Scanned by CamScanner Scanned by CamScanner Scanned by CamScanner Scanned by CamScanner Scanned by CamScanner Scanned by CamScanner Scanned by CamScanner Transportation Safety Engineering CE 599 Lecture I Mesfin Lakew, Ph.D., CPM BACKGROUND Globally: • 1.3 million people die per year as a result of traffic crashes •  3,287 death/day • 20-50 million get injured/disabled per year • Crash accounts to 2.2% of all deaths worldwide • Ranks 9th leading cause of death • Cost $518 Billion per year •  1 – 2% of countries GDP Cont’d BACKGROUND U.S.A. • • • • • • • 33,000 people die per year 2.25 million get injured/disabled Cost $230.6 billion/year  $25.6 billion in travel delay  $61 billion in lost market productivity  $59 billion in property damage  $32.6 billion in medical expenses FEDERAL PROGRAMS • Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFET-LU), signed into Law on August 10, 2005, • Highway Safety Improvement Program (HSIP) became a core Federal-aid program • Law included a guaranteed funding for highways, highway safety, and public transportation totaling $244.1 billion • HSIP received an average of $1,279 million each year for the following years. • Following the success of HSIP, Moving America in the 21st Century (MAP-21) was signed into law on July 6, 2012 by President Obama • The 1st long-term Highway Authorization program since SAFET-LU, 2005. MAP-21 allocates average annual funding of $2.4 billion to safety. FEDERAL PROGRAMS Before SAFETEA-LU 2005 • ISTEA 1991 - 1997 • TEA-21 1998 – 2003 Objective • Reduce the number and severity of crashes. Evolution in Safety Analysis Descriptive • traditional • include methods such as frequency, rate and EPDO • focus is on summarizing History of crash occurrence, type, severity at a location Predictive • quantitative • focus is on predicting expected number and severity at a site with similar volume, geometry, and operational characteristics. • existing condition, future, compare among alternative designs Human Factor • Human science such as psychology, physiology an kinesiology are helpful to design safe and effective systems. Goal is: • Understanding effects of human factor to reduce the probability & consequence of human error by designing systems with respect to human characteristics and limitations. • Driver error is a significant contributing factor in most crashes. • Errors can be due to: – – – – Judgment concerning speed, curve negotiation, ….. Inattentiveness Weariness Overloaded by information processing required to carry out multiple tasks simultaneously – Relying on previous knowledge results in mistakes when expectations are not met – Deliberate violation of traffic control devices and laws • Frequent mistakes because of human physical, perceptual, and cognitive limitations. Contributing Factors Driving Sub-Tasks Driving involves many sub-tasks and some of them need to be performed simultaneously. The three major sub-tasks: • Control - keeping the vehicle at a desired speed and heading within the lane • Guidance - interacting with other vehicles by following traffic control signs, markings, signals etc. • Navigation - follow path using guide signs, and landmarks (most complex sub-task). Driver Characteristics and Limitations Basic driver capabilities & limitations in performing driving tasks which can influence Safety: Attention and Information Processing • Ability to process info. Is limited • can create difficulties since driving requires attention between tasks • Overload of info. can result in neglect of critical piece of information Driver Characteristics and Limitations Traffic condition and operational situation can overload the driver. Therefore, Roadway design considerations for reducing workload include: • present info. in consistent manner • present info. sequentially • provide clues to help drivers prioritize the most important info. to assist in reducing workload by shedding extraneous tasks. Driver Characteristics and Limitations Designing roadway environments in accordance with driver expectation accommodate human info. processing limitation and help develop short and long term expectations. • Long-term: – Freeway exits on right-hand side – At major and minor road intersection, stop control on the minor rd. – Be on left lane to make left on approaching intersection • Short-term: – After driving a few miles on gently winding roadway, upcoming curves will continue to be gentle – On synchronized signalized arterial, drivers do not anticipate a location that operates different – Driver expect high speed to continue on road ahead similar to the one before Vision • Driver information use is 90% visual • Important aspects of vision are: – Visual Activity - ability to see detail at distance • Important for guidance and navigation tasks – Contrast Sensitivity – ability to detect slight differences in luminance (brightness of light) between an object and its background • Recognized as having a greater Impact on crash occurrence Vision – Peripheral Vision – ability to detect objects that are outside of the area most accurate vision within the eye. – Visual field is large 55 & 70 degrees above and below horizontal respectively & 90 degrees left and right. – However, accurate vision is very small area2 to 4 degrees from the focal point. • Targets detected in peripheral include: – Vehicle in intersecting path – Pedestrians, signs and signals – Movement in Depth – ability to estimate speed of another vehicle by the change of visual angle of the vehicle created at the eye. – Visual Search - ability to search the rapidly changing road scene to collect road info. Perception-Reaction Time (PRT) • Includes time to detect target, process info., decide on a response and initiate a reaction. • Commonly used 1.5 or 2.5 seconds to accommodate majority drivers. • Components of PRT: – Detection – Decision – Response Speed Choice • Speed limits influence driver speed • However, drivers select speed using perceptual and “road message” cues. • Perceptual cues come from – Peripheral vision – Streaming of information – Noise level – Speed adaptation – adjusting to lower speed after long Positive Guidance • Knowledge of human limitations in info. processing and human reliance on expectation to compensate for those limitations in info. processing , led to the “positive guidance” approach to highway design  lead to increase the likelihood of drivers to situation and information correctly and quickly. • Design that conform to long-term expectancies reduce driver chance of error – No traffic signals on freeways Positive Guidance – With Traffic Control devices, sample positive guidance approach emphasis : • Primacy - placement of signs according to importance of info. • Spreading – where all signs cannot put on one sign, spread the signs in small chunks • Coding - color and shape • Redundancy – display same thing in more than one way. • Stop sign - unique shape and message • No Passing – sign and pavement markings SAFETY FUNDAMENTALS GOAL is to provide background info : • prior to applying predictive method. • Use Crash Modification Factors • Apply Evaluation Method SAFETY FUNDAMENTALS • Difference between Objective and Subjective Safety • Definitions of Crash and related terms • Recognition that crash are rare and random events • Contributing factors and strategies Difference between Objective and Subjective Safety • Objective Safety – quantitative measures independent of the observer. • Subjective Safety – perception how safe an individual feels; vary between observers DEFINITIONS • Crash Frequency – No. of crashes per year • Crash Estimation – methodology used to forecast or predict crash frequency of: – An existing roadway for existing conditions during a past or future period – An existing roadway for alternative conditions during past or future period – A new roadway for given conditions for a future period DEFINITIONS • Predictive Method – a methodology used to estimate the “expected average crash frequency of a site, facility, or roadway under given geometry and traffic volume in a given period (years). • Expected Average Crash Frequency - used to describe the estimate of long-term average crash frequency DEFINITIONS • Crash Severity - defined using KABCO scale  possible differences between jurisdictions. • Crash Evaluation – determining effectiveness of a particular treatment or program. • Crash – rare  represent only small proportions of total events occurring in transportation system. – Random  occur as a function of a set of events influenced by several other factors; which are partly deterministic (can be controlled) and partly stochastic (random and unpredictable) • Note: as crashes are random, observed crash frequencies fluctuate over time and short period observations are not reliable indicator of crash frequencies expected under same condition for a long period of time. Can be done only if all conditions are controlled  rarely possible. CONTRIBUTING FACTORS Major Categories: • Human – age, judgment, driving skill, fatigue, …. • Vehicle – design, manufacture, maintenance • Roadway/Environment – geometry, alignment, traffic control devices, friction Haddon Matrix A framework relating series of crash events to contributing factors. • Helps create order when determining which contributing factor influence a crash and period. • Leads to identifying appropriate crash reduction strategies. • Reduction in no. and severity achieved thru changes in – human behavior – roadway condition – design and maintenance technology for both vehicle and roadway • Provision of medical treatment, post crash rehabilitation, . • Exposure to travel or level of transportation demand Data For Crash Estimation Data needed for Crash Analysis • Crash Data – overall characteristic of the crash location, date, type, severity • Facility Data – classification, no. of lanes, …. • Traffic Volume - AADT, ADT, VMT,… Limitation of Observed Crash Data Accuracy Can introduce bias and affect crash estimation reliability • Data quality and accuracy • Crash reporting thresholds and the frequencyseverity indeterminacy – Not all crashes correctly reported and severe crashes reported more reliably which creates frequencyseverity indeterminacy  creates difficulty in determining if change in reported crash is due to actual changes or a shift in severity proportion or a mixture of the two. • Differences in data collection methods and definitions used by jurisdictions Homework 1) Prepare a short write up regarding the role and responsibilities each of the following organizations in relation to Safety. a) NHTSA b) FHWA c) ITE d) AASHTO e) TRB 2) Compile Safety Facts, select one area and prepare a write-up to present to the class. (10-15minutes). TRANSPORTATION SAFETY ENGINEERING CE – 599 Lecture - II ROADWAY SAFETY MANAGEMENT • Process includes: – Network Screening – Diagnosis – Select Countermeasure – Economic Appraisal – Projects Prioritization – Safety Effectiveness Evaluation ROADWAY SAFETY MANAGEMENT • Process includes: – Network Screening – Diagnosis – Select Countermeasure – Economic Appraisal – Projects Prioritization – Safety Effectiveness Evaluation Network Screening • Is the first step in Roadway Safety Management • Is a process for reviewing transportation network to identify and rank site that can realize reduction in crash frequency thru the implementation of a countermeasure / program (increase enforcement). • Countermeasure ? Network Screening • Identified locations are further studied to: – Find causation – Crash patterns – Determine appropriate countermeasures • Used to formulate and implement policy – Such as prioritizing the replacement of nonstandard guardrail state wide for sites with high no. of run-off –the-road crashes. Network Screening • Today’s Class will cover – Steps for Network Screening Process – Performance Measures – Methods for conducting Network Screening Network Screening Process • Five major steps 1. Establishing Focus 2. Identification of Network and Establishing Reference 3. Selection of Performance Measures 4. Selection of Screening Methods 5. Screening and Evaluation of Results Major Steps: 1) Establish Focus – identify intended outcome of analysis • This will influence data needs • Selection of performance measures • Screening method that can be applied 2) Identify Network and Establish Reference Populations • Specify type of site being screened (intersection, segment, …) • Identify grouping of similar sites (signalized, four leg, two lane rural, etc.) Major Steps Cont’d • Characteristics used to establish reference population for Intersection: • • • • • Area Type (urban, rural,…) Functional Classification Traffic Control Traffic Volume Geometry/configuration Major Steps Cont’d • Characteristics used to establish reference population for Segments: – Functional Class – No. of lanes per direction – Traffic Volume – Median Type, width – Access density (driveway and intersection spacing) – Operating Speed – Adjacent land use Major Steps Cont’d 3) Select Performance Measures – Select as a function of the focus • Key consideration in selecting performance measures are: • Data availability – Typical Data Required » Facility information » Crash data » Traffic volume » Performance measures (some cases) Key consideration in selecting performance measures are: • Regression to the mean bias – Failure to account RTM introduces the potential for “RTM bias” » Crash frequency fluctuate » Significant variance in results between short and long term average freq. » Randomness indicate that short-term frequencies are not good estimator of long-term crash frequencies » A period with high frequency is statistically probable to be followed by lower crash frequency the following period (vice versa) and this tendency is known as Regression to the mean (RTM) Establish Thresholds –Threshold provides reference point for comparison –Method for determining depends on performance measure selected. –Can be subjectively assumed value or calculated. Performance measures • Each has its own strength and limitations – Average Crash Frequency – site with the greatest no. of crashes of a particular crash severity or type in a given time is given the highest rank. – Crash Rate – normalizes the frequency of crashes with the exposure. – EPDO – average crash frequency performance measure assigns weighting factors to crashes by severity to develop combined frequency and severity score. – Relative Severity Index - monetary crash costs are assigned to each crash type and the total cost of all crashes is calculated for each site and compared with average crash cost for each site. – Critical rate – observed crash rate at each site is compared to a calculated critical crash rate that is unique to each site. • Etc. Screening • Identifying location within the roadway segment or ramp that most likely benefit from a countermeasure. • The location within a segment that shows the most potential for improvement is used to specify the critical crash frequency of the entire segment for further investigation. Screening • Identifying location within the roadway segment or ramp that most likely benefit from a countermeasure. • The location within a segment that shows the most potential for improvement is used to specify the critical crash frequency of the entire segment for further investigation. Screening and Evaluation • Identifying location within the roadway segment or ramp that most likely benefit from a countermeasure. • The location within a segment that shows the most potential for improvement is used to specify the critical crash frequency of the entire segment for further investigation. 4) Select Screening Method • Three methods –Ranking –Sliding window –Peak searching • Sliding window and peak searching can be used to identify the segment which is more likely to benefit from the countermeasure. • Simple ranking method can also be used, even though performance measures are calculated for the entire Sliding Window • A window of specified length conceptually moved along the road segment from beginning to end in increments of specified size. Peak Search Method • Each individual roadway segment is subdivided into windows of similar length, potentially growing incrementally to the length of the roadway segment Simple Ranking Method • Applied to Nodes and Segments • Performance Measure calculated for all sites • Result ordered from high to low. • Simplicity is its strength • However, results are not reliable for segments as other screening method. Performance measures • Three categories –Segments use sliding window or peak searching • Requires identifying location within the segment that most likely to benefit from the countermeasure • This location is used to specify the critical crash frequency for the entire segment and make it more easier and efficient selecting countermeasure. –Nodes (intersections, ramps) use simple ranking –Facilities (combination of nodes and segments) use combination of nodes and segments screening method. Performance measures 5) Screen and Evaluate Results –Final step in the process is to conduct screening analysis and evaluate needs EXAMPLE • Performance Measure Calculations • Screening Method –Ranking –Sliding window –Peak searching Know: • CMF • Predictive model  N predicted = • Note – different model for different facility types • Advantage of predictive method: • Estimation of expected crash freq. using only observed or estimated crash freq. results in a reasonable estimate. • However, using both from a predictive model improve the statistical reliability. • Several stat. models do exist to compensate for the bias resulting from RTM. • EB applied here and uses a weighting factor which is a function of SPF. • EB is a method used to combine crash freq. data for a site with predicted data from many similar sites to estimate the expected crash freq. Facts • Effects of multiple CMFs applied a site are independent of each other. • CMFs are estimate values of change expected. • Standard error  serve as a measure of the reliability of the estimate. • Standard errors are also used to calculate the confidence interval for the estimated change. Predictive Method • Has 18 step procedure • Used to estimate avg. crash freq. • By: total crashes, severity, type for an intersection, segment, or network • Facility – contiguous set of intersections and segments  sites • Network  group of contiguous facilities. Predictive Method • Used to estimate expected avg. crash freq. of individual site. • Cumulative of all sites  estimate for facility/network • Estimate – good for given time period where no change in physical and/or data change. • Estimate relies on regression models developed from observed crash data for a number of similar sites. Steps for Predictive Method 1. Define the limits for the site - (a site, group of sites, or a corridor for network screening). 2. Define the period/years - (past or future) – Years of interest are determined by: • • • Availability of observed or forecast AADTs Observed crash data Geometric design data – Predictive method for past or future period depends on purpose of study. 3. Determine availability AADT and the availability of observed crash data to determine if EB method is applicable. – If data not available for all years  use interpolation/extrapolation as appropriate. – EB method is applicable only when reliable, observed crash data are available. – At least two years of observed crash freq. data are desirable to apply EB method. 4. Determine geometric design features, traffic control features, and site characteristics for all study sites. 5. Divide the roadway network under consideration into sites (individual segments, intersections. 6. Assign observed crashes to individual sites, if applicable. 7. Select the first or next individual site; if no site to be evaluated, go to step 15. 8. For the selected site, select the first or next year; if no more years to be evaluated, go to step 15. 9. Apply the appropriate SPF for the site’s facility type and traffic control features. 10. Multiply the result obtained in 9 by the appropriate CMFs to adjust to site-specific geometric design traffic control features. Multiply the result obtained in step 10 by the appropriate calibration factor. 12. If another year to be evaluated, return to step 8, otherwise go to step 13. 13. Apply site-specific EB (if applicable). 14. If another site to be evaluated, return to step 7, otherwise, proceed to step 15. 11. Apply project level EB (if the site-specific method is not applicable). 16. Sum all sites and years to estimate total crashes or avg. crash freq. for the network. 17. Determine if alternate design, treatment, or forecast AADT to be evaluated. 18. Evaluate and compare results. 15. Homework II 1) 2) 2) Compile list of Crash Types, causation, and possible countermeasures for each. List and provide strength and limitations of four (4) performance measures of your choice. Select two (2) and prepare a presentation for class on the process how they are calculated (performance measures). TRANSPORTATION SAFETY ENGINEERING CE – 599 Lecture III ROADWAY SAFETY MANAGEMENT • Process includes: – Network Screening – Last Class – Diagnosis – Today’s Class – Select Countermeasure – Economic Appraisal – Projects Prioritization – Safety Effectiveness Evaluation Last Class Network Screening Process • Five major steps 1. Establishing Focus 2. Identification of Network and Establishing Reference 3. Selection of Performance Measures 4. Selection of Screening Methods 5. Screening and Evaluation of Results DIAGNOSIS • Second step in Roadway Safety Management Process. • Provides an understanding of: – Crash Patterns – Past Studies – Physical Characteristics • INTENDED OUTCOME –Identification of the causes of the crashes –Potential safety concerns that can be further evaluated Diagnosis Procedure • Three Steps 1) Safety Data Review 2) Assess Supporting Documentation 3) Assess Field Condition Diagnosis Procedure 1) Safety Data Review – Review crash types, severities and environmental conditions • develop summary statistics for pattern identification – Pattern related to: » Time of day » Direction of travel prior to crashes » Weather conditions » Driver behavior Three to five year data is suggested to improve the reliability of the diagnosis • Safety Data Review Considers –Descriptive Statistics of Crash Conditions –Crash Locations • Descriptive Crash Statistics –Summarizes crash data into three categories: 1) Information about the crash 2) The vehicle in the crash 3) The people in the crash • Summaries Included: – Crash Identifiers – date, day of week, time of day – Crash Type – define by police officer at scene • • • • • • • • • • Turning Angle Rear-end Run-off-the-Road Sideswipe Head-on Fixed Object Animal Out-of-control Work zone Crash Severity – defined by KABCO –an injury scale defined by the National Safety Council. Acronym derived from – Fatal Injury (K) - Incapacitating Injury (A) - Non-Incapacitating Injury (B) - Possible Injury (C) - No Injury (O) Contributing Circumstances Parties Involved – vehicle only, ped. and vehicle,…… Road Condition – dry, wet, ice, ….. Lighting Condition – dawn, daylight, dusk, …. Weather condition – clear, cloudy, fog, ….. Impairments of parties involved – alcohol, drugs, fatigue Tabular Summaries, Bar Charts, pie charts are used to present descriptive crash statistics and make patterns visible. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% C B A K Hit Object Left Turn Angle Rear End Sequence of events Direction of travel Location of parties involved 2) Review Crash Locations – Review past studies and plans – Identify known issues, opportunities and constraints –Summarize using three tools: • Collision Diagram • Conditions Diagram • Crash Mapping • Collision Diagram – Two dimensional plan view representation of the crashes – Simplifies visualization of crash patterns – Clusters or patterns of crashes by collision type – May not reflect a quantitative or statistically reliable assessment of site trends – Provide indication if pattern exists • Condition Diagram – Purpose is to develop a visual site overview that can be related to the collision diagram findings. – The two diagrams can be overlaid to further relate crashes to the roadway conditions. – Plan view of the many characteristics of the site and include: – Roadway • • • • • • Lane configuration and traffic control Pedestrian, bicycle, and transit facilities Presence of roadway medians Landscaping Shoulder or type of curb and gutter Location of utilities (fire hydrant, light pole, …) • Pavement Conditions – Location of potholes, ponding, rut, …. • Land Uses – Type of adjacent land use (school, commercial, …) – Driveway access points serving these land uses. • Crash Mapping – Is evaluating crash locations and trends using GIS – Following are some crash analysis techniques and advantages of using GIS • Scanned police reports and video logs for each location can be related to GIS • Data analyses can integrate crash data (location, time, …) with other database info. such as presence of school, … • Crash database can be queried to report crash clusters; ie. Crashes within a specific distance of each other or land use. • Crash frequency or density can be evaluated along a corridor to provide indication of pattern. • Data entry quality control checks can be conducted easily and if necessary, correction can be made directly in the database. 2) Assess Supporting Documentation - Is the second step in the overall diagnosis of site - Goal is to obtain documented info. and people testimonies. - Obtain historical perspective of the site - For instance crash data may reveal frequency of left turn crash at a signalized intersection increased; associated documents may show a widening project completed at that time which may have led to the increased observed frequency due to travel speed. • Useful supporting documentations include: – Current traffic volume for all travel modes – As-built drawings – Relevant design criteria and pertinent guidelines – Relevant photos or video pictures – Inventory of field conditions – Maintenance logs. – Recent traffic operation studies – Land use mapping and traffic access control characteristics – Etc. 3) Assess Field Condition - Serve to validate safety concerns identified by review of crash data and/or supporting documentation – Visit site to review and observe multimodal transportation facilities and services in the area. – Firsthand site is obtained to help understand motorized and non-motorized travel. – A comprehensive field assessment involves travel through the site from all possible directions and modes. • Sample useful site review considerations – Roadway and Roadside Characteristics • Signing and striping • Posted speeds • Overhead lighting • Pavement Condition • Landscape Condition • Sight Distance • Shoulder Width • Roadside Furniture • Geometric Design • Sample useful site review considerations –Traffic Condition • Types of Facility • Travel Condition • Adequate queue storage • Excessive vehicular speed • Traffic Control • Adequate traffic signal clearance time • Sample useful site review considerations – Roadway Consistency – cross section,.. – Land Uses –Weather Conditions –Traveler Behavior • Drivers • Bicyclists • Pedestrians • Sample useful site review considerations –Evidence of Problems such as: • Broken glass • Skid marks • Damaged signs • Damaged guard rails • Damaged road furniture • Damaged landscape treatments • RSA – different, conducted by independent interdisciplinary audit team experts on existing or future facility. • IDENTIFY CONCERNS – Once field assessment, data review, supporting document assessment is completed – information compiled to identify specific crash patterns that can be addressed by a countermeasure. – Comparing observations may lead to observations that would not have otherwise been identified. – EX. Crash data review showed high crash frequency at a particular approach to an intersection; and field investigation showed potential site distance constraints; these two piece of info. may be related and may warrant further consideration. Alternatively, document assessment may reveal signal timing change in response to capacity concern. RESULT OF DIAGNOSIS • In General –Sets of pie charts summarizing crash data –Collision Diagram –Condition Diagram –Written assessment and summary of the diagnosis