• Save
ATIS: A Picture Book Approach to Traffic Signal Management
Upcoming SlideShare
Loading in...5
×
 

ATIS: A Picture Book Approach to Traffic Signal Management

on

  • 423 views

ITS Heartland 2012

ITS Heartland 2012
Annual Meeting
Kansas City, MO

Presented by Darcy Bullock, Professor at Purdue University

Statistics

Views

Total Views
423
Views on SlideShare
423
Embed Views
0

Actions

Likes
0
Downloads
2
Comments
0

0 Embeds 0

No embeds

Accessibility

Categories

Upload Details

Uploaded via as Microsoft PowerPoint

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment
  • Developing weekend plans is a challenge for most agencies because of the need to collect data on weekends.In this case, on SR 37 the weekend plan has not been updated for quite some time… so we felt it would be a good idea to take a look and see if there were any opportunities for improving it.Here are the progression diagrams… (talk through good/bad)Good progression at x, y, z shown by the distribution of vehicles coinciding with the green band.-Poor progression at a,b,c illustrated by the distribution of vehicles coinciding with the red.We have random arrivals at the northern entry point into the system… we also see what appear to be random arrivals occurring at 37 and Pleasant.
  • Here’s a closer look at the threecoord phases with poor quality of progression. Each of these shows a clear example of a case where we could improve the offsets. The vehicles appear to arrive in regular platoons, but those platoons are arriving during the red phase.
  • To illustrate this concept in detail, we will focus on the 2nd and 3rd intersections on SR 37.
  • Here’s an example where the approach on one end of the intersection has very good progression (POG = 80%), while the other on the opposite has rather poor progression (40%). Although we can change the offset at either intersection, there is only one relative offset between the two intersections that strongly influences the progression quality at both intersections. This is a classic example of a case where a tradeoff has to be made between one direction and the other…We’ll use progression diagrams to estimate what will happen as we change the offset… keeping track of total vehicle arrivals on green as the quantity we are trying to maximize.
  • Now that we have found some optimal offsets, we can use progression diagrams to estimate what the impact will be.First, we’ll put up the current weekend offsets and the actual time when vehicles are arriving during cycle for each intersection.
  • If we adjust the arrival times of vehicles according to the proposed offsets, this is the estimated impact on the system.(flip back and forth)As we can see here, we expect these adjustments to cause the coord platoons that are currently arriving in red to be shifted into green.The other offsets that were formerly good have not been negatively affected. In this case, we predict that southbound at 37/Town and Country will suffer a little, but it should be offset by substantial improvements at the other approaches.

ATIS: A Picture Book Approach to Traffic Signal Management ATIS: A Picture Book Approach to Traffic Signal Management Presentation Transcript

  • Operational Performance Measures and A PictureBased Outcome Book AssessmentApproach for Arterial Management using High Resolution to Traffic Data and DatabaseController Data, Probe Signal Management QA/QC Procedures Darcy Bullock darcy@purdue.edu Purdue University School of Civil Engineering
  • March 27, 2012: 0830
  • Why?
  • Advanced Transportation Information Systems (ATIS) Messages Opportunities toWhat gets measured gets done, Push the State ofwhat gets measured and fed back the Possiblegets done well,what gets rewarded gets repeated.– John E. JonesEnormous opportunities to fuse data from• traffic signal controllers and• probe data sources
  • Circa 1995 Screen Scrapes (PPTs were so old they did not load)LSU vehicle (25k miles) I-10/12 Split Segment Coding R D S an t nna eFM s ub ca r re r i G P S an t nna e S a t llit s gna l e e i 1001 7 1005 1004 1002 1006 101 1003 1008 1010 1009 7 101 10 100 3 10 11 14 1018 RD S T ri b e m l 10 D if fe ren ta lco r re c ton i i 15 10 300 0 P a ce r l 12 400 10 16 Lap t p o co pu t r m e DG PS da t a
  • Evolution LSU Early Controller CMU (Denardo Data (Brute Force Years)1990 2000 2010 Brute force Performance GPS Data Saban joins Measures LSU
  • October 2006 State of the practice
  • “If one wants to be outstanding inthere field, one must first spend sometime standing in the field” Bill Kloose
  • Contributor to this talk• INDOT (Infrastructure • Indiana LTAP Support and Agency – Neal Carboneau Perspective) – Jay Grossman (Elkhart – Jim Sturdevant County) – Jay Wasson • Andrew Nichols (Marshall – Ryan Gallagher University) – Greg Richards • Econolite (ASC 3 Data• Purdue University Logger ) – Chris Day – Gary Duncan – Tom Brennan – Eric Raamot – Ross Haseman – Lu Ta – Alex Hainen – Brian Griggs – Steve Remias – Darcy Bullock 12
  • Emerging Shared Vision Active Traffic Signal 1. Develop Management infrastructure and procedures to systematically prioritize investing AgenciesUniversities Vendors engineering resources 2. Assess that impact We are in a period where we need to re- introduce theory and fundamentals so we change how agencies spec. & operate traffic signals and what vendors provide and
  • True or FalseOn Average, we have no capacity problems at anysignalized intersection? 250 200 Counts per 15 minutes 150 100 50 0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 Time of Day
  • 15-Minute Counts (Phase “n”) 250 200Counts per 15 minutes 150 100 50 0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 Time of Day 15
  • ChallengeHow can we use controller and probe data to quantifywhat we can see and assess if we can improve.
  • Outline• Signal Management/ Challenges• Data Sources – Controller – Probe Vehicles• Fundamentals• Data ->Information to assess fundamentals• Results• Q&A..Dialog Don’t wait until the End
  • INDOT Signal NetworkQuestion• Where (and when) are the opportunities to improve signal operations? INDOT: 2,600 signals in 300 systems Nationwide: 350,000 signals Globally: who knows?18
  • ~350,000 Traffic Signals…We need systematicprocedures for identifying operationalproblems…and fixing them using controller andprobe data.
  • ChicagoTypical Corridor (22 Intersections) 80 65 65 65 Lafayette 74,000 Indianapo Parameters/int 70 N Map Area Isolated Indianapolis, IN 65 (Free) Operation Coordinated Operation 1 65 2 3 US30 d) 5 ft antenna 4 5 6 8 10 12 65 18 2000 ft 20 22 14 16 500 m e) 7.5 ft antenna US30 d) 5 ft antennaN 4,000 FT 1,000 M 7 9 11 65 15 13 17 19 21 e) 7.5 ft antenn 2000 ft 500 m US30 d) 5 ft antenna
  • ~74,000 ConfigurableParameters/Int 192M Statewide! ~1.6M Parameters Default Database Default Database Query of Each Intersection1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Longitudinal Query of Intersection Controller Databases
  • So how many do we really use? 3200 Isolated Coordinated Vendor Specific 3000 (Free) Operation Operation NTCIP 1202 2800 2600 2400 COUNT OF ALTERED PARAMETERS 2200 2000 1800 1600 1400 1200 1000 800 600 400 200 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 INTERSECTION ID
  • Traffic Signal Timing Process Opportunity Today’s Message I. II. III. IV. V. VI. Define Assembly Software Timing Design Deployment Assess Objectives, relevant data Modeling and Docu- Assess and to support mentation Prioritize timing andactivities by docu-Time of Day mentationand location objectives
  • Outline• Signal Management/ Challenges• Data Sources – Controller – Probe Vehicles• Fundamentals• Data ->Information to assess fundamentals• Results• Q&A..Dialog
  • Historical data collection (~1,000:1)averages out all of the information Typical less then 500 average values/day sent to Traffic ~500,000 Management Centers Events/day (and those are rarely archived)
  • Easter Weekend: SouthboundExample 40 min travel time sample Observed at 18:10 ~MM 244 ~40 ~MM 220 minutes Observed at 18:50
  • Easter WeekendSouthbound Travel Time (~24 miles) 140 Friday Saturday Sunday Monday 120 SB Travel Time (minutes) Approximately 1 hour of delay Diversions on 100 State/Local Roads 80 60 Normal Travel Time ~ 22 minutes 40 20 0 4/10/09 0:00 4/11/09 0:00 4/12/09 0:00 4/13/09 0:00 4/14/09 0:00
  • Sample SR 32 Arterial Data SR 32 Instrumented Arterial from SR 238 to SR 37
  • Econolite ASC 3Bluetooth Antenna with Indiana Data Logger Enabled SR 32 @ SR 238 Bluetooth Data Ethernet Switch Logger
  • SR-37 Travel Times 10 Perhaps 9 Opportunity to 8 improve on Saturday 7Travel Time (min) 6 5 4 3 2 1 Perhaps M T W Th F Sa Su M T W Th F Sa Su Opportunity to 0 6/15 6/16 6/17 6/18 6/19 6/20 6/21 6/22 6/23 6/24 6/25 6/26 6/27 6/28 6/29 improve on Saturday
  • Outline• Signal Management/ Timing overview• Data Sources – Controller – Probe Vehicles• Fundamentals• Data ->Information to assess fundamentals• Results• Q&A..Dialog
  • Highway Capacity Manual Delay Equation Oversaturation d d1 PF d2 d3 (Split Failures) 2 g ,a 1 P ,a 0.5C 1 PF ,a fPA Ca g ,ad1, ,a 1 g ,a C 1 min 1 X , ,a Ca1 Capacity Utilization (Volume-to-Capacity Ratio) Quality of Progression (Percent on Green) 2
  • Coordination: Split, Cycle, Offset 5L Cycle Really Hard 4L to get perfect for 22 IntersectionsDistance 3L 2L L 1 2 3 4 Time (cycles)
  • Coordination: Split, Cycle, Offset 5L Phone calls 4L work pretty wellDistance 3L 2L Split L 1 2 3 4 Time (cycles)
  • Coordination: Split, Cycle, Offset 5L Tuning is 4L labor intensive Distance 3L 2LOffset L 1 2 3 4 Time (cycles)
  • Coordination: Split, Cycle, Offset 74,000 parameters, 1000’s of opportunities to make mistakes 5L Cycle 4L Distance 3L 2LOffset Split L 1 2 3 4 Time (cycles)
  • Operational Performance Measures and A PictureBased Outcome Book AssessmentApproach for Arterial Management using High Resolution to Traffic Data and DatabaseController Data, Probe Signal Management QA/QC Procedures
  • Outline• Signal Management/ Challenges• Data Sources – Controller – Probe Vehicles• Fundamentals• Data ->Information to assess fundamentals• Results• Q&A..Dialog
  • Purdue Coordination Diagram Construction (PCD)LoopDetection Time in cycle 120 Cycle boundary 90 Red Green phase ends 70 Green Cycle begins window phase begins Cycle ends Green 50 Green time 0 0 sec 50 sec 90 sec 120 sec12:00:00 12:02:00 time of day 12:00:00 12:01:10 12:02:00 70 sec 12:01:10
  • a c383 c384 f c g b e d c385 c386 c387 c388 i c389 c390 h c391 c392 c393 c394 c395 c396 c397 Phase 3 Phase 4 Phase 1 Clearance Phase 2 Green52 Phase 2 Red
  • Timing Plan Pattern1 24 Hour 3Overview 2 4 5 6 7 8 20-pt. moving average b2 a2 a1 b1 53
  • NB @ Pleasant NB @ SR 238 Travel Times 10 9 8 7Travel Time (min) 6 Better Progression 5 4 3 2 1 Perhaps M T W Th F Sa Su M T W Th F Sa Su Opportunity to 0 6/1 6/2 6/3 6/4 6/5 6/6 6/7 6/8 6/9 6/10 6/11 6/12 6/13 6/14 6/15 improve on Saturday
  • Saturday Offset Adjustment SR 32 goodRandomarrivals Pleasant Noplatoons bad Town & Country good good bad Greenfield bad 55
  • Three Poor OffsetsNB @ 37/Pleasant SB @ 37/Greenfield NB @ 37/Greenfield 56
  • Offset Adjustments on Middle Segment SR 32 goodRandomarrivals Pleasant Noplatoons bad Town & Country good good bad Greenfield bad 57
  • Offset Adjustments on Middle Segment• Northbound at 37/Pleasant is bad. The platoon is captured in red.• However, any offset adjustments that we make will also impact Southbound progression at 37/Town and Country (intersection to south) by shifting arrivals.• We can mitigate any impacts at 32/37 (intersection to north) by adjusting its offset to keep it fixed relative to 37/Pleasant.NB @ 37/Pleasant SB @ 37/Town and Country POG = 40.1% POG = 80.2% 5069 arrivals on green (0600-2200) 58
  • Add 10 seconds at 37/Pleasant• Green times will occur 10 seconds earlier at 37 & Pleasant – Equivalent to vehicles arriving 10 seconds later• Southbound vehicles will arrive 10 seconds earlier at 37 & Town and CountryNB @ 37/Pleasant SB @ 37/Town and Country POG = 55.4% POG = 77.8% 5589 arrivals on green 59
  • Add 20 seconds at 37/Pleasant• Green times will occur 20 seconds earlier at 37 & Pleasant – Equivalent to vehicles arriving 20 seconds later• Southbound vehicles will arrive 20 seconds earlier at 37 & Town and CountryNB @ 37/Pleasant SB @ 37/Town and Country POG = 67.4% POG = 68.8% 5688 arrivals on green 60
  • Add 30 seconds at 37/Pleasant• Green times will occur 30 seconds earlier at 37 & Pleasant – Equivalent to vehicles arriving 30 seconds later• Southbound vehicles will arrive 30 seconds earlier at 37 & Town and CountryNB @ 37/Pleasant SB @ 37/Town and Country POG = 73.4% POG = 57.5% 5446 arrivals on green 61
  • Comparison with original (actual “before” case) SR 32 good Pleasant bad Town & Country good good bad Greenfield bad 62
  • Predicted Vehicle Distributions with Offset Adjustments SR 32 Unchanged Pleasant Better Town & Country Still OK Still OK Better Greenfield Better 63
  • Outline• Signal Management/ Timing overview• Data Sources – Controller – Probe Vehicles• Fundamentals• Data ->Information to assess fundamentals• Results• Q&A..Dialog
  • So..did it have an impact.
  • SB, SR 37 / SR 32 NB, SR 37 / SR 32Before d 1001 0600 1400 2200 0600 1400 2200 SB, SR 37 / Pleasant NB, SR 37 / Pleasant 1002 a 0600 1400 2200 0600 1400 2200 SB, SR 37 / Town & Country NB, SR 37 / Town & Country 1003 0600 1400 2200 0600 1400 2200 SB, SR 37 / Greenfield NB, SR 37 / Greenfield c 1004 b 0600 1400 2200 0600 1400 2200
  • SB, SR 37 / SR 32 NB, SR 37 / SR 32 eAfter d 1001 0600 1400 2200 0600 1400 2200 SB, SR 37 / Pleasant NB, SR 37 / Pleasant a 1002 z z 0600 1400 2200 0600 1400 2200 SB, SR 37 / Town & Country NB, SR 37 / Town & Country 1003 f 0600 1400 2200 0600 1400 2200 SB, SR 37 / Greenfield NB, SR 37 / Greenfield b c 1004 0600 1400 2200 0600 1400 2200
  • SB, SR 37 / SR 32 NB, SR 37 / SR 32 ePredicted d 1001 0600 1400 2200 0600 1400 2200 SB, SR 37 / Pleasant NB, SR 37 / Pleasant a 1002 0600 1400 2200 0600 1400 2200 SB, SR 37 / Town & Country NB, SR 37 / Town & Country 1003 f 0600 1400 2200 0600 1400 2200 SB, SR 37 / Greenfield NB, SR 37 / Greenfield b c 1004 0600 1400 2200 0600 1400 2200
  • Before 9 (Sample size=4797) 8 7 Long Saturday travel times compared to restMeasured Travel Time (min) of week 6 5 4 3 2 1 Mon Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun 0 6/15 6/16 6/17 6/18 6/19 6/20 6/21 6/22 6/23 6/24 6/25 6/26 6/27 6/28 6/29 Date/Time
  • After 9 (Sample size=5401) 8 7 Saturday travel timesMeasured Travel Time (min) comparable to rest of week 6 5 4 3 2 1 Mon Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun 0 7/13 7/14 7/15 7/16 7/17 7/18 7/19 7/20 7/21 7/22 7/23 7/24 7/25 7/26 7/27 Date/Time
  • Lets Evaluate the ImpactStatistically and Financially I. II. III. IV. V. VI. Define Assembly relevant Software Modeling Timing Design and Deployment Assess Objectives, Assess data to support Docu- mentation and Prioritize timing and docu- activities by Time mentationof Day and location objectives
  • NB: June 6, 2009 0900-1200 Travel Time Histograms 60 100% 90% 50 80% Frequency 70% 40 Cumulative % 60%Frequency 30 50% 40% 20 30% 20% 10 10% 0 0% 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 More Travel Time Bin (Minutes)
  • NB: July 18, 2009 0900-1200 Travel Time Histograms 60 100% 90% 50 80% Frequency 70% 40 Cumulative % 60%Frequency 30 50% 40% 20 30% 20% 10 10% 0 0% 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 More Travel Time Bin (Minutes)
  • Before (6/20/09) After (7/25/09) ~1 min Travel Time Reduction
  • Business Case: SR 37 Timing Improvements (Largest Cost Benefit/Reduction/Avoidance) USER SAVINGS • Travel time tests for SR37 Corridor 1 have improved Northbound Travel Time by ~ 1 Minute. 0.75 • ~8,500 Cars per Day Are Effected byCumulative Probability this benefit (NB). 0.5 • ~0.17 Cents per minute ($10/hour) saved for each driver in fuel costs and time value. 0.25 • ~1.0 Minutes are assumed saved on average for the intersection over 1- 0 Year with improvements. 0 1 2 3 4 5 6 7 8 9 Travel Time (min) • User benefit = 0600 -2200 (8,500 veh) (8,500 Veh/Day * $0.17/min * 1- min/Veh * 2*52 Days/Year )= $150,000/year for a 1.6 mile stretch of roadway is realized. 76
  • Southbound Northbound Baseline Optimized (Obj. III) Baseline Optimized (Obj. III)1.0 1.0 1.0 1.0 1/SB 1/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.0 Int. 11.0 1.0 1.0 1.0 2/SB 2/NB0.5 0.5 Int. 2 0.5 0.50.0 0.0 0.0 0.01.0 1.0 Int. 3 1.0 1.0 3/SB 3/NB0.5 0.5 0.5 0.50.0 0.0 Int. 4 0.0 0.01.0 1.0 1.0 1.0 4/SB 4/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.01.0 1.0 1.0 1.0 5/SB 5/NB0.5 0.5 Int. 5 0.5 0.50.0 0.0 0.0 0.01.0 1.0 Int. 6 1.0 1.0 6/SB 6/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.01.0 1.0 1.0 1.0 7/SB Int. 7 7/NB0.5 0.5 0.5 0.5 Int. 80.0 0.0 0.0 0.01.0 1.0 1.0 1.0 8/SB 8/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.0 0 29 57 86 114 0 29 57 86 114 0 29 57 86 114 0 29 57 86 114 Time in Cycle (s) Time in Cycle (s) Time in Cycle (s) Time in Cycle (s)
  • Travel Time between 4 intersections100% 100% IV IV I III75% 75% II II I50% 50% III Base Base25% 25% 0% 0% 1 2 3 4 5 6 7 1 2 3 4 5 6 7 (e) Southbound, Case B to Case C. (f) Northbound, Case C to Case B. Cumulative frequency diagrams of probe vehicle travel times for alternative objective functions, Saturday, 1500-1800.
  • Economic Impact (52 Saturdays) Daily Annual CO2 CO2 Total Time Emission Emission Saved Reduction CO2 User Multi- Reduction CO2 UserObjective (veh-min) (tons) Savings Benefits plier (tons) Savings Benefits (a) System 1, Northern Section I Min Delay 5032 0.71 $16 $1,697 52 37 $810 $88,233 II Min Delay and Stops 3813 0.54 $12 $1,286 52 28 $614 $66,864 III Max Ng 1760 0.25 $5 $593 52 13 $283 $30,855 IV Alt. Max Ng 7883 1.11 $24 $2,658 52 58 $1,268 $138,229 (b) System 2, Southern Section I Min Delay 24386 3.43 $75 $8,223 52 178 $3,924 $427,614 II Min Delay and Stops 25327 3.56 $78 $8,541 52 185 $4,075 $444,111 III Max Ng 25147 3.54 $78 $8,480 52 184 $4,046 $440,962 IV Alt. Max Ng 26338 3.70 $81 $8,882 52 193 $4,238 $461,845 (c) System 1 and System 2, Arterial I Min Delay 29418 4.14 $91 $9,920 52 215 $4,733 $515,847 II Min Delay and Stops 29140 4.10 $90 $9,826 52 213 $4,689 $510,976 III Max Ng 26907 3.78 $83 $9,073 52 197 $4,329 $471,817 IV Alt. Max Ng 34221 4.81 $106 $11,540 52 250 $5,506 $600,073 Controller PM + Probe Data
  • Emerging Shared Vision Active Traffic Signal 1. Develop Management infrastructure and procedures to systematically prioritize investing AgenciesUniversities Vendors engineering resources 2. Assess that impact We are in a period where we need to re- introduce theory and fundamentals so we change how agencies spec. & operate traffic signals and what vendors provide and
  • 1 $150,000 0.75 Cumulative Probability 0.5 0.25 0 0 1 2 3 4 5 6 7 8 9 Travel Time (min) 0600 -2200 (8,500 veh)What gets measured gets done,what gets measured and fed back gets done well,what gets rewarded gets repeated.– John E. Jones Darcy Bullock darcy@purdue.edu Purdue University School of Civil Engineering
  • Diagnostics (Jim Sturdevant)
  • SR37 Fishers/Noblesville- (9) Very good 2 way progression
  • Coordination: Split, Cycle, Offset 5L Cycle Really Hard 4L to get perfect for 22 IntersectionsDistance 3L 2L L 1 2 3 4 Time (cycles)
  • US 31 @ 126th (8 in system) Plan /cycle change discrepancy from north
  • Diagnostics (Andrew Nichols)
  • Thursday (12/8) Should we still be running coord @ 9pm?
  • Opportunities Active Traffic Signal 1. Hi Resolution Management Controller Data Performance Measure Picture BookUniversities Agencies Vendors Manual 2. Integrated (but independent) Probe Data We are in a period where we need to re- introduce theory and fundamentals so we Assessment change how agencies spec. & operate traffic signals and what vendors provide and
  • Scales well…4 more intersectionsadded to corridor in Summer 2010
  • Alternative Objective Functions 8 14 Performance Index (veh-h/h) 7 12 6 Delay (veh-h/h) 10 5 8 +44 4 +56 6 3 4 2 1 2 0 0 0 15 30 45 60 75 90 0 15 30 45 60 75 90 Offset Adjustment Offset Adjustment (a) Objective I: Estimated delay. (b) Objective II: Delay and stops. 600 400 350 +40 500 Alt. Number on Green Number on Green 300 400 +44 250 300 200 150 200 100 100 50 0 0 0 15 30 45 60 75 90 0 15 30 45 60 75 90 Offset Adjustment Offset Adjustment (c) Objective III: Number of arrivals on (d) Objective IV: Alternative number of green. arrivals on green. 1
  • Southbound Northbound Baseline Optimized (Obj. III) Baseline Optimized (Obj. III)1.0 1.0 1.0 1.0 1/SB 1/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.0 Int. 11.0 1.0 1.0 1.0 2/SB 2/NB0.5 0.5 Int. 2 0.5 0.50.0 0.0 0.0 0.01.0 1.0 Int. 3 1.0 1.0 3/SB 3/NB0.5 0.5 0.5 0.50.0 0.0 Int. 4 0.0 0.01.0 1.0 1.0 1.0 4/SB 4/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.01.0 1.0 1.0 1.0 5/SB 5/NB0.5 0.5 Int. 5 0.5 0.50.0 0.0 0.0 0.01.0 1.0 Int. 6 1.0 1.0 6/SB 6/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.01.0 1.0 1.0 1.0 7/SB Int. 7 7/NB0.5 0.5 0.5 0.5 Int. 80.0 0.0 0.0 0.01.0 1.0 1.0 1.0 8/SB 8/NB0.5 0.5 0.5 0.50.0 0.0 0.0 0.0 0 29 57 86 114 0 29 57 86 114 0 29 57 86 114 0 29 57 86 114 Time in Cycle (s) Time in Cycle (s) Time in Cycle (s) Time in Cycle (s)
  • Travel Time between 4 intersections100% 100% IV IV I III75% 75% II II I50% 50% III Base Base25% 25% 0% 0% 1 2 3 4 5 6 7 1 2 3 4 5 6 7 (e) Southbound, Case B to Case C. (f) Northbound, Case C to Case B. Cumulative frequency diagrams of probe vehicle travel times for alternative objective functions, Saturday, 1500-1800.
  • Economic Impact Daily Annual CO2 CO2 Total Time Emission Emission Saved Reduction CO2 User Multi- Reduction CO2 UserObjective (veh-min) (tons) Savings Benefits plier (tons) Savings Benefits (a) System 1, Northern Section I Min Delay 5032 0.71 $16 $1,697 52 37 $810 $88,233 II Min Delay and Stops 3813 0.54 $12 $1,286 52 28 $614 $66,864 III Max Ng 1760 0.25 $5 $593 52 13 $283 $30,855 IV Alt. Max Ng 7883 1.11 $24 $2,658 52 58 $1,268 $138,229 (b) System 2, Southern Section I Min Delay 24386 3.43 $75 $8,223 52 178 $3,924 $427,614 II Min Delay and Stops 25327 3.56 $78 $8,541 52 185 $4,075 $444,111 III Max Ng 25147 3.54 $78 $8,480 52 184 $4,046 $440,962 IV Alt. Max Ng 26338 3.70 $81 $8,882 52 193 $4,238 $461,845 (c) System 1 and System 2, Arterial I Min Delay 29418 4.14 $91 $9,920 52 215 $4,733 $515,847 II Min Delay and Stops 29140 4.10 $90 $9,826 52 213 $4,689 $510,976 III Max Ng 26907 3.78 $83 $9,073 52 197 $4,329 $471,817 IV Alt. Max Ng 34221 4.81 $106 $11,540 52 250 $5,506 $600,073
  • Assessing Deployments is Essentialand Possible….we just have to do it I. II. III. IV. V. VI. Define Assembly relevant Software Modeling Timing Design and Deployment Assess Objectives, Assess data to support Docu- mentation and Prioritize timing and docu- activities by Time mentationof Day and location objectives 9 8 7 Long Saturday travel times compared to rest Measured Travel Time (min) of week 6 5 4 V/C ratio, Northbound approach Phase 2, Noblesville, 07/25/2006 3 Equivalent Hourly Volume 2 2 Free AM Off- Mid- Off- PM Off- Free Peak Peak day Peak Peak peak 1 Mon Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun 0 6/15 6/16 6/17 6/18 6/19 6/20 6/21 6/22 6/23 6/24 6/25 6/26 6/27 6/28 6/29 Date/Time V/C ratio 1 0 0:00 6:00 12:00 18:00 0:00 Time of Day
  • Closing message• Let’s stop modeling what we can measure.• Let’s continue to enhance controllers to help us manage our systems.• Let’s close the “system performance measurement” loop.
  • References• Day, C.M., R.J. Haseman, H. Premachandra, T.M. Brennan, J.S. Wasson, J.R. Sturdevant, and D.M. Bullock, “Visualization and Assessment of Arterial Progression Quality Using High Resolution Signal Event Data and Measured Travel Time,” Transportation Research Board Paper ID:10- 0039, January 2010.• Bullock, D.M., R.J. Haseman, J.S. Wasson, and R. Spitler, “Anonymous Bluetooth Probes for Airport Security Line Service Time Measurement: The Indianapolis Pilot Deployment,” Transportation Research Board Paper ID:10-1438, January 2010.• Haseman, R.J., J.S. Wasson, and D.M.Bullock, “Real Time Measurement of Work Zone Travel Time Delay and Evaluation Metrics,” Transportation Research Board Paper ID:10-1442, January 2010.• Day, C.M., J.R. Sturdevant, and D.M. Bullock, “Outcome Oriented Performance Measures for Signalized Arterial Capacity Management,” Transportation Research Board Paper ID:10-0008, January 2010.• Smaglik E.J., A. Sharma, D.M. Bullock, J.R. Sturdevant, and G. Duncan, “Event-Based Data Collection for Generating Actuated Controller Performance Measures," Transportation Research Record, #2035, TRB, National Research Council, Washington, DC, pp.97- 106, 2007.
  • OperationalPicture Book A Performance Measures and Outcome Based Approach Assessment for Arterial Management to Traffic Signal using High Resolution Controller Data and Bluetooth Probes Management Jay Wasson & Jim Sturdevant, Indiana Department of Transportation Chris Day, Ross Haseman, Tom Brennan, Alex Hainen, Steve Remias & Darcy Bullock Purdue University School of Civil Engineering
  • •SR 37 Free
  • 400 1/14/2010 Clear PM Peak (Vehicle Arrival w/Reference to Upstream BOG) T&C to Pleasant 350 300 250Vehicles 200 150 100 50 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Travel Time (Seconds)
  • 400 1/7/2010 Snow PM Peak (Vehicle Arrival w/Reference to Upstream BOG) T&C to Pleasant 350 300 250Vehicles 200 150 100 50 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Travel Time (Seconds)
  • 32/37 January 23 37/T&C 37/Pleasant Observed 37/Greenfield 1 0.03 1 0.03 1 0.03 1 0.03 0.9 0.9 0.9 0.9 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Proportion of Vehicle Distribution Proportion of Vehicle Distribution Proportion of Vehicle Distribution Proportion of Vehicle Distribution 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 Probability of Green Probability of Green Probability of Green Probability of Green 0.6 0.6 0.6 0.6NB 0.5 0.4 0.015 0.5 0.4 0.015 0.5 0.4 0.015 0.5 0.4 0.015 0.01 0.01 0.01 0.01 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0 0 0 0 0 0 0 0 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 Time in Cycle (sec) Time in Cycle (sec) Time in Cycle (sec) Time in Cycle (sec) 32/37 37/Pleasant 37/T&C 37/Greenfield 1 0.03 1 0.03 1 0.03 1 0.03 0.9 0.9 0.9 0.9 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Proportion of Vehicle Distribution Proportion of Vehicle Distribution Proportion of Vehicle Distribution Proportion of Vehicle Distribution 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 Probability of Green Probability of Green Probability of Green Probability of Green 0.6 0.6 0.6 0.6SB 0.5 0.4 0.015 0.5 0.4 0.015 0.5 0.4 0.015 0.5 0.4 0.015 0.01 0.01 0.01 0.01 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0 0 0 0 0 0 0 0 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 Time in Cycle (sec) Time in Cycle (sec) Time in Cycle (sec) Time in Cycle (sec) 105
  • 100% 1 0.03 1 0.03 1 0.03 1 0.03 3% NB @ Int. 1001 NB @ Int. 1002 NB @ Int. 1003 NB @ Int. 1004 Observed – June 6, 2009 0.9 0.9 0.9 0.9Probability of Green 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Pct. of Vehicles 75% 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 2% 0.6 0.6 0.6 0.6 50% 0.5 0.015 0.5 0.015 0.5 0.015 0.5 0.015 0.4 0.4 0.4 0.4 0.3 0.01 0.3 0.01 0.3 0.01 0.3 0.01 1% 25% 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0% 0 0 0 0 0 0 0 0 0% 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 Time in Cycle Time in Cycle Time in Cycle Time in Cycle 140 NB @ Int. 1001 NB @ Int. 1002 NB @ Int. 1003 NB @ Int. 1004 105Time in Cycle 70 35 0 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Hour Hour Hour Hour 100% 1 0.03 1 0.03 1 0.03 1 0.03 3% 0.9 SB @ Int. 1001 0.9 SB @ Int. 1002 0.9 SB @ Int. 1003 0.9 SB @ Int. 1004Probability of Green 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Pct. of Vehicles 75% 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 2% 0.6 0.6 0.6 0.6 50% 0.5 0.015 0.5 0.015 0.5 0.015 0.5 0.015 0.4 0.4 0.4 0.4 0.3 0.01 0.3 0.01 0.3 0.01 0.3 0.01 1% 25% 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0% 0 0 0 0 0 0 0 0 0% 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 Time in Cycle Time in Cycle Time in Cycle Time in Cycle 140 SB @ Int. 1001 SB @ Int. 1002 SB @ Int. 1003 SB @ Int. 1004 105Time in Cycle 70 35 0 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Hour Hour Hour Hour
  • 100% 1 0.03 1 0.03 1 0.03 1 0.03 3% NB @ Int. 1001 NB @ Int. 1002 NB @ Int. 1003 NB @ Int. 1004 Predicted Optimized – June 6, 2009 Data 0.9 0.9 0.9 0.9Probability of Green 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Pct. of Vehicles 75% 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 2% 0.6 0.6 0.6 0.6 50% 0.5 0.015 0.5 0.015 0.5 0.015 0.5 0.015 0.4 0.4 0.4 0.4 0.3 0.01 0.3 0.01 0.3 0.01 0.3 0.01 1% 25% 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0% 0 0 0 0 0 0 0 0 0% 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 Time in Cycle Time in Cycle Time in Cycle Time in Cycle 140 NB @ Int. 1001 NB @ Int. 1002 NB @ Int. 1003 NB @ Int. 1004 105Time in Cycle 70 35 0 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Hour Hour Hour Hour 100% 1 0.03 1 0.03 1 0.03 1 0.03 3% 0.9 SB @ Int. 1001 0.9 SB @ Int. 1002 0.9 SB @ Int. 1003 0.9 SB @ Int. 1004Probability of Green 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Pct. of Vehicles 75% 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 2% 0.6 0.6 0.6 0.6 50% 0.5 0.015 0.5 0.015 0.5 0.015 0.5 0.015 0.4 0.4 0.4 0.4 0.3 0.01 0.3 0.01 0.3 0.01 0.3 0.01 1% 25% 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0% 0 0 0 0 0 0 0 0 0% 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 Time in Cycle Time in Cycle Time in Cycle Time in Cycle 140 SB @ Int. 1001 SB @ Int. 1002 SB @ Int. 1003 SB @ Int. 1004 105Time in Cycle 70 35 0 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Hour Hour Hour Hour
  • 100% 1 0.03 1 0.03 1 0.03 1 0.03 3% NB @ Int. 1001 NB @ Int. 1002 NB @ Int. 1003 NB @ Int. 1004 Observed (After Changes) – July 25, 2009 0.9 0.9 0.9 0.9Probability of Green 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Pct. of Vehicles 75% 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 2% 0.6 0.6 0.6 0.6 50% 0.5 0.015 0.5 0.015 0.5 0.015 0.5 0.015 0.4 0.4 0.4 0.4 0.3 0.01 0.3 0.01 0.3 0.01 0.3 0.01 1% 25% 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0% 0 0 0 0 0 0 0 0 0% 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 Time in Cycle Time in Cycle Time in Cycle Time in Cycle 140 NB @ Int. 1001 NB @ Int. 1002 NB @ Int. 1003 NB @ Int. 1004 105Time in Cycle 70 35 0 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Hour Hour Hour Hour 100% 1 0.03 1 0.03 1 0.03 1 0.03 3% 0.9 SB @ Int. 1001 0.9 SB @ Int. 1002 0.9 SB @ Int. 1003 0.9 SB @ Int. 1004Probability of Green 0.025 0.025 0.025 0.025 0.8 0.8 0.8 0.8 Pct. of Vehicles 75% 0.7 0.7 0.7 0.7 0.02 0.02 0.02 0.02 2% 0.6 0.6 0.6 0.6 50% 0.5 0.015 0.5 0.015 0.5 0.015 0.5 0.015 0.4 0.4 0.4 0.4 0.3 0.01 0.3 0.01 0.3 0.01 0.3 0.01 1% 25% 0.2 0.2 0.2 0.2 0.005 0.005 0.005 0.005 0.1 0.1 0.1 0.1 0% 0 0 0 0 0 0 0 0 0% 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 0 25 50 75 100 114 Time in Cycle Time in Cycle Time in Cycle Time in Cycle 140 SB @ Int. 1001 SB @ Int. 1002 SB @ Int. 1003 SB @ Int. 1004 105Time in Cycle 70 35 0 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Hour Hour Hour Hour
  • Audience Participation on Following Two Slides (Eye Doctor A/B)
  • SB on SR 37 on Jan 19 (Clear)
  • SB on SR 37 on Jan 26 (Snow)
  • Purdue Coordination Diagrams as Changes Were Implemented CR17 and Missouri Elkhart County, IN Ross Haseman
  • N PPD Before Change, Φ1 Φ2 Φ3 Φ4 Phase 6, 02/17/09 Φ5 Φ6 Φ7 Φ8 NA NB Green Time Red Time200190180 End of170160 Green150140 Platoon130120110100 90 Start of 80 70 Green 60 50 40 30 20 10 00:00:00.0 3:00:00.0 6:00:00.0 9:00:00.0 12:00:00.0 15:00:00.0 18:00:00.0 21:00:00.0 0:00:00.0
  • N Predicted PPD Φ2 Φ3 Φ1 Φ4 After Change, Phase 6 Φ5 Φ6 Φ7 Φ8 02/17/09 NA NB Green Time Red Time200190 Calc Offset180 35s+19=54s170160150140130120110100 90 80 70 60 50 40 30 20 10 00:00:00.0 3:00:00.0 6:00:00.0 9:00:00.0 12:00:00.0 15:00:00.0 18:00:00.0 21:00:00.0 0:00:00.0
  • N PPD After Change, Φ1 Φ2 Φ3 Φ4 Phase 6, 02/24/09 Φ5 Φ6 Φ7 Φ8 NA NB Green Time Red Time200190180 Time Change Was170 Implemented160150140130120110100 90 80 70 60 50 40 30 20 10 00:00:00.0 3:00:00.0 6:00:00.0 9:00:00.0 12:00:00.0 15:00:00.0 18:00:00.0 21:00:00.0 0:00:00.0
  • N PPD Φ1 Φ2 Φ3 Φ4 Phase 6, 02/25/09 Φ5 Φ6 Φ7 Φ8 NA NB Green Time Red Time 200 190 54s offset ok 180 Subsequently Fixed with 60s cycle 170 160 150 140 130 120 110 100 90 80 70 6054s offset fails 50 40with 50s cycle 30 20 10 0 0:00:00.0 3:00:00.0 6:00:00.0 9:00:00.0 12:00:00.0 15:00:00.0 18:00:00.0 21:00:00.0 0:00:00.0
  • N PPD Φ1 Φ2 Φ3 Φ4 Phase 6, 03/07/09 Φ5 Φ6 Φ7 Φ8 NA NB Green Time Red Time200190180 Not set back to TOD170 after correction, fixed160150 3/09140130120110100 90 80 70 60 50 40 30 20 10 00:00:00.0 3:00:00.0 6:00:00.0 9:00:00.0 12:00:00.0 15:00:00.0 18:00:00.0 21:00:00.0 0:00:00.0
  • N PPD Φ1 Φ2 Φ3 Φ4 Phase 6, 03/20/09 Φ5 Φ6 Φ7 Φ8 NA NB Green Time Red Time200190180170160150140130120110100 90 80 70 60 50 40 30 20 10 00:00:00.0 3:00:00.0 6:00:00.0 9:00:00.0 12:00:00.0 15:00:00.0 18:00:00.0 21:00:00.0 0:00:00.0
  • Percent of Cycles with Ped Phases, Wednesday100% 2 4 Before (1/9/08) After (1/30/08)50% 0%100% 6 850% 0% 0:00 12:00 24:00 0:00 12:00 24:00
  • 24 Hour Counts by phase…dependent upon Cycle 60 P1 P2 P3 P4Vehicle Detections per Cycle 30 0 60 P6 P5 P7 P8 30 0 0:00 12:00 24:00 0:00 12:00 24:00 0:00 12:00 24:00 0:00 12:00 24:00 Time of Day 120
  • 24 Hour Green Time by phase 90 P1 P2 P3 P4 45Green Time (sec) 0 90 P6 P5 P7 P8 45 0 0:00 12:00 24:00 0:00 12:00 24:00 0:00 12:00 24:00 0:00 12:00 24:00 Time of Day ga cactual s C 121
  • V/C Ratios by Phase, 24 Hours P1 P2 P3 P4 1.0Volume-to-Capacity Ratio 0.5 0.0 P6 P5 P7 P8 1.0 0.5 0.0 0:00 12:00 24:00 0:00 12:00 24:00 0:00 12:00 24:00 0:00 12:00 24:00 Time of Day 122
  • 24-Hour Plot of Intersection Saturation Showing Critical v / s C X Path c ci i C L v /s c Critical Ratio of Volume to Saturation 1.2 C Cycle Length (s) 1234 L Lost Time (s) 1278 1.0 5634 5678 20 pt. Mov. Avg.Intersection Saturation 0.8 0.6 0.4 0.2 0.0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00 Time of Day 123
  • 24-Hour Plot of Intersection Saturation With Split Failures Indicated C Xc v/s ci 1.2 i C L Phase 1 Phase 2 1.0 Phase 3 Phase 4 Phase 5 Phase 6Intersection Saturation 0.8 Phase 7 Phase 8 20 pt. Mov. Avg. 0.6 0.4 0.2 0.0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00 Time of Day