Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA Home
Federal Highway Administration Research and Technology: Coordinating, Developing, and Delivering Highway Transportation Innovations

ELCSI-PFS: Overview | Seven Phases of ELCSI-PFS | Technical Advisory Committee Members | Technical Advisory Committee Meetings | Publications | Treatment Databases | NCHRP Project 17-25: Crash Reduction Factors forTraffic Engineering and Its Improvements |
NCHRP Project 17-35: Evaluation of Safety Strategies at Signalized Intersections

 

 

Evaluations of Low Cost Safety Improvements Pooled Fund Study

 

Low Cost Safety Improvements Pooled Fund Study Technical Advisory Committee Meeting Notes

Tuesday, June 19, 2007 and Wednesday, June 20, 2007
National Highway Institute

Day 1: Tuesday, June 19, 2007

Welcome, Introduction, and Objectives for Meeting

  • Carol Tan (FHWA) and Roya Amjadi (FHWA) greeted the Technical Advisory Committee (TAC) and Roya presented the outline of the meeting.

Simulation Studies

  • Tom Granda (FHWA) presented on the potential for simulation–based studies.
    • The question was posed of whether or not treatments may be evaluated from a pedestrian perspective. In particular, there was some concern that people with visual or hearing impairments may react differently to certain treatments (especially at intersections).
      • Tom Granda responded that the simulation depends on the situation, but treatments may be considered from either a driver or pedestrian perspective.
    • This discussion led to a brief discussion of newer cars and electric cars that are quieter than older cars. These cars may present a problem for pedestrians (particularly the hearing impaired) and future research needs to focus on this issue.
    • Tom Granda discussed the results of a simulation–based speed study of horizontal curve treatments and indicated that speeds were only reduced by 4 to 5 mph.
      • Bhagwant Persaud (Persaud and Lyon) commented that a 4 or 5 mph reduction is a dramatic reduction in speeds for horizontal curve treatments.

Results from First Four Evaluations

  • Bhagwant Persaud (Persaud and Lyon) provided a background presentation on the analytical basis for the Low Cost Study design.
    • Bhagwant Persaud discussed the use of the empirical Bayes (EB) approach for the Low Cost studies
    • A Technical Advisory Committee (TAC) member questioned whether the full Bayes (FB) or EB is better and if there are any reasons for using one over the other?
    • Tom Welch (IA) Iowa indicated that both provide good results.
    • Paul Jovanis (PSU) indicated that FB is typically used for situations where there are limited data; FB may require less data and might be a better option if data are limited. Paul also commented that it is uncertain whether the FB will give better results than the EB for large samples.
    • There was a general consensus that the FB procedure is more complex than the EB approach. FB still requires a significant amount of data and similar variables to the EB approach (e.g., ADT); therefore, this is not necessarily an option for reducing the required number of sites.
  • Bhagwant Persaud (Persaud and Lyon) presented the results of Stop Signs with Increased Retroreflectivity.
    • There was a question regarding the possibility of regression to the mean; a TAC member asked whether or not the sites were identified for treatment based on high crash locations. Bhagwant Persaud and Nancy Lefler (VHB) indicated that the strategy was implemented as a blanketed treatment in both States.
    • Gary Modi (PA) indicated that he was not sold on the STOP sign treatment based on the initial results. Gary said that the method does not appear to be justifiable based on limited safety funds. He said that strategies with a greater potential effect would be used before increasing the retroreflectivity.
      • Bhagwant Persaud indicated that he would explain the cost–effectiveness in the economic analysis section of his presentation.
    • Jana Lynott (AARP) asked about the effect of increased retroreflectivity on older drivers.
      • Bhagwant indicated that the effects of increased retroreflectivity may be more pronounced for older drivers, but there was not proper data to answer this question.
    • A question was asked regarding the type of upgrade (e.g., poor sign to good sign, differential in reflectivity between signs).
      • Nancy Lefler (VHB) indicated that the signs were upgraded from engineering grade to high type sign, but the differential between the old and new sign because the condition of the sign before the upgrade was not known.
    • Gary Modi (PA) brought–up another related issue. He indicated that there may be a problem with drivers stopping in the wrong location and therefore have limited sight distance. Gary advocated the need for a treatment to help drivers get closer to the roadway or stopping in the appropriate location to avoid clear zone issues.
      • David Engstrom (MN) supported Gary 's point and indicated that a similar situation occurs in Minnesota. The STOP sign may be located closer or further from the intersection than the stop bar due to other variables (e.g., obstructions).
    • Shawn Troy (NC) indicated that some States need to apply a crash modification factor to install a treatment. Results from the Stop Sign study may not be usable to States because the crash modification factors need to compare favorably to other projects for funding purposes. Although the State may recognize the potential benefits, they need to justify the cost with a limited budget.
    • A question was asked regarding the effectiveness of increased retroreflectivity at night. This question was considered during the analysis, but Bhagwant indicated that the night effect is hidden by the fact that we do not have enough crashes at night to detect a difference.
    • Tom Welch (IA) indicated that Iowa gives free signs to their districts so cost should not be an issue.
    • Pat Brady (FL) indicated that Florida implemented this strategy about 10 years ago and conducted a very simple before–after study. They did not notice any changes at the time of the treatment and did not report any significant reductions. Eventually, Florida decided to make this strategy a standard application.
    • John Smith (MS) indicated that there may be a problem using this strategy in urban areas because ambient lighting may wash away the brightness of the retroreflective stop sign.
    • Scott Jones (UT) indicated that Type IX sheeting has a longer life and produces a better benefit–cost ratio than Type III sheeting.
  • Raghavan Srinivasan (HSRC) presented the results of the flashing beacon evaluation.
    • A question was asked about the installation sites; how were sites selected for treatment in North Carolina and South Carolina?
      • Shawn Troy (NC) indicated that any reason for installing treatments was based on high accident locations. If this varied by State, then there may be a larger effect at certain locations.
      • Amelia Glisson (SC) could not recall the factors that went into installing the treatments.
    • Shawn Troy (NC) indicated that the flashers on the STOP signs were placed to keep people pacified. This was a temporary treatment while signals were being studied or designed.
    • Shawn Troy (NC) indicated that North Carolina has a few places where they apply the treatment to all–way stop–control (AWSC) locations, but this study focused on two–way stop–control (TWSC).
    • John Smith (MS) indicated that enforcement may help to maintain the long term effect of the treatment after installation.
  • Frank Gross (VHB) presented the results of the STOP AHEAD pavement marking evaluation.
    • Joseph Santos (FL) asked what distance was used to classify crashes as "intersection crashes".
      • A 250' radius was used to classify intersection crashes, which was consistent with the other studies.
    • Gary Modi (PA) added a comment regarding the effect of pavement markings. His general conclusion was that drivers started paying more attention once they started implementing STOP AHEAD pavement markings.
    • Frank Gross added that drivers may change their behavior based on general warning messages.
    • Jeff Wolfe (KY) indicated that any time they do something different or unusual they find larger reductions than standard treatments; there may be some novelty effect.
    • Gary Modi (PA) mentioned that Pennsylvania has tried a similar treatment (SLOW and XX MPH), which had a tremendous effect. Eventually, this effect wore off. There was a brief discussion of enforcement to maintain the effectiveness.
    • David Engstrom (MN) commented that the STOP AHEAD pavement markings came from one rural county in Minnesota. He asked if all the data were used in the analysis.
      • Nancy Lefler (VHB) responded that only those sites with sufficient traffic volumes were included in the analysis.
      • David added that during a recent road safety audit (RSA), the consultant noted that trunk highway intersections were different than non–trunk highways. MnDOT came up with a combination of pavement markings for trunk highway intersections to address the issue of drivers stopping in the wrong location or disregarding the STOP sign (e.g., STOP AHEAD pavement markings and sign, stop signs, stop bars with correct placement).
    • The conversation continued on the topic of stopping location and Gary Modi (PA) indicated that Pennsylvania has tried one more treatment (painting a broken line at the edge of the pavement) to more clearly designate the edge of the road (30 or 40 locations). This strategy was implemented to help drivers stop closer to the intersection where visibility is better. Gary noted that Pennsylvania is going to deploy this strategy at 200 intersections.
    • Brad Steckler (IN) asked if the reduction in impact over time related to lack of enforcement at the speed markings? The TAC discussed that the issue of time is a complicated issue.
    • Frank Gross noted that over time the visibility of these markings may deteriorate which could also contribute to the reduction in impact and added that proper maintenance is critical with this type of strategy.
    • Gary Modi (PA) commented that PennDOT looked at STOP AHEAD signs, which also lost value as time went by. He added that enforcement helped to maintain the effect of the treatment.
    • Scott Jones (UT) commented that they touch up their pavement messages twice a year in Utah. This comment was directed at the economic analysis where it was assumed that latex markings would last 2 years. States may have to apply a service life that is applicable to their State's practice when computing the benefit–cost ratio.
    • Cliff Reuer (SD) added a comment regarding the long term effects of treatments, indicating that locals obey them at first but do not continue to pay attention after the novelty wears–off. Many of these strategies, however, are focused on the unaware or unfamiliar driver.
  • Craig Lyon (Persaud and Lyon) presented the results of the two–way left–turn lanes for two–lane roads (TWLTL) evaluation.
    • Tom Welch (IA) commented that 4 to 3 lane conversions were broken down by crash type in Iowa. He also mentioned that a report was released by FHWA that notes the risk of this type of conversion. Tom requested that FHWA send a retraction on their previous publication so the public is better informed at public meetings.
    • Pat Brady (FL) indicated that the conversion depends on land use and conflict points. He also noted that this needs to be applied on a case–by–case basis and cannot be implemented blindly.
      • This issue was discussed by the TAC and it was noted that the conversion is probably a good idea for an interim strategy by re–striping; however, this is not recommended new construction because traffic volumes may change substantially and locations may need a different approach in the long term.
    • David Engstrom (MN) commented that Minnesota does not use this strategy on roads with posted speeds of 45 mph or greater. He added that the strategy may be okay on low volume roads, but not for high volume roads because people turning left onto the mainline may use the TWLTL as an acceleration lane.
    • Pat Brady (FL) commented that pedestrians need to be considered when applying TWLTLs. He added that ADT has reached capacity in some parts of Florida, which leads to insufficient gaps for left–turning vehicles; FLDOT has made it a priority to remove 5 to 4 lane conversions where it may become unsafe due to increases in volumes.
    • A discussion was held regarding the need to consider the differences in effect by ADT volumes. Specifically, is there a maximum volume where safety effects of TWLTLs fall–off? States would like to know the ADTs that were used in the study.

Two–way Left–turn Lane Experience in Iowa

  • Tom Welch (IA) gave a presentation on the use of TWLTLs in Iowa.
    • Pat Brady (FL) commented that a study was done by the access management team in Florida, but the installations were implemented for safety reasons rather than strictly operational. He noted that there was a publication from this study, but he needed to get the report. Pat also added that other districts have completed analyses with their own money, which may not make it to publication.
    • Brad Steckler (IN) commented that Indiana has mostly employed this strategy based on traffic flows. He wondered if we obtained any kind of data on left–turn volumes and if so, could we include it in the results. This generated some discussion about what conditions are important to the installation of TWLTLs. Brad also asked why a 2:1 benefit–cost ratio was used. It was decided that the 2:1 ratio was selected somewhat arbitrarily, but would allow for inflation and other factors. These installations will be built in the future and costs may increase over time so it helps to use a higher benefit–cost ratio to absorb these costs. It was also noted that there are many variations across States with respect to the benefit–cost ratio used.
    • A comment was made that driveway density is an important factor when installing or evaluating TWLTLs; however, it was added that the turning volume is also important and often not available for these access points.
  • Kim Eccles (VHB) led an open discussion about new strategies, strategies of interest, and general concerns.
    • Advance (transverse) rumble strips at stop sign locations were discussed as a potential for evaluation.
      • Tom Welch (IA) and Dave Engstrom (MN) indicated that they have a few locations where transverse rumble strips are installed, but there are some data issues.
    • One concern was raised regarding the target crashes for these strategies; do States code "ran stop sign" so we can analyze specific crash types?
      • Pat Brady (FL) indicated that the type of traffic control is very obvious and likely coded correctly in most cases; however, the contributing cause is a little more suspect in Florida. He added that all of the data gets coded in the master database.
      • James Montgomery (OK) indicated that Oklahoma codes specific intersection crashes as "fail to stop" and "fail to yield".
      • Stephen Read (VA) indicated that Virginia codes specific crashes as "disregard for stop sign" or stop light. In Virginia the type of violation is fairly good, but the type of control is more suspect.
      • Mike Doody (NY) commented that New York has two codes; type of control and driver violation. It was added that both data fields are suspect.
    • Kim Eccles posed a question regarding new strategies; are there any States that are interested or planning to implement new strategies?
      • David Engstrom (MN) indicated that Minnesota will be installing rumble stripes (edgeline). There would be a lot on 2–lane rural roads where the lane width will normally remain the same, but the lane width may be reduced to increase the shoulder for installation purposes.
      • J–turns were discussed as a strategy for reducing left–turn crashes. J–turns prohibit drivers from making a left hand turn, but instead, force the drivers to turn right, make a u–turn, and then proceed straight thru the intersection. It was noted that there was an NCHRP study on J–turns, but it was not very thorough with respect to safety.
      • Pat Brady (FL) indicated that Florida uses the edge line rumble strip as a "rain line" to give a vertical face for visibility. They are considering this strategy on the centerline "skip line" and trying to make it a standard for certain roads (e.g., 2–lane roads).
      • One State indicated that countdown pedestrian signals are now standard and there are 25 included in an analysis (behavioral rather than crash based).
      • Janice Benton (CA) indicated that they have shoulder rumble strip applications. She also added that urban freeway signing may be considered in the low cost study and California has a few applications of this treatment. Another strategy of interest in California is access to the HOV lanes (full vs. limited) and a study to evaluate the safety impacts.
      • Tom Welch (IA) indicated that Iowa is replacing old yellow signs with florescent yellow signs. Tom added that the old signs blend into the background (e.g., corn fields) and become difficult to see. This strategy is intended to focus on problem areas (e.g., sharp curves) and older drivers.
      • John Carey (CT) indicated that they have 38 installations of the florescent yellow signs from a few years ago.

Simulation Demonstration

  • The TAC went to the Turner–Fairbank Highway Research Center to tour the facilities and see the capabilities of the driving simulator. The TAC then returned to the National Highway Institute to discuss the potential of simulation–based studies in the evaluation of low cost strategies.
  • John Molino (SAIC) led a discussion on strategies that could possibly be used in the simulator. The discussion was broken into two parts. Part 1 focused on traffic calming treatments and Part 2 focused on rural, two–lane strategies. The discussion was very open and everyone was asked to provide their input.
    • A probing question was asked to generate some discussion; what is important and what is practical when considering the treatments?
    • Rick Pain (TRB) was curious about catering the experiments to older or younger drivers.
      • John Molino responded that there is the possibility to consider any populations in the simulator studies.
    • John Molino asked the States to volunteer specific locations where traffic calming strategies have been installed or are planned for the future.
      • Tom Welch (IA) is looking at traffic calming devices in small town locations. Particularly on short segments entering towns where traffic speeds are changing from high speed to low speed. Tom indicated that these locations are six to seven blocks in length.
      • John Smith (MS) commented on the drainage of the roadway.
    • John Molino noted that the simulation studies are a back–up plan when the crash data are not available or sufficient. Evaluations can not focus on older drivers or specific intersection crashes because the crash data are not detailed enough, but the simulator may be able to study specific populations. The simulator cannot study specific crash types though because reliable surrogates have not been identified.
    • Tom Welch (IA) indicated that the biggest problem for advance signs in Iowa is on high speed roads with multiple lanes and high traffic density. They need to provide drivers with advance warning so they can get into the appropriate lane.
    • There was a question regarding the capability of the simulator; can it simulate the rear–view mirror as well as other vehicles on the road? John Molino indicated that there is the capability to simulate both scenarios, but the funds are probably not adequate to get to this level of detail. The traffic may be more readily available and probably could be included at a reasonable cost.
    • A description of the general simulation study was provided by John Molino. He indicated that the total time in the experiment is about 1.5 hours with about 15 minutes per direction. He added that all ages have been considered for the simulation experiment.
    • Cliff Reuer (SD) raised a concern regarding the standards for the delineation used in the simulator studies; whose standards will be used to define the delineation?
    • John Molino was discussing the simulator study and reiterated that, hopefully, a systems engineering approach will be apparent in this study.
      • John Molino is looking for urban/suburban locations to test the strategies using the simulator.
      • David Engstrom (MN) indicated that he may know of a town in the suburbs in Minnesota where an old town has a lot of truck traffic and there have been many aspects of traffic calming proposed. The location is ¾ to 1 mile in length.
      • Stephen Read (VA) commented that there may be a high speed scenario in the suburbs in Northern Virginia.
      • There was quite a bit of discussion about the different scenarios that could be included in the simulation study. John Molino was trying to keep the number of scenarios to just 2, but there was interest in 3 scenarios. It may be necessary to prepare 3 scenarios for the simulation but there could be 2 scenarios within a single simulation; two of the scenarios are combined in different portions of the simulator. The combined portion of the simulation could have on a scenario with advance street name signs where the driver is on a multiple lane highway and the study records lane changing procedures. The second scenario occurs when the vehicle turns; it eventually reaches a small town with traffic calming measures. In the control scenario there will be no advance street name signs.
    • John Molino asked which traffic calming measures are of most interest and which are planned?
      • Tom Welch (IA) indicated that painted shoulders, painted neck–downs, narrow lanes, and speed humps were most important to him.
      • Cliff Reuer (SD) added that painted chicanes are of interest.
      • Stephen Read (VA) would like to see physical rather than painted treatments
      • There was a general discussion about painted or physical treatments and it was decided that bulb–outs, chicanes, and raised medians at intersections need to be physical treatments.
      • Pat Brady (FL) wanted to know whether the simulator scenarios are curb–and–gutter or open drainage. He advocated the use of a curb and gutter setting because it more closely replicates a small town.
    • There was a short discussion of transverse markings on the edge and centerlines, which give the perception that the driver is traveling fast. A comment was made regarding the use of painted lines that were lengthened rather than closer together to give the tunnel effect.
    • Paul Jovanis (PSU) was concerned that the simulation–based studies would not produce reliable results that could be used to develop crash reduction factors (CRFs). He mentioned that earlier studies (Evaluations 1–4) reported the safety effectiveness of countermeasures. He was concerned that we do not know enough about surrogate measures to reflect safety effectiveness.
      • John Molino noted that the surrogates are very close to pre–crash validity.
      • Paul Jovanis (PSU) added that we first need to identify crash surrogates that mimic crash data.
    • John Molino raised a concern that the crash data used in the evaluations is suspect and noted that the simulation is more valid.
    • A discussion was held regarding the types of treatments that would be appropriate for the simulator. It was noted that during the previous meeting, curve treatments were identified as a potential strategy for the simulator study because they are relatively rare. The TAC members seemed to agree that the advance street signs are more common and we may be able to get additional data and evaluate with a crash–based study.
    • Roya Amjadi (FHWA) expressed her concern that the simulation may be used to confirm safety decisions, but should not be used as a sole measure to support the implementation of strategies.
    • Gary Modi (PA) suggested the use of simulation to decide whether people like or dislike a particular treatment and how they react to it. This is important for relatively new and innovative treatments before they are installed in the field.
    • The discussion turned to Part 2 of the simulation discussion – strategies for rural 2–lane roads.
    • Ten strategies were presented for horizontal curves and again the question was asked: what is important and practical?
    • Tom Welch (IA) indicated that Iowa has been replacing existing chevrons with signs of higher florescence. He added that some type of incremental analysis would be interesting to determine whether there is such a thing as too much treatment?
    • LED and illuminated edge lines were discussed as potential treatments for the simulator, but were not considered because they are less likely to be implemented due to high costs. Cross–hatching the shoulders was also included in this discussion, but was not a popular strategy because it is very maintenance intensive.
    • There was a mixed opinion of actuated signs. Some felt that the signs are high cost, but others feel that they can be very effective and are relatively low cost compared to curve realignment.
    • Janice Benton (CA) indicated that the costs of activated signs are not too high so they need to be considered.
    • There was then a discussion about the usefulness of simulation. It was noted that many States are hesitant to install new strategies because of the unknown effects. Simulation may spur the implementation of these devices because there is a conservative attitude in general until some reason exists to implement the strategies. It was also noted that various designs (e.g., size and brightness) of treatments could be evaluated in the simulator to see what is more effective.
    • Tom Welch (IA) voiced his concern that the first curve is the most important because drivers adjust their behavior as they progress through the simulation. He would like this to be considered in any type of simulation for curve treatments. Tom also mentioned the interest in positive guidance and issues with impaired drivers for possibilities in the simulator.
    • Rick Pain (TRB) mentioned the use of secondary tasks to distract drivers and hit them with new conditions to determine the range of reactions and reaction times.
    • The discussion was brought back to the situation where a road is leading into a small, rural town and the speed limit is decreasing from high speed to low speed. The example was given for a 2–lane highway with moderate traffic going into a small town and speeding has been noted as an issue.

June 19th meeting adjourned for the day.

Day 2: Wednesday, June 20, 2007

FHWA Products

  • Ray Krammes (FHWA) presented on recent FHWA products available.

Marketing Plan

  • Vicki Glenn discussed the Draft Marketing Plan for Low Cost Evaluations1–4.
  • She led a discussion to get suggestions from the TAC on who would be the potential users of the results of the evaluations.
    • TAC members prioritized intended users to help the measures get adopted within the local transportation community.
    • Several TAC members voiced their opinion that it depends on structure of state:
      • Decentralized states need local engineers and maintenance on board.
      • Centralized states need CEO's on board.
    • Based on inputs from the TAC members the primary users were identified as people who make the decisions on what projects get put in place. This is different groups for different DOTs and ranges from CEOs to traffic engineers in districts.
    • Additional primary users were identified as:
      • Professional Organizations – all equally important especially AASHTO.
      • LTAP managers do good job of getting the projects to the local engineers.
      • City and local officials – MPOs/cites/counties.
    • Secondary users were identified as:
      • Government state safety officials since they are already on board, they are a secondary audience.
      • UTC/Academia – just by publishing will get to them.

Status of Next Four Evaluations

  • Frank Gross (VHB) presented the preliminary results of the lane/shoulder width evaluation.
    • Jana Lynott (AARP) asked if there were any sites with high bicycle volumes on rural roads.
      • These data are currently not available for the datasets used in the analysis.
    • Gary Modi (PA) asked whether the analysis was related to paved or unpaved shoulder width.
      • The results were related to total paved width including the paved shoulder width; the unpaved shoulder width was included as an additional variable in the model.
    • Jana Lynott (AARP) asked if it is correct to interpret the results as shoulder width having a greater effect than lane width.
      • Frank Gross clarified that he was not saying that increasing shoulder width has a greater effect than increasing the lane width, but the results indicate that, for a given lane width, crash risk decreases as shoulder width increases.
    • Gary Modi (PA) commented that the results for the 34 foot pavement width are based on a small sample size and, therefore, we should not read too much into that result.
      • Frank Gross agreed that he was correct and added that the results are preliminary and more analysis needs to be completed before drawing any solid conclusions.
    • Frank Gross posed a question to the TAC regarding the anomalies in the results for pavement widths with 10' lanes and 7' shoulders. He asked if there were any ideas why 10' lanes with 7' shoulders have a much higher crash risk than all other pavement widths and combinations.
      • There were no responses to this question.
    • Shawn Troy (NC) asked if all the roads included in the study were rural.
      • All segments included in the analysis were rural, 2–lane roads with an average speed around 45 mph.
    • Tom Welch (IA) asked if we have any data on the traffic volumes or percent of heavy truck traffic.
      • The case–control pairs are matched using ADT, so ADT is accounted for in the analysis. The analysts did not look at truck traffic, but can if it is available through HSIS for Washington. Gary Modi (PA) confirmed that the data are available for Pennsylvania.
    • Tom Welch (IA) asked if curvature was included in the analysis and indicated that the variables are very important.
      • Curvature and grade are potential confounding variables that were both included in the analysis of the Washington data. The data are not available for Pennsylvania, but they can include these variables in the disaggregate analysis for Washington.
    • Shawn Troy (NC) asked the TAC what are the typical pavement cross–sections used?
      • Tom Welch (IA) indicated that they are trying to use 28' cross–sections with 2 ft gravel shoulders.
      • Shawn Troy (NC) added that most states will have a lot of interest in 28' and 30' pavement widths because there are few rural roads in North Carolina with pavement widths greater than 30'.
    • Frank opened the discussion to the TAC and asked what their thoughts were on this issue; what lane widths do you typically employ?
      • Shawn Troy (NC) commented that most States do not use much else than 28' or 30' pavement widths so it might not be worth it to find a large enough sample size for the pavement widths greater than 30'.
      • Brad Steckler (IN) commented that pavement widths are all over the place for their State roads, but this study also takes into consideration local roads. He also asked if the States report lane and shoulder widths to FHWA.
      • Carol Tan (FHWA) indicated that the States report on their Highway Performance Monitoring System (HPMS), but it is only a sample of the entire population.
      • Brad Steckler (IN) added that they would have as many 36' pavement widths on rural roads as they do other widths and would be interested in these results.
    • Kim Eccles (VHB) commented that perhaps colder states have wider roads?
      • The States indicated that the pavement width has more to do with the topography and available right–of–way.
    • Paul Jovanis (PSU) indicated that the anomalous results for the 34' pavement width were may be due to a few peculiar sites.
    • John Smith (MS) mentioned there is a minimum standard of 30' pavement width for new construction in Mississippi.
    • Frank Gross asked if there is a minimum cut–off that we should not look any further below.
    • Brad Steckler (IN) added that this may be applicable to new roads, but not existing roads.
    • Tom Welch (IA) indicated that there are a lot of 24' pavement widths with 12' lanes and gravel shoulders in Iowa. He said that if they could move the lane in a foot to create an 11' lane with a 1' paved shoulder, it would reduce maintenance costs. He added that the results from this study would help convince maintenance folks that this is a good idea.
    • Shawn Troy (NC) asked if there in any interest in looking at paved widths of 26'?
      • If the project team can obtain the data, they can evaluate paved widths of 26'.
    • The question was posed to the TAC if there are States that would want less than 26' feet?
      • No state other than NC really spoke up about it.
    • Cliff Reuer (SD) asked if there be any benefit of have more than 2 states.
      • Yes, if there is additional data VHB would be happy to add it to the study.
      • Frank Gross explained that data from PA was used because they offered it at last year's TAC meeting and WA was used because they are an HSIS state.
    • There was a brief discussion led by Carol Tan (FHWA) to explain what the Highway Safety Information System (HSIS) is to those in the meeting who were not familiar with it.
    • Gary Modi (PA) asked if the project team was planning on doing an economic analysis.
      • Yes, restriping is very low cost; it is virtually free.
    • Bhagwant Persaud (Persaud and Lyon) asked the TAC if any of the sates have any limited before/after data with reallocating the lane width.
      • Tom Welch (IA) is in middle of study looking at variations of lane widths.
      • Shawn Troy (NC) will have some. There may be a problem as there is only a handful funded with some pushed off indefinitely because of funding.
  • Jim Jenness (Westat) gave a presentation on advanced street name signs.
    • Pat Brady (FL) asked what types of crashes will be included in the evaluation.
      • All crashes involving older drivers, whether they caused it or were hit by someone else would be included in the evaluation.
    • Jana Lynott (AARP) asked what VHB considered an older driver.
      • Age 65 and above.
    • Kim Eccles (VHB) asked Jana Lynott what AARP would define as older driver.
      • Jana Lynott (AARP) responded that age 65 and older is appropriate.
    • Jim Jenness emphasized to the TAC that more installations need to be collected in order to meet the required sample size.
    • Stephen Read (VA) asked if the street name is also mounted on a traditional street sign.
      • In NC yes, in MA, no.
    • Shawn Troy (NC) commented that based on type of crash that would be targeted with this strategy; the evaluators may need to look at crashes more than 250 feet away from the intersection.
    • Pat Brady (FL) noted that distance might be more than 500 ft.
    • Kim Eccles (VHB) posed the question to the TAC – What distance do you put your signs in at?
      • Scott Jones (UT) responded that the speed of the road would determine the distance.
      • Gary Modi (PA) noted that for high–speed roads PA uses 500 ft.
    • Brad Steckler (IN) asked why the evaluation was focusing on urban roads.
      • Those were the only locations for which data were available.
      • Brad Steckler (IN) noted that how far from the intersection that should be considered depends on urban/rural environment.
      • Kim Eccles (VHB) noted that if data were available for rural locations they would be included in the study.
      • Tom Welch (IA) stated that Iowa installs them in rural locations.
    • Brad Steckler (IN) noted that Indiana has the problem of not knowing when the signs were installed.
      • As long as the year they were installed is known, they can be included in the evaluation.
    • Shawn Troy (NC) asked if sample size was a concern, what about just doing comparison study rather than a before/after study?
    • Jim Jenness and Frank Gross (VHB) stated that it would require an even larger sample size and Bhagwant Persaud (Persaud and Lyon) agreed.
    • Tom Welch (IA) asked if a cross–sectional study across states could be done.
      • Jim Jenness indicated that there is only major ADT in Massachusetts.
      • Pat Brady (FL) noted that the evaluation would focus on crashes on the major road anyways.
      • Jim Jenness agreed that although that was true, minor road ADT data are still necessary for the evaluation.
  • Nancy Lefler (VHB) presented on the status of the offset left–turn lanes data collection effort.
    • Nancy Lefler emphasized that the evaluation is focusing on those locations where the left–turn signal is not fully protected.
    • John Smith (MS) noted that large trucks especially logging trucks are an issue in Mississippi and that limits the use of the offset turn lanes in that state.
    • Jeff Wolfe (KY) asked if using a flashing yellow is an alternative to offsets.
      • Shawn Troy (NC) stated that North Carolina is monitoring the performance of flashing yellows; instead of a green ball the driver sees a flashing yellow. These are not necessarily at offset left–turn lanes.
    • Joseph Santos (FL) noted that there may potential difficulty of allowing a U–turn with offsets left–turn lanes but that would only an issue with narrow medians.
    • Brad Steckler (IN) noted that in Indiana they employ a different design. They install slanted left–turn lanes but have some concerns with rear–ended vehicles being pushed into oncoming traffic with this design.
  • Frank Gross (VHB) presented on the status of the data collection effort for the dual application of rumble strips on the centerline and edgeline.
    • Tom Welch (IA) suggested that the evaluation separate out the curve crashes & older and younger drivers. Some states have opposition to installing these, but if we can show the effectiveness on curves states might install them just on curves if not the whole roadway.
    • A TAC member asked if all of the installations include resurfacing.
      • Yes, they do.
    • Jeff Wolfe (KY) asked if they have had opposition to it from motorcyclists.
      • Pat Brady (FL) stated that in Florida the lane lines are more of an issue with motorcycles. The Florida rumble stripes are build–up rather than milled–in.
    • The States asked if the evaluation would be able to account for the different types of installations.
      • That can be accounted for in the study.
    • Dave Engstrom (MN) indicated that Minnesota did some research on motorcycles with centerline rumble strips.
    • Stephen Read (VA) noted that the faster motorcycles go the less of an issue rumble strips are.
      • Janice Benton (CA) indicated that California has an issue on rural two–lane roads with rumble strips. Motorcycles hit rumble strips and lose control because they are going too fast.
    • Janice Benton (CA) indicated that California might have sites with dual applications.
    • Jeff Wolfe (KY) noted that noise is an issue with rumble strips.
      • Janice Benton (CA) stated that California has issues with noise too, but bikes are more of an issue. She added that they did a study and designed rumble strips to pacify the biking community.
    • Dave Engstrom (MN) asked if there is a way to compare combination rumble strips with just edgeline rumble strips.
      • Yes, the analysis can account for different designs if the data are available.
      • Kim Eccles (VHB) mentioned that we can work with the NCHRP 17–32 team and see what we can do to evaluate dual applications where edgeline rumble strips were installed prior to centerline rumble strips. She added that we may be able to have them present next year at the TAC meeting.
    • Tom Welch (IA) indicated that Jeni Behar from iTRANS has completed a study of edgeline rumble strips and there is high confidence in the results.
    • Shawn Troy (NC) told an anecdotal story where a person complained about the noise associated with rumble strips. NCDOT conducted a noise study and determined that the rumble strips were no louder than the current truck traffic on the road, but it was just a different type of noise. In the end, the State had to repave the centerline rumbles strips because of one complaint. This incident has affected the installation of centerline rumble strips through the entire State of North Carolina.
    • Kim Eccles (VHB) mentioned that there were some TAC States, surveyed as part of the NCHRP 17–32 project, which may have combination rumble strip applications.
    • Reed Henry (AZ) indicated that Arizona should have some data related to combination rumble strips.
    • Jeff Wolfe (KY) indicated that Kentucky has had shoulder rumble strips for years, but the centerline installations are relatively new.
    • Frank Gross commented that we may be able to compare just edgeline rumble strips with combination (edgeline and centerline) rumble strips the appropriate data is available.
  • Kim Eccles (VHB) and Tom Welch (IA) presented on possible additional data sources and data issues faced by the states.
    • Kim Eccles and Tom Welch indicated that local governments (e.g., county and city engineers) may be a good source of additional data. They asked if there were any other suggestions for data.
    • David Engstrom (MN) indicated that a county engineer responded to the request for STOP AHEAD pavement marking locations.
    • Tom Welch (IA) commented that the Highway Safety Improvement Program (HSIP) requires States to track their safety improvements that are funded with HSIP funds. He added that this may be a good source of project locations and would require little additional effort.
    • Pat Brady (FL) mentioned that in Florida the involvement with local agencies only occurs when the local agencies want to be included in a special solution; typically when the agencies want answers to a specific question.
    • Pat Brady (FL) asked how a State is included in HSIS.
      • Carol Tan (FHWA) responded that due to funding constrains there is a limit to how many states they can include. There is currently only enough funding to maintain what they have.
    • Brad Steckler (IN) wanted on know what kind of data tracking would be good? What would be unnecessary? What countermeasures should states be tracking?
      • VHB developed a database for States to use to track data. They will send a copy to Brad and anyone else who would like a copy.
    • Roya Amjadi (FHWA) stated that based on her experience with construction management system software it would be possible to just add a column to the existing database to track safety projects. When a construction engineer is entering the project information into the software they can just check off that it was done for safety. Another way to track safety projects is through the Highway Management software. They can just add another column for safety. The same can be done for traffic crew, pavement markers, and bridge management systems. Every field within the DOT can do this within management systems that are already being tracked.
    • Tom Welch (IA) noted that the biggest problem is resources of the staff. He suggested that states can use students, FHWA interns, or the consultants will come get the data.
      • Kim Eccles agreed that VHB can come do the work to collect the data if states think there are data to collect.
    • Frank Gross (VHB) asked the TAC who does the markings for each of the states.
      • Pat Brady (FL) noted that Florida hires a private company. Florida does not have the records.
    • Tom Welch (IA) noted that the study results are only as good as the data the TAC submits to the evaluators.
    • Brad Steckler (IN) asked what types of data / projects they should be tracking.
      • Tom Welch (IA) noted that HSIP projects should be tracked first.
      • Brad Steckler (IN) indicated that Indiana does not always label these projects as "safety projects" – so how would he know what to track.
      • Pat Brady (FL) stated that Florida tracks both those safety projects and those that improve safety but are not necessarily "safety projects".
    • Joseph Santos (FL) asked if there are ever general safety studies done. For example, for the street name signs, can an evaluation of general crash statistics of a state be conducted?
      • Roya Amjadi (FHWA) responded that some have been done. She emphasized how important it was that the States track and keep all data.
    • Carol Tan (FHWA) noted that it is not just research projects that need data – everyone needs better data. There is a need for better more detailed data.
    • Clayton Chen informed the TAC that as of last Friday, as part of SAFETLU – Model Minimum Uniform Crash Criteria (MMUCC) and Model Minimum Inventory of Roadway Elements (MMIRE) are available as recommended guidelines for data.
  • Clayton Chen from the FHWA discussed the efforts he is undertaking that relate to the Low Cost project.
    • Regarding the lane widths shoulder widths study, in approximately 2003 that was an NCHRP project that produced software called the Resurfacing Safety Resource Allocation Program. The program can evaluate the trade–off between existing travel lane from 10–12 ft for resurfacing/restriping projects. It did not have capability with trade–offs of fixed travel lane but it did have cost benefit and operational capabilities.
    • There is a task Clayton is currently managing which is a desktop reference that provides a compilation of existing crash reduction factors (CRF).
    • Intersection Issue Brief Number 8 contains crash reduction factors.
      • Tom Welch (IA) offered to post the issue brief on the safety listserv.
    • The first release of the Highway Safety Manual will come out soon. Factors are presented as CRFs not accident modification factors (AMFs). There was some debate within panel on this issue, but decided to use CRFs because that is what engineers are most likely to use in the field.
  • Raghavan Srinivasan (HSRC) gave a presentation on the NCHRP 17–35 Low Cost Signalized project.
    • Tom Welch (IA) asked if the strategies can be segregated by speed.
      • That is a good point, yes they can.
  • Kim Eccles (VHB) gave a presentation on the NCHRP 17–25 Crash Reduction Factors for Traffic Engineering and ITS Improvements project.
    • Jeff Wolfe (KY) asked if using LEDs on traffic signals are a concern for any of the states.
      • John Carey (CT) stated that there have been some problems in Connecticut with closely spaced intersections drawing people through.
      • David Engstrom (MN) stated that Minnesota has had problems with snow since the LEDs don't put off the heat to melt the snow.
      • Scott Jones (UT) noted that some LED manufacturers are making them with a heat component that you can turn on to address that problem.
    • A TAC member noted that none of the strategies that are being evaluated are ITS strategies.
      • Although the name includes ITS strategies, the project team could not find any studies or data for strategies that could be used in an evaluation.
    • John Smith (MS) stated that due to Hurricane Katrina, Mississippi lost about 100 signals that were span wire and were not reimbursed for them. The state is now only installing mast arms.
    • Stephen Read (VA) indicated that the Virginia DOT is finishing up a study in multiple sites of going from protective–permissive to protected left–turn phasing in Northern VA.
  • Michael Trentacoste (FHWA) thanked the attendees for all of their invaluable contributions to the project discussed the direction of the project.
  • Roya Amjadi (FHWA) led a discussion on the future direction of the project.
    • Tom Welch (IA) stated that Wisconsin has installed alternating passing lanes. This was a strategy presented in Phase I of the project but was not evaluated due to lack of data.
    • Stephen Read (VA) asked how the heavy truck strategies overlap with what is being done at the Federal Motor Carriers Safety Administration (FMCSA).
      • Amy Hauser (FMCSA) knows a little of what is being done. Some of the strategies really tie in especially those that are related to the on–board safety systems. There is definitely possibility of working together. Relaying information over the air to drivers regarding hazardous locations (work zones, crashes, traffic, etc) is an important strategy.
    • States were given lists of strategies; some were from last two phases, some were from new books. The States were asked to rank the strategies they were most interested in seeing get evaluated from 1–5.
  • James Montgomery (OK) gave a presentation on Oklahoma 's experience with wire rope cable barrier.
    • Jeff Wolfe (KY) asked if the barrier was on same side of the road for the entire length of the project.
      • Yes, it was on the same side of the road throughout.
    • Brad Steckler (IN) asked if there were some things that did not work quite right.
      • They would have spent a little more money and would have put it closer to the median.
    • Brad Steckler (IN) asked when the DOT might adopt another non–rigid barrier for the median.
      • They would use guardrail on narrow median.
    • Jeff Wolfe (KY) noted the importance of educating emergency personnel when responding to a crash that involved the cable barrier.
    • Reed Henry (AZ) asked how they dealt with the drainage structure – did they have to modify the slope.
      • They raised the drainage structure.
    • Mike Trentacoste (FHWA) asked how they are determining nuisance hits.
      • For those that are not reported, police would see posts down and report them.
  • Shawn Troy (NC) presented on evaluations of low cost countermeasures conducted in North Carolina.
    • A TAC member asked how much time the evaluations take.
      • They take a maximum of 24 working hours. The most time is spent traveling from the DOT headquarters in Raleigh to these sites.
    • Jeff Wolfe (KY) asked when does the DOT revisit the projects.
      • Currently, they are working on 2002 projects.
  • Due to time constraints, a presentation on advanced warning flashers developed by Ron Lipps form the Maryland Department of Transportation could not be given.
    • Scott Jones (UT) noted that Utah is conducting a study on advanced warning flashers.
    • Kim Eccles encouraged him to share the results of the study with the TAC when it is complete.

Meeting Adjourned