Mercer SCOOT Adaptive Signal Control Karl Typolt, Transpo Group PSRC RTOC July 6th, 2017
Our mission, vision, and core values Mission: deliver a high-quality transportation system for Seattle Vision: connected people, places, and products Committed to 5 core values to create a city that is: Safe Interconnected Affordable Vibrant Innovative
Presentation Overview What is SCOOT adaptive signal control? Project background Project phases and more future SCOOT corridors SCOOT User Interface Performance measures 3
What is SCOOT Adaptive Signal Control? Adjusts signal timing in real-time to match traffic patterns. Combines data from multiple sources. SCOOT is the algorithm we have selected as recommended by System Engineering report. 4
Traffic Volume and Cycle Length Comparison 5
Project Background In 2015, initiated planning process including systems engineering evaluation. Selected adaptive system in accordance with compatibility to existing signal management platform. Developed a phased implementation plan that packages adaptive deployment into 3 separate phases. 6
Project Phases 7
Daily Operations Insert text 8
Performance and Regions RFV RSC1 Insert readable map RSC RMF1 RMF 9
Performance and Regions 10
Performance and Regions 11
Special Events SCOOT cycle length for RSC1 Special event Normal Day 12
Special Events Ped volume at parking garage Special event Normal Day 13
PERFORMANCE MEASURES Data Sources Field Equipment Databases Internal Performance Measures Internal to SDOT staff Detailed reporting to evaluate system performance Public Facing Performance Measures External to users of the corridor High-level metrics to show benefits experienced by user
INTERNAL PERFORMANCE MEASURES Outcome-Oriented Metrics for an Arterial Metric Types Investment-Oriented: number of engineers/technician invested, amount of money spent on equipment upgrades, amount of time since last re-timing, etc. Outcome-Oriented Layers Intersection: signal data unique to the intersection. Performance measures can include red/green allocation, occupancy ratio, arrival data, etc. Outcome-Oriented: determines whether the control plan is effective, whether equipment needs maintenance, where capital investments might be needed, etc. Arterial: aggregate intersection data and travel time data. Performance measures can include travel times, before and after studies, corridor progression, etc. Public-Oriented: complaints, etc. Network: aggregate arterial data. Performance measures can include origin-destination studies, travel patterns, etc.
DATA SOURCES Roadside Equipment Acyclica WiFi Antennas Cabinet installed 10 installed along the corridor 22 additional installed in SLU Travel times and reliability EDI icite Data Aggregators Installed alongside Acyclica devices Monitors signal controller operations Arrivals on green Sensys Magnetometers Primary method of vehicle detection Vehicle throughput
DATA SOURCES Transit Databases King County Metro Transit Data GPS point whenever a bus stops within ½ mile buffer of Mercer Street Used for stop to stop travel time analysis Routes analyzed include RapidRideC, RapidRide D, 62, 40, and 70 SDOT Street Car Data Timestamp point when streetcar reaches stop Provides headway data at each stop Goal is to compare streetcar headways before and after SCOOT Goal is to measure transit travel times before and after SCOOT on and crossing Mercer St
INTERNAL PERFORMANCE MEASURES Travel Time and Travel Time Reliability Travel Time Comparisons Acyclica WiFi data used to measure travel times Eastbound and Westbound across the corridor Before SCOOT condition: 3+ weeks of data from March 2017 before adaptive turn on After SCOOT condition: 12 weeks of data after deployment date (March 27, 2017 - present) Internal and public facing report produced weekly Measuring average, 5 th and 95 th percentile travel times to demonstrate reliability improvements
Travel Time (minutes) INTERNAL PERFORMANCE MEASURES Mercer St Westbound Performance 20 18 16 14 12 10 8 6 4 2 0 WEEKDAY Travel Time on Mercer St Westbound from Fairview Ave to Queen Anne N Time of Day 5% No SCOOT 95% 5% SCOOT 95% Shows Average, 5 th and 95 th percentile WB travel times before and after SCOOT Minor travel time savings during both midday and PM periods Slightly increased reliability throughout the day
Percent of Trips INTERNAL PERFORMANCE MEASURES Mercer St Westbound PM Performance CDF Chart 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Mercer Westbound PM Peak Period Travel Times 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Travel Time (minutes) Shows what percentage of trips made during PM peak period occur at or below given travel time Steeper is better Farther left is better NO SCOOT SCOOT
Travel Time (minutes) INTERNAL PERFORMANCE MEASURES Mercer St Eastbound Performance 40 35 30 25 20 15 10 5 0 WEEKDAY Travel Time on Mercer St Eastbound from Queen Anne Ave N to Fairview Ave Time of Day Shows Average, 5 th and 95 th percentile EB travel times before and after SCOOT Biggest savings during PM peak Increased average speed Greatly improved reliability 5% No SCOOT 95% 5% SCOOT 95%
Percent of Trips INTERNAL PERFORMANCE MEASURES Mercer St Eastbound Performance CDF Chart 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Mercer EB PM Peak Period Travel Times 0 5 10 15 20 25 30 35 40 45 Travel Time (minutes) Shows what percentage of trips made during PM peak period occur at or below given travel time Steeper is better Farther left is better SCOOT NO SCOOT
INTERNAL PERFORMANCE MEASURES Arrivals on Green Dexter Ave & Mercer St 120% 100% 80% 60% 40% 20% 0% WB Arrivals without SCOOT 120% 100% 80% 60% 40% 20% 0% WB Arrivals with SCOOT
INTERNAL PERFORMANCE MEASURES Vehicle Throughput Fairview Ave & Mercer St Sensys Volume Data Exit detection Three detectors WB Four detectors EB
INTERNAL PERFORMANCE MEASURES Vehicle Throughput Sensys Volume Data Westbound
INTERNAL PERFORMANCE MEASURES Vehicle Throughput Sensys Volume Data Eastbound
Travel Time (min) INTERNAL PERFORMANCE MEASURES KC Metro Transit Transit Travel Time Comparison RapidRide D Line Analysis Extents: Compares stop to stop transit travel times for routes operating on or across Mercer St Measured for AM, PM and Daily Before SCOOT data: Weekdays in February 2017 Rapidride D Line Results: After SCOOT case: Weekdays in April 2017 Improved transit travel times both NB and 3.0 2.5 2.0 1.5 1.4 1.4 2.0 1.8 1.9 1.5 2.4 1.7 1.6 1.5 1.9 1.7 SB for D Line 1.0 0.5 0.0 Rapidride D NB Rapidride D SB Rapidride D NB Rapidride D SB Rapidride D NB Rapidride D SB AM PM Daily No SCOOT SCOOT
5% Average 95% 5% Average 95% 5% Average 95% Headway (min) 5% Average 95% 5% Average 95% 5% Average 95% Headway (min) INTERNAL PERFORMANCE MEASURES SDOT Street Car Streetcar Directional Headways Headways averaged northbound and southbound during AM, PM and daily Results show minor increase in headways at all time periods 35 30 25 20 15 10 5 0 Northbound Streetcar Headways 32.8 28.3 22.8 23.0 19.0 18.8 13.8 14.8 10.8 10.9 11.7 12.2 3.9 4.3 3.7 4.0 3.9 4.2 AM PM Daily No SCOOT SCOOT Southbound Streetcar Headways SB 35 30 25 20 15 10 5 0 18.2 18.3 10.7 11.0 13.5 5.1 4.8 3.8 4.1 28.4 13.9 29.0 4.5 21.5 21.8 11.612.0 4.5 NB AM PM Daily No SCOOT SCOOT
INTERNAL PERFORMANCE MEASURES Side Street Delay Measure before and after travel times along Dexter, 9 th Ave, Westlake, and Fairview Most side streets experience minor improvements in travel time
INTERNAL PERFORMANCE MEASURES Pedestrian Delay Measure % cycle time allocated for pedestrian phases (before and after SCOOT) Measure delay between pushbutton activation and pedestrian service
INTERNAL PERFORMANCE MEASURES Automated Web-Based Platform Brings in multiple data feeds to one platform (Acyclica, INRIX, transit providers, and others) Automates the data summary Makes it possible for clients to access data directly
PUBLIC FACING PERFORMANCE MEASURES Mercer St Website Data Spreadsheet
PUBLIC FACING PERFORMANCE MEASURES SDOT s Mercer St Website Summary of Benefits Travel time improvements in both directions in both AM and PM peak periods (3 hour avg) Reliability improvements in AM & PM for westbound, and PM eastbound Average commuter (WB in AM and EB in PM) savings in terms of time, C0 2, and gallons of gas
PUBLIC FACING PERFORMANCE MEASURES Return on Investment per Purdue Report Calculations User Cost Calculation assumes SOV, no trucks, peak hours only (WB AM, EB PM) User Cost per Year = Change in Travel Time x Value of Time per Hour = 23 hours (change in travel time) x $17.67 (value per hour) = $406.41 per vehicle per year. $406.41 (per vehicle) x 14,000 (vehicle throughput) = $5,689,740 yearly savings Fuel Consumption Fuel Consumption = Change in Travel Time x Average Fuel Consumption = 23 hours (change in travel time) x 0.87 gal/hour (avg. fuel cons.) = 20 gals, per person per year. 20 gals (per vehicle) x 14,000 (vehicle throughput) = 280,000 gallons Assume $3 per gallon = $840,000 CO2 Emissions CO2 Emissions = Fuel Consumption x CO2 Emitted per Gallon of Gasoline = 280,000 gallons (peak period vehicles) x 19.17lb/gal (per EPA) = 5,367,600 lbs = 2,683 tons CO2. Carbon Cost Carbon Cost = CO2 tons x Social Cost of CO2 = 2,683 tons CO2 (peak period vehicles) x $36 per ton (per EPA) = $96,617 TOTAL = $6,626,357
WSDOT ATSPMS GUIDE Additional Support Services One page summary for each performance measure currently being reported across the nation Support ATSPM deployment by identifying detection/communication needs, visualization tools, and methods to improve signal operations
Questions? Band.Sittikariya@seattle.gov Karl.Typolt@transpogroup.com www.seattle.gov/transportation