82nd Annual Meeting Transportation Research Board January 13, 2003 Robert L. Bertini Department of Civil & Environmental Engineering Ahmed El-Geneidy School of Urban Studies and Planning Portland State University Using Archived Data to Generate Transit Performance Measures
29
Embed
82nd Annual Meeting Transportation Research Board January 13, 2003 82nd Annual Meeting Transportation Research Board January 13, 2003 Robert L. Bertini.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
82nd Annual MeetingTransportation Research BoardJanuary 13, 2003
Robert L. BertiniDepartment of Civil & Environmental Engineering
Ahmed El-GeneidySchool of Urban Studies and Planning
Problem StatementProblem StatementProblem StatementProblem Statement
• Importance of transit serviceImportance of transit service• New ITS monitoring and management systemsNew ITS monitoring and management systems• Performance monitoring—real time & in retrospectPerformance monitoring—real time & in retrospect• PastPast
– Limited scope and durationLimited scope and duration– Aggregate measuresAggregate measures– Costly data collectionCostly data collection
• NowNow– Unlimited coverage and continuous durationUnlimited coverage and continuous duration– Design, extract and test specific measuresDesign, extract and test specific measures– Actual system performanceActual system performance– Data management/processing challengesData management/processing challenges– Need for generating relevant measuresNeed for generating relevant measures
• Importance of transit serviceImportance of transit service• New ITS monitoring and management systemsNew ITS monitoring and management systems• Performance monitoring—real time & in retrospectPerformance monitoring—real time & in retrospect• PastPast
– Limited scope and durationLimited scope and duration– Aggregate measuresAggregate measures– Costly data collectionCostly data collection
• NowNow– Unlimited coverage and continuous durationUnlimited coverage and continuous duration– Design, extract and test specific measuresDesign, extract and test specific measures– Actual system performanceActual system performance– Data management/processing challengesData management/processing challenges– Need for generating relevant measuresNeed for generating relevant measures
3
Let Knowledge Serve the City
ObjectivesObjectivesObjectivesObjectives
• Describe how archived dispatch system database can be Describe how archived dispatch system database can be
used to generate performance measures.used to generate performance measures.
• Improve service standards and effectiveness.Improve service standards and effectiveness.
• Begin process for developing, testing, using and Begin process for developing, testing, using and
incorporating performance measures into daily incorporating performance measures into daily
operations.operations.
• Focus on experimental set (pilot) of measures.Focus on experimental set (pilot) of measures.• Part of larger transit operations research program under Part of larger transit operations research program under
Great Cities’ Universities Coalition and partially funded Great Cities’ Universities Coalition and partially funded
by Trimet.by Trimet.
• Describe how archived dispatch system database can be Describe how archived dispatch system database can be
used to generate performance measures.used to generate performance measures.
• Improve service standards and effectiveness.Improve service standards and effectiveness.
• Begin process for developing, testing, using and Begin process for developing, testing, using and
incorporating performance measures into daily incorporating performance measures into daily
operations.operations.
• Focus on experimental set (pilot) of measures.Focus on experimental set (pilot) of measures.• Part of larger transit operations research program under Part of larger transit operations research program under
Great Cities’ Universities Coalition and partially funded Great Cities’ Universities Coalition and partially funded
by Trimet.by Trimet.
4
Let Knowledge Serve the City
FrameworkFrameworkFrameworkFramework
Cost E
ffective
ness
Cost E
ffective
nessC
ost
Effi
ciency
Cost
Effi
ciency
Service EffectivenessService Effectiveness
Service InputsService InputsLabor, Capital, FuelLabor, Capital, Fuel
Service ConsumptionService ConsumptionPax, Pax-Miles, RevenuePax, Pax-Miles, Revenue
Service OutputsService OutputsVeh-Hrs, Veh-MilesVeh-Hrs, Veh-Miles
• Measuring system performance is the first step toward Measuring system performance is the first step toward
efficient and proactive management.efficient and proactive management.• Increasing attention to transit performanceIncreasing attention to transit performance
– Transit Capacity and Quality of Service ManualTransit Capacity and Quality of Service Manual» Quantitative/qualitativeQuantitative/qualitative» Passenger point of viewPassenger point of view» Linked to agency operating decisionsLinked to agency operating decisions
– NCHRP Performance Based Planning ManualNCHRP Performance Based Planning Manual» AccessibilityAccessibility» MobilityMobility» Economic DevelopmentEconomic Development
• Measuring system performance is the first step toward Measuring system performance is the first step toward
efficient and proactive management.efficient and proactive management.• Increasing attention to transit performanceIncreasing attention to transit performance
– Transit Capacity and Quality of Service ManualTransit Capacity and Quality of Service Manual» Quantitative/qualitativeQuantitative/qualitative» Passenger point of viewPassenger point of view» Linked to agency operating decisionsLinked to agency operating decisions
– NCHRP Performance Based Planning ManualNCHRP Performance Based Planning Manual» AccessibilityAccessibility» MobilityMobility» Economic DevelopmentEconomic Development
• 62 million annual bus trips62 million annual bus trips• 600 square miles600 square miles• 1.2 million population1.2 million population• 700 vehicles700 vehicles• 98 routes98 routes• 9,000 bus stops9,000 bus stops
• 62 million annual bus trips62 million annual bus trips• 600 square miles600 square miles• 1.2 million population1.2 million population• 700 vehicles700 vehicles• 98 routes98 routes• 9,000 bus stops9,000 bus stops
9
Let Knowledge Serve the City
TriMet Bus Dispatch SystemTriMet Bus Dispatch SystemTriMet Bus Dispatch SystemTriMet Bus Dispatch System
• Bus Dispatch System (BDS) tracks bus location and Bus Dispatch System (BDS) tracks bus location and schedule adherence.schedule adherence.
• Automatic vehicle location (AVL) using global Automatic vehicle location (AVL) using global positioning system (GPS).positioning system (GPS).
• Automatic passenger counters (APCs) on most Automatic passenger counters (APCs) on most vehicles.vehicles.
• Bus Dispatch System (BDS) tracks bus location and Bus Dispatch System (BDS) tracks bus location and schedule adherence.schedule adherence.
• Automatic vehicle location (AVL) using global Automatic vehicle location (AVL) using global positioning system (GPS).positioning system (GPS).
• Automatic passenger counters (APCs) on most Automatic passenger counters (APCs) on most vehicles.vehicles.
Smart Bus Concept
10
Let Knowledge Serve the City
TriMet Bus Dispatch SystemTriMet Bus Dispatch SystemTriMet Bus Dispatch SystemTriMet Bus Dispatch System
• Real time operating informationReal time operating information• Stop level data archived on vehicle, available for later Stop level data archived on vehicle, available for later
analysis on system-wide basisanalysis on system-wide basis• Each stop geo-codedEach stop geo-coded• New data added for each stopNew data added for each stop
– Scheduled arrival time (important meta data)Scheduled arrival time (important meta data)– Actual Arrive/door open timeActual Arrive/door open time– Number of boardings and alightingsNumber of boardings and alightings– Depart/door close timeDepart/door close time– Lift useLift use
• Schedule adherence reported to operator/dispatcherSchedule adherence reported to operator/dispatcher
• Real time operating informationReal time operating information• Stop level data archived on vehicle, available for later Stop level data archived on vehicle, available for later
analysis on system-wide basisanalysis on system-wide basis• Each stop geo-codedEach stop geo-coded• New data added for each stopNew data added for each stop
– Scheduled arrival time (important meta data)Scheduled arrival time (important meta data)– Actual Arrive/door open timeActual Arrive/door open time– Number of boardings and alightingsNumber of boardings and alightings– Depart/door close timeDepart/door close time– Lift useLift use
• Schedule adherence reported to operator/dispatcherSchedule adherence reported to operator/dispatcher
System Level TPMsSystem Level TPMsSystem Level TPMsSystem Level TPMs
• System level TPMs can include all data System level TPMs can include all data procesed for external reporting:procesed for external reporting:– RidershipRidership– BoardingsBoardings– RevenueRevenue– Expenditures of the overall system. Expenditures of the overall system.
• Route level measures can be aggregated Route level measures can be aggregated over the entire transit network. over the entire transit network.
• System level TPMs can include all data System level TPMs can include all data procesed for external reporting:procesed for external reporting:– RidershipRidership– BoardingsBoardings– RevenueRevenue– Expenditures of the overall system. Expenditures of the overall system.
• Route level measures can be aggregated Route level measures can be aggregated over the entire transit network. over the entire transit network.
• Time distribution between trip time and Time distribution between trip time and layover timelayover time
• Route 12 during one weekday of service Route 12 during one weekday of service (January 24, 2002). (January 24, 2002).
• At the route level, using the archived BDS At the route level, using the archived BDS data, it is possible to create a daily report for data, it is possible to create a daily report for each route. each route.
• Need to control layover time (non-revenue)Need to control layover time (non-revenue)• One day 9% of time at layoversOne day 9% of time at layovers
• Time distribution between trip time and Time distribution between trip time and layover timelayover time
• Route 12 during one weekday of service Route 12 during one weekday of service (January 24, 2002). (January 24, 2002).
• At the route level, using the archived BDS At the route level, using the archived BDS data, it is possible to create a daily report for data, it is possible to create a daily report for each route. each route.
• Need to control layover time (non-revenue)Need to control layover time (non-revenue)• One day 9% of time at layoversOne day 9% of time at layovers
• Daily report for Route 14Daily report for Route 14• Actual/scheduled hours of serviceActual/scheduled hours of service• Actual/scheduled tripsActual/scheduled trips• Actual/scheduled milesActual/scheduled miles• Actual/scheduled layoverActual/scheduled layover• Passengers carriedPassengers carried• Boardings/alightingsBoardings/alightings• Dwell time analysisDwell time analysis• DelayDelay• Average passenger loadAverage passenger load• Passengers per milePassengers per mile• Scheduled/actual speedScheduled/actual speed• Number of operatorsNumber of operators
• Inbound/outboundInbound/outbound• Peak/offpeakPeak/offpeak• Study longitudinally over many days/yearsStudy longitudinally over many days/years
• Daily report for Route 14Daily report for Route 14• Actual/scheduled hours of serviceActual/scheduled hours of service• Actual/scheduled tripsActual/scheduled trips• Actual/scheduled milesActual/scheduled miles• Actual/scheduled layoverActual/scheduled layover• Passengers carriedPassengers carried• Boardings/alightingsBoardings/alightings• Dwell time analysisDwell time analysis• DelayDelay• Average passenger loadAverage passenger load• Passengers per milePassengers per mile• Scheduled/actual speedScheduled/actual speed• Number of operatorsNumber of operators
• Inbound/outboundInbound/outbound• Peak/offpeakPeak/offpeak• Study longitudinally over many days/yearsStudy longitudinally over many days/years
702446Number of operators
11.111.310.9Average speed mile/hour
11.712.011.3Average scheduled speed mile/hour
14.810.38.9Number of passengers per mile
116.780.173.2Average passenger load during the trip
15,8758,3317,544Total boarding and alighting
7,9374,1653,772Number of passengers carried
1,650.3806.8843.6Number of actual miles operated
1,677.2862.4814.8Number of scheduled miles
207104103Number of actual trips
212107105Number of scheduled trips
0131440557101872Scheduled hours of service
2853148401571483777Hours of service
Seconds
Minutes
Hours
Seconds
Minutes
Hours
Seconds
Minutes
Hours
TotalOutboundInbound
702446Number of operators
11.111.310.9Average speed mile/hour
11.712.011.3Average scheduled speed mile/hour
14.810.38.9Number of passengers per mile
116.780.173.2Average passenger load during the trip
• Transit Availability—key measure of quality of serviceTransit Availability—key measure of quality of service• One sample census tractOne sample census tract
– 1.5 square miles1.5 square miles– 7,900 population (2000)7,900 population (2000)– 0.25-mile buffer around each bus stop0.25-mile buffer around each bus stop– 38% of area within walking distance38% of area within walking distance
• Transit Availability—key measure of quality of serviceTransit Availability—key measure of quality of service• One sample census tractOne sample census tract
– 1.5 square miles1.5 square miles– 7,900 population (2000)7,900 population (2000)– 0.25-mile buffer around each bus stop0.25-mile buffer around each bus stop– 38% of area within walking distance38% of area within walking distance
• Important for passenger Important for passenger attractiveness and attractiveness and operating efficiencyoperating efficiency
• Observe how speed Observe how speed varies with time and varies with time and spacespace
• Example using Example using instantaneous instantaneous speed/location for speed/location for express bus on freeway express bus on freeway corridor (highlights corridor (highlights bottleneck)bottleneck)
Transit Operating Transit Operating SpeedSpeed
• Important for passenger Important for passenger attractiveness and attractiveness and operating efficiencyoperating efficiency
• Observe how speed Observe how speed varies with time and varies with time and spacespace
• Example using Example using instantaneous instantaneous speed/location for speed/location for express bus on freeway express bus on freeway corridor (highlights corridor (highlights bottleneck)bottleneck)
• Shift from relying on few, general, aggregate measures Shift from relying on few, general, aggregate measures to detailed, specific measures.to detailed, specific measures.
• Challenges in data collection deployment and archiving—Challenges in data collection deployment and archiving—demonstration of value.demonstration of value.
• Difficulties in converting large quantities of data into Difficulties in converting large quantities of data into meaningful, useful information.meaningful, useful information.
• Connections to service standards.Connections to service standards.• Importance of performance measurement for planning, Importance of performance measurement for planning,
system design/modification and operations.system design/modification and operations.• Support development of TCQSM.Support development of TCQSM.• Experiment with new TPMs and track them over time.Experiment with new TPMs and track them over time.• Introduce into daily operations environment.Introduce into daily operations environment.
• Shift from relying on few, general, aggregate measures Shift from relying on few, general, aggregate measures to detailed, specific measures.to detailed, specific measures.
• Challenges in data collection deployment and archiving—Challenges in data collection deployment and archiving—demonstration of value.demonstration of value.
• Difficulties in converting large quantities of data into Difficulties in converting large quantities of data into meaningful, useful information.meaningful, useful information.
• Connections to service standards.Connections to service standards.• Importance of performance measurement for planning, Importance of performance measurement for planning,
system design/modification and operations.system design/modification and operations.• Support development of TCQSM.Support development of TCQSM.• Experiment with new TPMs and track them over time.Experiment with new TPMs and track them over time.• Introduce into daily operations environment.Introduce into daily operations environment.