Mercer SCOOT Adaptive Signal Control
Karl Typolt, Transpo Group
PSRC RTOC
July 6th, 2017
Our mission, vision, and core values
Committed to 5 core values to create a city that is:
• Safe
• Interconnected
• Affordable
• Vibrant
• Innovative
Mission: deliver a high-quality
transportation system for Seattle
Vision: connected people,
places, and products
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?
4
• 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.
5
Traffic Volume and Cycle Length Comparison
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
7
Project Phases
Daily Operations
• Insert text
8
Performance and Regions
Insert readable map
9
RSC1
RSCRMF1
RMF
RFV
Performance and Regions
10
Performance and Regions
11
Special Events
12
Special event
Normal Day
SCOOT cycle length for RSC1
Special Events
13
Special event
Normal Day
Ped volume at parking garage
• Field Equipment
• Databases
Data Sources
PERFORMANCE MEASURES
• Internal to SDOT staff
• Detailed reporting to evaluate system performance
Internal Performance Measures
• External to users of the corridor
• High-level metrics to show benefits experienced by user
Public Facing Performance Measures
Investment-Oriented: number of
engineers/technician invested, amount of
money spent on equipment upgrades,
amount of time since last re-timing, etc.
Outcome-Oriented: determines whether
the control plan is effective, whether
equipment needs maintenance, where
capital investments might be needed, etc.
Public-Oriented: complaints, etc.
Metric Types Outcome-Oriented Layers
INTERNAL PERFORMANCE MEASURES
Outcome-Oriented Metrics for an Arterial
Intersection: signal data unique to the
intersection. Performance measures can
include red/green allocation, occupancy
ratio, arrival data, etc.
Arterial: aggregate intersection data and
travel time data. Performance measures
can include travel times, before and after
studies, corridor progression, etc.
Network: aggregate arterial data.
Performance measures can include
origin-destination studies, travel patterns,
etc.
• Cabinet installed
• 10 installed along the corridor
• 22 additional installed in SLU
• Travel times and reliability
Acyclica WiFi Antennas
• Installed alongside Acyclica devices
• Monitors signal controller operations
• Arrivals on green
EDI iCITE Data Aggregators
DATA SOURCES
Roadside Equipment
• Primary method of vehicle detection
• Vehicle throughput
Sensys Magnetometers
• 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
• Goal is to measure transit travel
times before and after SCOOT on
and crossing Mercer St
King County Metro – Transit Data
• Timestamp point when streetcar
reaches stop
• Provides headway data at each stop
• Goal is to compare streetcar headways
before and after SCOOT
SDOT – Street Car Data
DATA SOURCES
Transit Databases
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, 5th and 95th percentile travel times to demonstrate reliability improvements
Travel Time Comparisons
INTERNAL PERFORMANCE MEASURES
Travel Time and Travel Time Reliability
INTERNAL PERFORMANCE MEASURES
Mercer St Westbound Performance
Shows Average, 5th
and 95th percentile
WB travel times
before and after
SCOOT
Minor travel time
savings during both
midday and PM
periods
Slightly increased
reliability
throughout the day
0
2
4
6
8
10
12
14
16
18
20
Tra
ve
l T
ime
(m
inu
tes)
Time of Day
WEEKDAY Travel Time on Mercer St Westbound from Fairview Ave to Queen Anne N
5% No SCOOT 95% 5% SCOOT 95%
INTERNAL PERFORMANCE MEASURES
Mercer St Westbound PM Performance – CDF Chart
Shows what percentage
of trips made during PM
peak period occur at or
below given travel time
“Steeper is better”
“Farther left is better”
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Pe
rce
nt o
f T
rip
s
Travel Time (minutes)
Mercer Westbound PM Peak Period Travel Times
NO SCOOT SCOOT
INTERNAL PERFORMANCE MEASURES
Mercer St Eastbound Performance
0
5
10
15
20
25
30
35
40
Tra
ve
l T
ime
(m
inu
tes)
Time of Day
WEEKDAY Travel Time on Mercer St Eastbound from Queen Anne Ave N to Fairview Ave
5% No SCOOT 95% 5% SCOOT 95%
Shows Average, 5th
and 95th percentile
EB travel times
before and after
SCOOT
Biggest savings
during PM peak
• Increased
average speed
• Greatly
improved
reliability
INTERNAL PERFORMANCE MEASURES
Mercer St Eastbound Performance – CDF Chart
Shows what percentage
of trips made during PM
peak period occur at or
below given travel time
“Steeper is better”
“Farther left is better”
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 5 10 15 20 25 30 35 40 45
Pe
rce
nt o
f T
rip
s
Travel Time (minutes)
Mercer EB PM Peak Period Travel Times
SCOOT NO SCOOT
INTERNAL PERFORMANCE MEASURES
Arrivals on Green – Dexter Ave & Mercer St
0%
20%
40%
60%
80%
100%
120%
WB Arrivals without SCOOT
0%
20%
40%
60%
80%
100%
120%
WB Arrivals with SCOOT
Sensys Volume Data
INTERNAL PERFORMANCE MEASURES
Vehicle Throughput – Fairview Ave & Mercer St
• Exit detection
• Three detectors WB
• Four detectors EB
Sensys Volume Data – Westbound
INTERNAL PERFORMANCE MEASURES
Vehicle Throughput
Sensys Volume Data – Eastbound
INTERNAL PERFORMANCE MEASURES
Vehicle Throughput
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
After SCOOT case: Weekdays in April 2017
Improved transit travel times both NB and
SB for D Line
Transit Travel Time Comparison
INTERNAL PERFORMANCE MEASURES
KC Metro Transit
1.4
2.01.9
2.4
1.6
1.9
1.4
1.8
1.51.7
1.51.7
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Rapidride DNB
Rapidride DSB
Rapidride DNB
Rapidride DSB
Rapidride DNB
Rapidride DSB
AM PM Daily
Tra
vel T
ime (
min
)
Rapidride D Line Results:
No SCOOT SCOOT
Headways averaged northbound and
southbound during AM, PM and daily
Results show minor increase in headways
at all time periods
Streetcar Directional Headways
INTERNAL PERFORMANCE MEASURES
SDOT Street Car
3.9
10.8
19.0
3.7
13.8
28.3
3.9
11.7
22.8
4.3
10.9
18.8
4.0
14.8
32.8
4.2
12.2
23.0
0
5
10
15
20
25
30
35
5%
Avera
ge
95%
5%
Avera
ge
95%
5%
Avera
ge
95%
AM PM Daily
Headw
ay
(min
)
Northbound Streetcar Headways
No SCOOT SCOOT
5.1
10.7
18.2
3.8
13.5
28.4
4.5
11.6
21.5
4.8
11.0
18.3
4.1
13.9
29.0
4.5
12.0
21.8
05
101520253035
5%
Avera
ge
95%
5%
Avera
ge
95%
5%
Avera
ge
95%
AM PM Daily
Headw
ay
(min
)
Southbound Streetcar Headways
No SCOOT SCOOT
NB
SB
• Measure before and after travel times along Dexter, 9th Ave, Westlake, and Fairview
• Most side streets experience minor improvements in travel time
INTERNAL PERFORMANCE MEASURES
Side Street Delay
• Measure % cycle time
allocated for
pedestrian phases
(before and after
SCOOT)
• Measure delay
between pushbutton
activation and
pedestrian service
INTERNAL PERFORMANCE MEASURES
Pedestrian Delay
• 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
INTERNAL PERFORMANCE MEASURES
Automated Web-Based Platform
PUBLIC FACING PERFORMANCE MEASURES
Mercer St Website Data Spreadsheet
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, C02,
and gallons of gas
Summary of Benefits
PUBLIC FACING PERFORMANCE MEASURES
SDOT’s Mercer St Website
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
User Cost – Calculation assumes SOV, no trucks, peak hours only (WB AM, EB PM)
PUBLIC FACING PERFORMANCE MEASURES
Return on Investment per Purdue Report Calculations
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
Carbon Cost = CO2 tons x Social Cost of CO2 =
2,683 tons CO2 (peak period vehicles) x $36 per ton (per EPA) = $96,617
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
.
CO2 Emissions
Carbon Cost
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?
www.seattle.gov/transportation