Evaluation of Effectiveness of Automated Workzone Information Systems

Evaluation of Effectiveness of Automated Workzone Information Systems Lianyu Chu CCIT, University of California Berkeley Hee-Kyung Kim, Yonshik Chung, Will Recker University of California Irvine

OUTLINE • Introduction • Framework and Operation of CHIPS • Safety Effects • Diversion Effects • Driver Survey • Conclusion 2

Background n work zones have become one of source of traffic congestion Central Controller n n ITS AWIS Benefits Traffic Sensors Changeable Message Signs provide traffic information to travelers potentially: -> increase safety -> improve the efficiency of traffic system 3

Background n Example of AWIS Ø ADAPTIR Ø CHIPS Ø Smart Zone Ø TIPS n Evaluation studies Ø Most studies: system functionality and reliability Ø Few studies: effectiveness of AWISs 4

Objectives & approach n Evaluation of CHIPS Ø Ø n Developed by ASTI Deployed in southern California focus: effectiveness Ø Ø Diversion effects Ø n Safety effects Drivers’ acceptance Approach: before and after study 5

OUTLINE • Introduction • Framework and Operation of CHIPS Ø System Structure Ø Study Area Ø System Setup • Safety Effects • Diversion Effects • Driver Survey • Conclusion 6

System Structure 7

Study Area n Site Location Ø City of Santa Clarita, 20 miles north of Los Angeles, on freeway I-5 Ø I-5: 4 -lane freeway with the closure of one lane on the median side Construction zone: 1. 5 miles long Ø Parallel route: the Old Road Ø n System Configuration - 3 RTMSs - 5 PCMSs - 3 CCTV cameras 8

System Setup Queue Detector Scenario CMS Combo Message RTMS-1 RTMS-2 RTMS-3 PCMS-1 PCMS-2 PCMS-3 PCMS-4 SBS 01 F F F CMB 01 SBS 02 T F F CMB 02 CMB 03 CMB 05 SBS 03 T T F CMB 06 CMB 07 CMB 03 CMB 10 SBS 04 T T T CMB 06 CMB 07 CMB 08 CMB 09 PCMS-5 CMB 11 T = Queue being detected, F = No queue being detected n Scenario SBS 04: all three RTMSs have congestion, the following messages are shown on PCMSs: Ø CMB 06 : SOUTH 5/TRAFFIC/JAMMED, AUTOS/USE NEXT/EXIT Ø CMB 07 : JAMMED/TO MAGIC/MOUNTAIN, EXPECT/10 MIN/DELAY Ø CMB 08 : JAMMED/TO MAGIC/MOUNTAIN, EXPECT/15 MIN/DELAY Ø CMB 09 : TRAFFIC JAMMED TO MAGIC MTN, AVOID DELAY USE NEXT EXIT Ø CMB 11: SOUTH 5 ALTERNAT ROUTE, AUTOS USE NEXT 2 EXITS 9

OUTLINE • Introduction • Framework and Operation of CHIPS • Safety Effects Ø Data Collection Ø Traffic Throughput Ø Travel Speed • Diversion Effects • Driver Survey • Conclusion 10

Data Collection n Collection locations Ø RTMS-1: 0. 15 mile before construction Ø RTMS-2: 1. 19 miles before construction n Collection time Ø Before scenario : Aug. 17 th, 2003 Ø After scenario : Sep. 1 st , 2003 n Collection Methods Ø Jamar DB-100 counters and Bushnell Speed Guns 11

Traffic Volume Variance of traffic volume based on 1 -min data n Total Lane 1 Lane 2 Lane 3 Lane 4 RTMS-1 Before After 44. 9 11. 6 25. 4 6. 5 11. 6 6. 5 15. 7 6. 3 3. 4 4. 5 7. 8 7. 6 RTMS-2 Before After 37. 4 31. 2 4. 4 4. 2 2. 8 4. 6 3. 9 4. 3 1. 8 2. 4 Difference between before and after values is significant (90% confidence level) Ø With the grouped traffic data, the difference of variance was significant at RTMS-1, which means that the variance of the after scenario was statistically smaller than that of the before scenario Ø With lane-based traffic data, the significant differences of variances were found for lane 1 and lane 2 at RTMS-1 12

Speed Mean and Variance # of Samples Sample Mean Standard Deviation Sample Variance RTMS-1 Before After 979 970 29. 9 30. 6 8. 9 7. 1 80. 0 50. 2 RTMS-2 Before After 1, 186 993 21. 2 21. 0 5. 7 32. 4 13. 5 Difference between before and after values is significant (90% confidence level) RTMS-1 RTMS-2 13

OUTLINE • Introduction Old Road I-5 • Framework and Operation of CHIPS • Safety Effects Lake Hughes Off-ramp • Diversion Effects Ø Data Collection Ø Calculation of Diversion Ø Diversion Estimation Ø Travel Time Analysis Hasley Canyon Off-ramp SR-126 Rye Canyon Off-ramp Magic Mountain On-ramp • Driver Survey Valencia On-ramp • Conclusion Old Road I-5 14

Data Collection n Collection Methods Ø I-5 mainline traffic volume : Pe. MS database Ø Off-ramp traffic volume : Tube counter n Collection Periods Ø Before scenario : May 13 th ~ May 18 th, 2003 May 18 th Ø After scenario : Independence Holiday weekend (June 30 th ~ July 7 th, 2003) July 6 th Labor Holiday weekend (Aug. 30 th ~ Sep. 2 nd, 2003) Sep. 1 st 15

Calculation of Diversion n Proportion-based method Ø Ø I-5 S Voff Old road Diversion rate = Pa - Pb = Ø V Proportion Voff P= V Voffa Va Voffb Vb a : after scenario b : before scenario Diversion traffic volume Vd = Va 16

Diversion Estimation n Hasley Canyon off-ramp traffic proportions 17

Diversion Estimation of diversion traffic volume Based on Caltran’s traffic report regarding Maximum Delay On July 6 th 15: 30 ~ 17: 30 On Sep. 1 st 17: 30 ~ 20: 00 18

Travel Time Analysis n Comparison of travel times - July 6 th , 2003 by GPS-based probe vehicles survey 19

Driver Survey § Method : Postcard-based survey § Location : Lake Hughes and Hasley Canyon off-ramp § Date : 1: 40~4: 30 PM, Sunday, July 6 th , 2003 § Response rate : 25% (100/400) 20

Driver Survey § Did the traffic signs influence route choice? Ø Yes : 78% of people who saw the PCMS message § Why did you get off the I-5 south? Ø 73% : avoid traffic Ø 22% : buy gas and foods Ø 5% : arrived at destination § Did you find these signs useful? (check all that apply) Ø 70% : useful for providing information Ø 63% : useful for taking alternative routes Ø 53% : useful for avoiding delay Ø 48% : useful for reducing anxiety Ø 9% : NOT useful 21

Conclusion Three aspects of effectiveness studies were conducted, including traffic diversion, safety effects, and responses from travelers n The results of these studies showed that: n Ø Obvious diversion were observed on two evaluation dates, July 6 th and September 1 st Ø Based on the study of the effects of traffic flow, the driving environment after the use of CHIPS seemed safer Ø Positive responses about the system were obtained based on driver surveys. 22

Conclusion • The safety has been enhanced – Stable traffic condition (speed and volume variance) • Network performance improved – 12% of diversion was observed – Alternative was still faster than mainline • Driver response – 70% of drivers expressed the system to be useful • Direct quantification was not made, but found positive effectiveness of the system. 23