Crash Reduction Factors for Countermeasures Issue Briefs and
Crash Reduction Factors for Countermeasures Issue Briefs and Desktop Reference Maurice Masliah, Ph. D. Senior Researcher i. TRANS Consulting Prepared by i. TRANS Consulting (905) 882 -4100 ext 5295 mmasliah@itransconsulting. com December 2007
Updated Issue Briefs Intersections Traffic Signals April 2004 September 2007 December 2007 FHWA 2
New Issue Briefs Pedestrian Roadway Departure December 2007 FHWA 3
New Product: Desktop Reference December 2007 FHWA 4
What is a CRF? • Crash Reduction Factor • Percentage crash reduction that might be expected after implementing a countermeasure – CRF of 25 means a 25% reduction in crashes – CRF of -20 means a 20% increase in crashes December 2007 FHWA 5
Issue Briefs: Purpose • One CRF for each countermeasure • The “best” available CRF (accuracy and precision checks) • Additional information relevant to the CRF (traffic control, area type, etc. ) December 2007 FHWA 6
Desktop Reference: Purpose • Multiple CRFs for the same countermeasure (where available) • CRF functions • Additional information such as study type, number of observations, CRF range (high - low) December 2007 FHWA 7
Poll Question 1 How useful do you think the Issue Briefs will be to you? L K J ¶ Not very useful Somewhat useful Very useful Best thing since sliced bread December 2007 FHWA 8
Product Purpose Iowa DOT Tom Welch State Transportation Safety Engineer Iowa Dept of Transportation December 2007 FHWA 9
Roadway Departure Issue Brief • Purpose • CRFs • How to Use December 2007 FHWA 10
Roadway Departure CRF Example • CRF = 38(10)17 December 2007 FHWA 11
CRF Example Expected crashes without treatment CRF Expected crashes with treatment 38 Remove or relocate fixed objects outside the clear zone December 2007 FHWA 12
CRF Example • CRF = 38(10)17 Ø Bold type means a rigorous methodology was used to estimate the CRF, and the standard error is relatively small Ø (10 ) is the standard error for this CRF Ø 17 is the reference number* * details at the end of the Issue Brief December 2007 FHWA 13
CRF Confidence Interval • Estimation of confidence interval [38(10)17] 2 standard errors from the CRF 18% (38 - 2 X 10) and 58% (38 + 2 X 10) • Expected safety effect of the countermeasure: between 18% and 58% December 2007 FHWA 14
Poll Question 2 How many of you use CRFs in your work? • I use CRFs • I don’t but I think there are others in my agency who do • I don’t use and don’t know anyone in my agency who does use CRFs December 2007 FHWA 15
Example Application Douglas County Keary B. Lord Assistant Director/ Traffic Operations Division Manager December 2007 FHWA 16
Accuracy • Accuracy is the proximity of the estimate of the CRF to the true value • Two types of bias affect accuracy: 1. Regression to the mean bias 2. Traffic volume bias December 2007 FHWA 17
Precision • Precision is the degree to which repeated estimates of the CRF are similar to each other • Measure of precision: standard error of CRF December 2007 FHWA 18
Reliability • Reliability is defined here as the combination of precision and accuracy • CRFs that pass the reliability test are shown in bold font December 2007 FHWA 19
Reliability Selection Criteria and Thresholds December 2007 FHWA 20
Number of CRFs and Bold CRFs ** Very few published CRFs are considered reliable! December 2007 FHWA 21
Poll Question 3 Where do you get your CRFs from? • Internal documentation • FHWA Issue Briefs • Other published sources • Other (please use message board to explain) December 2007 FHWA 22
CRF Proper Use and Limitations Colorado DOT TBD – Bryan Allery/Jake Kononov Colorado Dept of Transportation December 2007 FHWA 23
Highway Safety Improvement Program (HSIP) • Planning • Implementation • Evaluation (3) A comparison of accident numbers, rates, and severity observed after the implementation of a highway safety improvement project with the accident numbers, rates, and severity expected if the improvement had not 2007 been made. FHWA December 24
HSIP & CRFs • Evaluation legislation specifies the information needed to produce CRFs • Our review of HSIPs has not found any states which tie their evaluation work with CRF generation • Potential to link tasks currently being conducted with improving knowledge of effectiveness of countermeasures December 2007 FHWA 25
Strategic Highway Safety Plan (SHSP) • CRFs should be part of the process of selecting countermeasures • Intersections as an emphasis area within a SHSP • Task team identifies left-turn crashes at intersections as an overrepresented crash type • What are appropriate countermeasures? December 2007 FHWA 26
Strategic Highway Safety Plan (SHSP) • Review of Intersection Issue Brief for left-turn crash CRFs – Provide protected left turn phase 16(2)19 – Install left-turn lane (physical channelization) 249 – Install cameras to detect red-light running 45(6)27 • Identify intersection locations and conduct a cost-benefit analysis December 2007 FHWA 27
Future Direction – What’s Next? • Accident Modification Factors and Functions (rather than CRFs) • Increase in number of studies that evaluate safety effects of countermeasures • More stringent criteria in adoption of AMFs • New tools December 2007 FHWA 28
What is an AMF? • Accident Modification Factor • Multiplier representing the expected change in crashes – CRF of 25 equals an AMF of 0. 75 – CRF of -20 equals an AMF of 1. 20 December 2007 FHWA 29
Accident Modification Functions • Today, CRFs / AMFs for most countermeasures are noted as constants • However, CRFs / AMFs are really functions of environmental variables: – Traffic volume – Traffic mix (trucks, pedestrians, bicyclists) – Road geometry – Operational conditions December 2007 FHWA 30
More Stringent Inclusion Criteria • Highway Safety Manual inclusion process: – Accuracy and precision of AMFs – Standard error of 0. 1 or less indicates AMF that is sufficiently accurate, precise, and stable – Companion AMFs with standard errors of 0. 2 - 0. 3 are also included • indicating the potential safety effects of the treatment on other facilities, or other crash types and severities December 2007 FHWA 31
Future Direction • Periodical Update of Issue Briefs and Desktop Reference • Search new information: – current research projects: http: //rip. trb. org/ – government-funded documents published: http: //www. ntis. gov/search/index. asp? – bibliographic database http: //ntlsearch. bts. gov/tris/index. do December 2007 FHWA 32
New Tools and New AMFs • Low Cost Safety Improvements Pooled Funds Study • Highway Safety Manual (HSM) • Safety. Analyst • Interactive Highway Safety Design Model (IHSDM) December 2007 FHWA 33
Thank You Questions? December 2007 FHWA 34
CREDITS The Technical Working Group (TWG) FHWA Clayton Chen Project Manager FHWA Marcee Allen Nick Artimovich Mark Doctor Kenneth Epstein Ray Krammes Jim Growney Dean Larsen Iowa DOT Tom Welch ITE Ed Stollof December 2007 FHWA Matt Lupes Mary Mc. Donough David Morena Fred Ranck Tamara Redmon Ed Rice Gabe Rousseau 35
- Slides: 35