC survey that were not included on Tri-Met’s survey. A copy of SAIC’s Transit Tracker intercept survey can be found in Appendix A.
For survey administration, two teams of two surveyors went to different bus stops during the morning peak (7 – 9 a.m.), over the lunch hour (11:30 a.m. – 1 p.m.), and during the evening peak (4 – 6 p.m.). Table 2‑1 illustrates the mean weekday boardings from Spring 2001 by time of day at each bus stop surveyed. Table 2‑2 shows the number of completed surveys obtained by time of day at each bus stop. In all, 240 completed surveys were obtained.
Overall, bus riders were extremely receptive and cooperative. In fact, about 9 out of 10 people approached agreed to participate in the survey. The only difficulty experienced in survey administration was the ability to complete the survey before the bus arrived. In about 1 or 2 out of 10 riders surveyed, surveys were not completed before the bus arrived and therefore could not be used in the analyses. This was particularly problematic on rainy days at small shelters, as many riders arrived at the stops just before the bus arrived.
Table 2‑1. Ridership Data for Four Bus Stops in Portland
|
Bus
Stop |
Mean
Weekday Boardings by Time of Day (Spring 2001) |
|||
|
7 – 9 a.m. |
9 a.m. – 4
p.m. |
4 – 6 p.m. |
9 p.m. –
service end |
|
|
Barbur
Transit Center |
294 |
150 |
58 |
9 |
|
Weidler @ Lloyd Center |
12 |
163 |
81 |
9 |
|
Burnside and
28th |
70 |
158 |
44 |
16 |
|
Burnside and
Grand |
25 |
86 |
42 |
8 |
Table 2‑2. Number of Completed Baseline Customer Satisfaction Surveys
|
Bus Stop |
Number of Completed
Surveys by Time of Day (Jan. 2002) |
|||
|
7 – 9 a.m. |
9 a.m. – 4 p.m. |
4 – 6 p.m. |
Total |
|
|
Barbur
Transit Center |
109 |
0 |
0 |
109 |
|
Weidler
@ Lloyd Center |
0 |
18 |
35 |
53 |
|
Burnside
and 28th |
40 |
11 |
11 |
62 |
|
Burnside
and Grand |
0 |
9 |
7 |
16 |
At the time of administration of the SAIC baseline survey, Transit Tracker signs were not present at the four bus stops surveyed. However, there were Transit Tracker signs that had already been installed at four other bus stops elsewhere in Portland and on several of the light rail platforms on the MAX line to the airport. Thus, there was a possibility that riders surveyed as part of this evaluation had seen the signs at other locations, and that this awareness may have some impact on their survey responses.
To account for the previous installation of Transit Tracker signs at other locations, several steps were taken to avoid bias in the survey and to identify if bias existed. First, survey questions were worded to pertain to the stop at which the survey was being administered. For example, one question asked, “At this bus stop, how satisfied are you with bus adherence to the posted schedules?” By phrasing questions in this manner, riders were asked to focus on that bus stop when responding to questions, and not another stop that may have a Transit Tracker display.
In addition, so as to be able to test for bias in the survey responses, the last question on the survey inquired about whether riders had waited at one of the stops where Transit Tracker had been installed. The results showed that about 55 percent of those surveyed reported that they had seen a Transit Tracker sign at another location. Many respondents, however, reported that they had not really used the information on the sign (they had just seen it in passing), and only about 27 percent correctly identified a bus stop where a sign had been installed (others reported seeing a sign where one did not exist or had seen a sign only on the light rail system).
To determine if the signs that had already been installed had an affect on riders’ responses to survey questions, the responses of those who had seen a sign were compared to the responses of those who had not seen a sign. Comparisons were made for two of the important measures of effectiveness, perception of on-time performance and overall satisfaction with the system. The results showed no statistically significant differences in the responses of the two groups of riders. Therefore, it is assumed that the presence of the Transit Tracker signs at other locations did not impact the responses of the participants in the baseline Transit Tracker survey, and all survey responses are included in these analyses.
Of the 240 bus riders surveyed at the four stops, 43 percent were male, and 57 percent were female. The age distribution of the riders surveyed by SAIC is illustrated in Figure 2‑1 and is compared to the age distribution of 112 riders surveyed by Tri-Met in January 2001 (exactly one year earlier) at different locations. Figure 2‑2 shows a comparison of the SAIC survey conducted in January 2002 and the Tri-Met survey conducted in January 2001 to another survey conducted by Tri-Met in May 2000 in which 830 riders were surveyed (at the same locations as their 2001 survey). Age categories were aggregated for the 2002 and 2001 surveys to match the numbers reported in the 2000 survey.
Figure 2‑1 illustrates that every age category was represented in the SAIC survey, and that the age distribution of respondents was very similar to that of the 2001 Tri-Met survey. The SAIC survey did capture a greater percentage of riders in the 25 – 34 age category and fewer riders in the under 25 age category than the Tri-Met survey.
When comparing the 2002 and 2001 surveys to the 2000 Tri-Met survey (Figure 2‑2) that surveyed over 800 riders (using the 2000 Tri-Met survey’s age aggregation), age distributions are nearly identical for the three surveys.
Figure 2‑1. Age Distribution of Survey Respondents Compared to 2001 Tri-Met Survey
Figure 2‑2. Age Distribution of Survey Respondents Compared to 2000 and 2001 Tri-Met Surveys
Figure 2‑3 illustrates the reported frequency with which respondents ride the bus. Just over 70 percent of respondents reported that they ride the bus nearly every day. About 26 percent indicated that they ride the bus one to four days per week, and only about 3 percent reported that they ride the bus less than one day per week. It should be noted that those respondents who take the bus five days per week for work (but usually not on weekends) are represented in the nearly every day category.
Figure 2‑3. Frequency with Which Respondents Ride the Bus
Respondents were also asked to indicate for which trip purposes they most frequently ride the bus. Trip purposes included: work, school, shopping, recreation, other, and for most all trips. The distribution of responses as to the most frequent trip purpose are shown in Figure 2‑4.
Figure 2‑4 illustrates that just over half of respondents (51
percent) indicated that they ride the bus most frequently for work trips. About 12 percent of respondents reported that they ride the bus
most frequently for school trips, and
another 12 percent indicated that they ride the bus for most all their trips. Five
percent or less of the respondents indicated that they ride the bus most frequently
for each of the other trip purposes. (It should be noted that nearly half of
the completed surveys were obtained from the Barbur Boulevard Transit Center,
which is a major transit hub for downtown workers who drive to the
park-and-ride and ride the bus to work).
Figure 2‑4. Distribution
of Most Frequent Trip Purposes
When asked whether or not they had an automobile available for their use, 62 percent of the respondents indicated that they do have an automobile available, and 38 percent indicated that they do not have an automobile available to them. In other words, about one-third of the bus riders surveyed are transit dependent.
To determine if bus ridership in Portland is related to automobile ownership, an analysis of frequency of bus use versus automobile ownership was performed. Table 2‑3 shows the number of auto and non-auto owners by frequency of bus use. A chi-squared test was performed on the distribution of responses to determine if frequency of bus use is related to automobile ownership. Chi-squared tests test for the independence of two variables. The calculated chi-squared value is compared to chi-squared table values. If the calculated value is higher than the table value, the two variables are not independent of one another.
The calculated chi-squared value for this test was 5.17. For one degree of freedom, the chi-squared values at the 5 percent and 1 percent level of significance are 6.64 and 3.84, respectively. Because 5.17 is greater than 3.84, but less than 6.64, the test is significant at the 5 percent level. In other words, frequency of transit use is not independent of auto ownership. From the data, it can be seen that non-auto owners tend to take transit more frequently than auto owners.
Table 2‑3. Automobile Ownership Versus Frequency of Bus Use
|
Auto Ownership |
Frequency of Bus Use |
||
|
Every day |
Less than 5 days per week |
Total |
|
|
Auto
Owner |
96 |
51 |
147 |
|
Non-auto
Owner |
72 |
19 |
91 |
|
Total |
168 |
70 |
238 |
As previously discussed, one of the goals of the Transit Tracker evaluation is to assess riders’ use of trip planning information. The results of how riders use fixed-schedule information (and it’s perceived usefulness) will be compared to how riders use the real-time information provided by Transit Tracker in Phase III of the evaluation. Several questions on the survey probed respondents about the frequency with which they use different types of fixed-schedule information, as well as their perception of the accuracy of the information. Schedule information included: printed brochure schedules, schedules posted at bus stops, the on-line Internet schedule, and the 238-RIDE phone number. The frequency with which riders reported using these types of schedules is presented in Figure 2‑5.
Figure 2‑5. Frequency With Which Respondents Use Bus Schedule Information
More respondents indicated that they rarely or almost never use the schedule information than those who indicated that they sometimes use the information or than those who indicated that they frequently or almost always use the information. Schedules posted at bus stops was rated by more respondents (34 percent) as being frequently or almost always used than any other type of information, while 238-RIDE was rated by the fewest respondents (7 percent) as being frequently or almost always used.
Considering these results, one might wonder how transit riders obtain information about their trips, if few report that they frequently or almost always use these types of schedule information. To try and get a better idea of how riders plan for and schedule their transit trips, two additional questions were asked with regard to riders’ trip-planning behaviors. The questions are shown in Table 2‑4. These questions considered two reasons why riders may not need schedule information: because they just go to the stop and wait and because they have their routes and times memorized. The results are shown in Figure 2‑6.
The responses to these questions offer some insight as to why many respondents rarely or almost never use the available bus schedule information. Forty-five percent of respondents indicated that they frequently or almost always just go to the bus stop and wait for the next bus to arrive (not knowing the scheduled arrival time), and about 19 percent indicated that they do so sometimes. Over half of respondents indicated that they do not use schedule information because they frequently or almost always have their routes and times memorized.
Table 2‑4. Survey Questions About Riders’ Trip
Planning Behavior
|
Please rate HOW OFTEN
the following statements are TRUE: |
|
I generally do not use the Tri-Met
schedule information—I just go to the bus stop and wait for the next bus to
arrive. Almost Always Frequently
Sometimes Rarely Almost Never 1 2 3 4 5
o--------------------o--------------------o------------------o------------------o |
|
I generally do not use the Tri-Met
schedule information, because I have most of my times/routes memorized. Almost Always Frequently
Sometimes Rarely Almost Never 1 2 3 4 5
o--------------------o--------------------o------------------o------------------o |
Figure 2‑6. Alternative
Trip Planning Behaviors
Respondents who reported that they rarely or almost never just go to the bus stop and wait or who reported that they rarely or almost never have their routes and times memorized were compared to the overall average response to the frequency of use of trip-planning information. This comparison was made to determine if these riders do in fact use information more often. Table 2‑5 shows the percentages of respondents in each category that frequently or almost always use schedule information.
Considering the printed brochures, overall only 30 percent of respondents reported that they frequently or almost always use them when planning their transit trips. However, 50 percent of riders who reported that they rarely/almost never just go to the stop and wait reported that they frequently or almost always use the printed brochures. Considering the schedules posted at stops, overall only 34 percent of respondents reported that they frequently or almost always use them when planning their transit trips. However, 56 percent of riders who reported that they rarely/almost never have their routes/times memorized reported that they frequently or almost always use the guides posted at stops. These results are similar for on-line schedules and the 238-RIDE phone number. These results show that different riders have different needs when it comes to scheduling their transit trips; some riders tend to have their times/routes memorized, others tend to just go to the stop and wait, and others tend to rely on the available schedule information.
Table 2‑5. Comparison of Use of Trip-Planning Information
|
Information Type |
Percent Who Frequently or Almost Always Use Schedule
Information |
||
|
Overall Average |
Rarely/almost never Just
go to stop and wait |
Rarely/almost never have
times/routes memorized |
|
|
Printed brochures |
30% |
50% |
42% |
|
Posted at stop |
34% |
41% |
56% |
|
On-line |
16% |
24% |
10% |
|
238-Ride |
7% |
10% |
15% |
One possible reason that riders may not use schedule information could be because they feel that the information is inaccurate. One question on the survey inquired about respondents’ perception of the accuracy of the schedule information they use. The results are shown in Figure 2‑7. The results indicate that, in fact, nearly 70 percent of respondents indicated that the schedule information that they use is frequently or almost always accurate, while only about 7 percent indicated that the information is only rarely or almost never accurate. In other words, inaccuracy of schedule information is not the reason most riders’ report not using the information.
Other possible reasons for not using schedule information, while speculative, could be the lack of knowledge that the information exists or the lack of usefulness or availability (in the case of the Internet) of the information.
Figure 2‑7. Respondents’ Perceptions of Accuracy of Schedule Information
These analyses were taken one step further to try and determine if there was a group of riders that uses schedule information more frequently than others. Two categories were considered in these analyses: auto ownership and age. It was hypothesized that auto ownership may affect the use of schedule information. For example, it is likely that people who own automobiles take transit for only very specific trips (such as going to work or school). While they may need to use schedule information once to determine which bus they will take to get to work/school on time, they probably will not need to refer to the information again unless they change jobs, change work schedules, or the bus schedules change. It was shown previously that people who do not own automobiles are more reliant on transit for most or all their trips (i.e., take transit every day). It is less likely then that they would have all their routes and times memorized, as it is much more difficult to remember bus times for all trips than for two trips per day (to and from work). They would therefore need to refer to schedule information more often.
As for age, it was hypothesized that age may have an effect on riders’ use of schedule information, as riders of different ages may have different needs when it comes to riding the bus (e.g., older riders may need to know exact times and know of more time/route options than younger riders) and/or may like to obtain their schedule information from different sources (e.g., Internet versus telephone).
Chi-squared (Χ2) tests were used to determine if there is a relationship between use of schedule information and auto ownership and age. The results are shown in Table 2‑6 and Table 2‑7, respectively. The relationship between age and use of schedule information was found to be independent for all four types of schedule information. In other words, age does not affect the use of schedule information, contrary to what was hypothesized. However, it was found that auto ownership and use of schedule information are not independent for three of the four types of schedule information (shown in bold in Table 2‑6). The use of guides posted at bus stops is dependent on auto ownership (at the 1% significance level); the use of on-line Internet schedules is dependent on auto ownership (at the 5% significance level); and the use of the 238-RIDE phone number is dependent on auto ownership (at the 1% significance level). The use of printed brochure schedules, however, was found to be independent of auto ownership.
To help further explain these results, the distribution of responses to use of schedule information versus auto ownership is shown in Figure 2‑8. For guides posted at bus stops, the distribution of responses of the frequency with which auto owners use the information is nearly the inverse of the distribution of the frequency with which non-auto owners use the information. About half of non-auto owners reported using the guides posted at the bus stops frequently or almost always, while only 26 percent of auto owners reported frequently or almost always doing so. Likewise, about half of auto owners reported rarely or almost never using the guides posted at the bus stops, while only about 27 percent of non-auto owners reported rarely or almost never using the guides. Therefore, the hypothesis was correct, non-auto owners tend to use guides posted at stops more frequently than auto owners.
Table 2‑6. Χ2 Results for Auto Ownership Vs. Use of Schedule Information
|
Information Type |
Degrees of Freedom |
Χ 2 Calculated |
Χ 2 5% sig. level |
Χ 2 1% sig. level |
|
Printed brochures |
4 |
2.64 |
9.49 |
13.29 |
|
Guides posted at stop |
4 |
22.38 |
9.49 |
13.28 |
|
On-line |
4 |
11.10 |
9.49 |
13.29 |
|
238-Ride |
3 |
14.26 |
7.82 |
11.34 |
Table 2‑7. Χ2 Results for Age Vs. Use of Schedule Information
|
Information Type |
Degrees of Freedom |
Χ 2 Calculated |
Χ 2 5% sig. level |
Χ 2 1% sig. level |
|
Printed brochures |
12 |
6.88 |
16.92 |
21.67 |
|
Guides posted at stop |
9 |
15.31 |
21.03 |
26.22 |
|
On-line |
9 |
12.42 |
16.92 |
21.67 |
|
238-Ride |
6 |
9.22 |
12.59 |
16.81 |
While the most frequent response to the use of Internet schedules was almost never, many more non-auto owners reported that they almost never use the Internet than auto owners, and over twice as many auto owners than non-auto owners reported that they sometimes use the Internet to obtain information. This result is the opposite of what was hypothesized. This could be because people who own automobiles use transit for specific trips such as work and therefore may have better access to the Internet before making their transit trips (for instance, before their trip home from the office). Thus, it may be a question of accessibility to this type of information that affects its frequency of use.
For the 238-RIDE number, more auto owners reported that they almost never use the information than non-auto owners. Likewise, more non-auto owners reported that they almost always, frequently, or sometimes use the phone number than auto owners. This could be a question of the need for information and the accessibility of the information. While the 238-RIDE number does not seem like the preferred source of information, it may be more accessible to non-auto owners than the on-line Internet schedules.
When asked if the bus is usually on time at the stop, about 73 percent of respondents said yes, that the bus is usually on time. Only 10 percent said no, that the bus is not usually on time, and about 16 percent reported that they did not know if the bus is usually on time (either because they had never been to the stop before, or because they did not know the scheduled arrival time). This same analysis was performed separately for frequent (every day) versus less frequent (less than 5 days per week) riders. These results showed that slightly more less-frequent riders reported that the bus is usually on time (91 percent versus 86 percent overall), maybe because they are less aware of the schedule than those who ride every day. In addition, more less-frequent riders reported that they did not know if the bus is usually on time (24 percent versus 13 percent overall).
Figure 2‑8. Comparison of Auto Ownership and Use of Schedule Information
When asked how long they usually wait for the bus at the stop, 26 percent of the respondents gave a range (e.g., 5 – 10 minutes), while the remaining 74 percent reported an integer value. The distribution of responses for those reporting an integer is shown in Figure 2‑9.
