Lane Departure Strategic Action Plan

Sample Plan

March 23, 2005

Prepared for:

DOT logo

Federal Highway Administration
Office of Safety

Prepared by:
SAIC logo

Executive Summary
Lane Departure Crash Overview
The Cost of Lane Departure Crashes
Summary of Major Concerns
Lane Departure Goal
Approach
Statewide Initiatives to Improve Safe Driving Behavior
Other Actions
Summary of Proposed Initiatives
Key Strategic Direction Issues

Executive Summary

Over the past five years nearly 5,000 people have perished on highways within the State and, of these, nearly 3,000 died in run-off-the-road (ROR) and head-on (HO) collisions. This is an extraordinary number of tragic events.
In past years a gradually improving infrastructure has helped keep the numbers from increasing, but higher traffic volumes have counteracted any real reductions in numbers of deaths. Two major programs account for a gradually improving safer infrastructure

First, the Hazard Elimination Program has improved between 60 and 70 high-hazard locations each year. Approximately 50% of these high-hazard locations are intersections and 30% are curves.
Second, the 1R and 3R Programs have annually widened about 40 miles of pavement, widened and paved over 80 miles of shoulders, and converted over 120 miles of stabilized shoulders to paved shoulders.
In addition, a number of driver safety efforts have helped cap the number of deaths on the highways. These efforts consist of education and enforcement initiatives to increase safety belt usage, reduce drinking and driving, and reduce aggressive driving and speeding. However, if we want to begin to reverse the flat trend and create a substantive reduction in the number of statewide lane departure deaths, other strategic safety efforts are needed.
This Action Plan concentrates on identifying those strategic infrastructure-related initiatives that can substantively reduce lane departure deaths. Three principles have been established to better insure success of the plan.

Each strategy must be acceptable to the vast majority of the public.
Each strategy must be cost effective in terms of crash reductions.
Each strategy must substantively contribute to achieving the goal – reducing serious injuries and deaths associated with lane departure.

It is important to note that improving safe driver behavior is also crucial, since a large majority of lane departure deaths involve some form of unsafe driver behavior or error. In addition to proposed infrastructure improvements, actions to reduce drinking and driving and aggressive driving, increase safety belt usage, improve safe driving by young drivers, and increase the survivability of severe rural crashes by enhancing statewide and corridor-based emergency medical services (EMS) are also being pursued.
The plan demonstrates that a minimum 10% reduction in lane departure fatalities and serious injuries – more than 300 lives saved over the next five years – can be achieved through the strategic application of infrastructure safety initiatives and in accordance with the principles described above. However, this plan also requires addressing new challenges and taking new approaches to achieve success. As such, it includes the following:

This plan estimates the following annual resource needs for each of the next four years to fund the level of cost effective improvements needed to prevent over 330 lane departure deaths and serious injuries annually:

State highway system – approximately $25-30 million.
Local government safety improvements – $4 million.
Safe driver improvements on safety corridors – $5.0 million.
Considering the competing needs of other priorities, and recognizing the potential for additional safety funding in the upcoming federal reauthorization, some level of supplemental funding will be needed to realize a substantive reduction of lane departure deaths and serious injuries.

Wide use of two key strategies which are not being applied statewide – center/shoulder/edge rumble strips and select tree removal in rural areas – is both important in attaining the lane departure goal and potentially controversial. Steps will be taken to implement the improvements as extensively and cost effectively as possible while minimizing controversial actions/decisions.

With rumble strips, it is important that we are sensitive to potential noise concerns of adjacent residents and riding activity levels and issues of bicyclists. As an initial step, it is suggested that meetings with the State Pedestrian/Bike Coordinator, the Pavement Design Engineer, Design Engineer, and Maintenance Engineer be held to establish tentative parameters for application on the routes that have a ROR crash problem, resolving any rumble strip design issues. We should then strive to install approximately five to 10 center and edge rumble strip sections on agreeable non-controversial rural two-lane route sections that have a ROR crash problem during the 2005 construction season to gain initial experience with the strategy. We can then expand implementation, learning from the experiences on these test sections.
With the selective rural tree removal initiative, it is important to institute a policy and process identifying vulnerable trees that are of value environmentally, to the community, or to adjacent residents. This initiative must also have the flexibility to allow alternate improvements to removal, such as tree delineation, that preserve environmental and community values while addressing the crash problem. We should apply this policy on approximately 10 high tree crash locations to gain experience and potentially enhance it.

Many of these low cost improvements need to be considered for inclusion in 1R and 3R design projects, particularly if a ROR or head-on crash problem exists within the limits of work. The 1R and 3R design process needs to be revisited to determine under what circumstances low-cost safety improvements mentioned in this plan should be included in the contract.
Approximately one-third of all serious injuries and deaths occur off the State highway system. Realistically, improvements will not be extensively implemented on local roads unless federal funds and technical transfer assistance are extended to the municipality. In the past, funding to municipalities has been limited to specific programs, such as the Transportation Enhancement Program, that have targeted funds. We need to develop guidelines for when federal funds can be made available to municipalities to fund safety improvements identified in this plan.

In conclusion, we believe a substantive reduction in lane departure fatalities and serious injuries can be achieved through the judicious application of select strategies. It will, however, require the Department to: (1) invest a moderate level of funding for the strategies, and (2) have the fortitude to initiate new directions and safety strategies in a way that is acceptable to the vast majority of the public and remain cost effective.

Lane Departure Crash Overview

Over the last five years (1999 to 2003) highway fatalities, serious injuries, and crashes in the State have remained relatively stable with a very slight upward trend.

Table 1. Highway Fatalities, Injuries, and Crashes

	1999	2000	2001	2002	2003
Fatalities	962	982	974	998	1,003
Serious Injuries	4,561	4,432	4,619	4,481	4,596
Reportable Crashes	98,762	97,561	99,213	100,415	100,312

Lanedeparture crashes in which a vehicle unintentionally departs from its lane and crashes with another vehicle, rolls over, or hits a fixed object are a substantial portion of the statewide crash problem. In 2003, these crashes accounted for approximately 60% of the statewide serious injuries and deaths and 40% of all crashes.

Table 2. 2003 Lane Departure Crashes

	Single Vehicle Run-Off-the-Road (ROR)	Head-on, Non-Freeways	Head-on, Freeways
Fatalities	413	187	15
Serious Injuries	1,934	756	72
Reportable Crashes	35,622	2,836	216

Over the past five years, lane departure deaths and serious injuries have remained relatively constant even though vehicular travel has increased.

Table 3. Lane Departure Deaths and Serious Injuries

	1999	2000	2001	2002	2003
Lane Departure Deaths and Serious Injuries	3,261	3,301	3,274	3,278	3,290
VMT (times 100 million)	587	608	628	647	667
Lane Departure Death and Serious Injury Rate	5.55	5.42	5.21	5.06	4.93

Thus, even though the number of lane departure fatalities and serious injuries has remained relatively constant, the rate has declined. There are three primary factors responsible for the decline:

A Gradually Improving Infrastructure. Two major programs account for a safer infrastructure. The High Hazard Program has improved between 60 and 70 high hazard locations each year. Approximately 50% of these are intersections and 30% are curves. The 1R and 3R Programs have annually widened about 40 miles of pavement, widened and paved over 80 miles of shoulders, and converted over 120 miles of stabilized shoulders to paved shoulders.
Safer Drivers. Over the past five years safety belt usage has climbed from 68% in 1999 to 75% in 2003. The passage of driving under the influence (DUI) legislation in 1999, which lowered the blood alcohol concentration (BAC) level from 0.10 to 0.08, is another deterrent to drinking and driving.
The Unbalanced Growth of Traffic. Traffic is growing at a rate of approximately three percent per year. On close examination, this growth is occurring predominantly during daylight and early evening hours. The fatality rate of highways is about four times greater during late night-early morning hours than in daylight driving hours. In addition between 40% and 50% of the fatalities and serious injuries occur during this time period. Thus if nothing else occurs but a growth of traffic, predominantly during daylight and early evening hours, the overall rate will naturally decline.

This initiative will address run-off-the-road and head-on non-freeway crashes. It is to be noted that the head-on crashes also include opposing flow sideswipe crashes. Freeway head-on crashes will be addressed when the national guidance document for this area is available.

The crash problem needs to be analyzed from a road ownership and urban/rural perspective since these elements have a profound influence on the application of many of the strategies. The following two tables illustrate the distribution of crashes and fatalities by system ownership.

Table 4. Distribution of Crashes by System Ownership

		ROR Crashes			Head-On Crashes
	Miles	Crashes	Serious Injuries	Fatalities	Crashes	Serious Injuries	Fatalities
Rural State Highways	28,612	15,312	1,002	239	1,239	486	141
Urban State Highways	8,514	9,199	384	60	762	42	10
Rural Local Highways	62,113	6,641	412	84	414	161	24
Urban Local Highways	14,710	4,470	136	30	421	67	12
Totals		35,622	1,934	413	2,836	756	187

Table 5. Distribution of Fatalities by System Ownership

	Percentage of Lane Departure Deaths (ROR)	Percentage of Lane Departure Deaths (Head-On)	Total Percentage
Rural State Highways	40	23	63
Urban State Highways	10	2	12
Rural Local Highways	14	4	18
Urban Local Highways	5	2	7

Obviously the rural State highways are the predominant concern, but the other three categories collectively add up to a considerable portion of the problem with lane departure crashes.

Run-Off-The-Road Crashes – Serious Injuries and Deaths on Curves

A substantial portion of ROR crashes occurs on curves as indicated in Table 6.

Table 6. Distribution of ROR Serious Injury or Death Crashes

	Total	Curve	Straight
Rural State Highways	1,241	718	523
Urban State Highways	444	365	79
Rural Local Highways	496	342	154
Urban Local Highways	166	138	28

In addition, 72% of the serious injury and fatal curve crashes occurred at night. Thus a considerable portion of the problem occurs on curves at night.

Collisions with Fixed Objects

The types of collision where vehicles run-off-the-road and result in serious injury or death are shown in Table 7. In rural areas, trees, rollovers, and ditches are the primary concerns. In urban areas trees, utility poles, and rollovers are the most common collision types.

Table 7. Distribution of Collisions with Fixed Objects – Serious Injury or Death Crashes

	Rural State Highways	Urban State Highways	Rural Local Highways	Urban Local Highways
Rollover	206	52	81	41
Tree	312	75	191	31
Utility Pole	65	64	42	38
Bridge	32	12	30	8
Ditch	162	10	64	12
Guard (Guide) Rail	84	21	41	10
Other	380	210	47	26
Total	1,241	444	496	166

Illegal Driver Actions

Illegal driver actions play a significant role in many of these severe crashes. Three major types are as follows:

Drivers who lose control due to moving violations – these include speeding, reckless driving, and failure to obey a traffic control device (TCD),.
Alcohol related crashes – only fatal crashes are used in this analysis since other crashes are considered under reported relative to alcohol consumption.
Unbuckled drivers and occupants.

Table 8. 2003 Crashes Involving Illegal Driving Actions

		ROR Crashes	Head-On Crashes
Crashes Involving Illegal Driving – Serious Injuries and Fatalities	Total	1,261	445
	% Of All Lane Departure Serious Injuries and Fatalities That Involved Illegal Driving Actions	53	47
Alcohol-Related Fatalities	Total	240	93
Alcohol-Related Fatalities	% Of All Lane Departure Fatalities Involving Alcohol	58	50
Unbuckled Driver and Occupant Deaths	Total	310	108
Unbuckled Driver and Occupant Deaths	% of All Lane Departure Fatalities That Were Unbuckled	75	58

Because many of these crashes had multiple factors (e.g., an intoxicated, unbelted driver who is also speeding crashes into a tree), the percentages in the table exceed 100%. Obviously, illegal driving actions constitute a significant portion of the problem.

In addition to illegal driving actions, drivers who are distracted, fatigued, inattentive, or who fall asleep are also considered a significant factor in lane departure crashes. Information from the crash data system is considered to underreport these factors significantly. Many of these drivers are not likely to admit the cause of the crash and/or the police officer who arrives at the scene some time after the crash occurred has no or minimal evidence to make such a conclusion.

The Cost of Lane Departure Crashes

The costs of lane departure crashes may vary significantly dependent on specific factors involved. Two key measurements are needed in the upcoming analysis.

Average cost of specific crash types expressed in dollars. This is needed to determine if an improvement may be cost effective.
Expected deaths and serious injuries per 100 crashes for various crash types. This is needed to determine the extent to which an improvement helps achieve a serious injury and fatality goal.

The monetary costs of crashes are based upon FHWA cost estimates:

Fatality – $ 3,000,000.
Serious injury – $209,000.
Non-serious injury – $42,000.
Possible injury – $22,000.
Property damage – $2,300.

The average crash costs and death and serious injury rates for various crash conditions as extracted from the crash data and applying the FHWA cost estimates are as follows.

Table 9. Average Crash Costs and Deaths/Serious Injury Rates

Condition	Average Costs ($000)	Deaths and Serious Injuries Per 100 Crashes
ROR State Rural	61	8.1
ROR State Rural -Tree	84	10.0
ROR State Urban	28	3.4
ROR Local Rural	58	7.5
ROR Local Urban	24	3.0
Head-On State Rural	198	37.6
Head-On State Urban	46	14.8
Head-On Local Rural	168	29.4
Head-On Local Urban	38	14.2

Summary of Major Concerns

In reviewing the lane departure overview and the costs of lane departure crashes, the prime concerns are as follows:

State and local rural highways with emphasis on rural State highways.
Curves, particularly under night conditions.
Trees, rollovers, ditches, and to a lesser extent utility poles and guard (guide) rails.
Illegal driving actions relating to loss of control, alcohol, and unbelted drivers and occupants.
Drivers who are inattentive, fatigued, and falling asleep.

Lane Departure Goal

After careful analysis of the data, and determining that the serious injury classification in the crash data system is relatively accurate, the combination of fatalities and serious injuries will be used as the performance measure for lane departure. This combination was selected because it provides more data than just fatalities and often times a serious injury could be a fatality except for a few minor changes in the crash parameters.

The reduction levels for rates in the national goal (1.5 to 1.0) and the date to reach the goal (2008) have also been adopted. Recognizing that separate initiatives will be underway to increase safety belt usage and reduce drinking and driving and aggressive driving (which will reduce lane departure deaths and serious injuries), this initiative will concentrate on infrastructure-related strategies that can reduce lane departure deaths and serious injuries. We will attempt to reduce the lane departure fatality and serious injuries from 3,290 in 2003 to no more than 3,000 in 2008.
Over the period 2003 to 2008, traffic volumes are expected to continue growing at a rate of three percent per year. Thus in 2008, if the goal of no more than 3,000 lane departure fatalities and serious injuries is achieved, the rate would drop from 4.93 to 3.88 (3,000/667*1.16).

The formalized lane departure goal is then as follows:

Using infrastructure safety strategies reduce the lane departure serious injury and deaths from 3290 in 2003 to no more than 3000 in 2008 and also reduce the lane departure death and serious injury rate from 4.93 to 3.88 per 100 million vehicle miles traveled.

Approach

Preventing 290 more lane departure deaths and serious injuries each year is a challenge, particularly since the existing safety efforts have been able only to maintain the existing level of almost 3,300 annual serious injury and deaths. New, effective strategies have to be skillfully employed if major reductions are to be achieved.

A review of the crash data and the information in the NCHRP Report 500 guidance documents yield the following opportunity areas as the most productive to achieve the goal:

System deployment of low-cost, cost-effective improvements.
Targeted medium- to high-cost improvements at higher crash locations.
Local road lane departure safety initiatives.

The implementation of potential strategies must satisfy three principles as follows:

They must be acceptable to the vast majority of the public.
They must be cost effective in terms of crash reductions.
They must substantively contribute to achieving the goal – reducing serious injuries and deaths associated with lane departure.

System Deployment of Low-Cost, Cost-Effective Improvements

The low cost improvements in NCHRP Report 500 were analyzed in conjunction with the crash data and the following types were selected for system deployment consideration:

Rumble strips – shoulder, edge, and centerline.
Advanced curve marking warning systems.
Tree removal/delineation.
Guard (guide) rail upgrades.

Each of these low-cost improvements is assessed in terms of actions to better insure public acceptance, deployment levels and associated characteristics, threshold levels to achieve cost effectiveness, estimated impacts in terms of lane departure reductions of serious injuries and deaths, and implementation costs.

Application Conditions for Acceptability

Shoulder and Edge Rumble Strips

Initially limit to rural highways on the State system where there is a history of ROR crashes that can be reduced by the installation of shoulder or edge rumble strips or similar sites to those that have a ROR crash history and there are no concentrations of homes that may be affected by noise issues.
Receive input from the State Pedestrian and Bicycle Coordinator.
Consider the use of bicycle tolerable rumble strips.
Where the State Pedestrian and Bicycle Coordinator anticipates moderate to high bicycle traffic, defer placement of edge rumble strips or shoulder rumble strips on paved shoulders less than six feet in width until a six foot paved shoulder can be placed adjacent to the pavement as part of a 3R project.
Limit applications to hot mix surfaces less than four years old. However, edge rumble strips may be placed on older pavements with the Pavement Design Engineer’s approval.
Only apply edge rumble strips on pavement structures of sufficient strength to deter edge break up.
Edge rumble strips may be placed on narrow pavements between 18 and 22 feet in width with a history of ROR crashes.

Centerline Rumble Strips

Initially limit to rural highways on the State system where there are no concentrations of homes that may be affected by noise issues and a history of head-on or opposite direction sideswipe crashes or similar sites to those that have a head-on or opposite direction crash history.
Limit applications to hot mix surfaces less than four years old. However, edge rumble strips may be placed on older pavements with the Pavement Design Engineer’s approval.
If center line rumble strips are perceived to be a problem in passing zones for passing maneuvers, consideration may be given to reducing the depth of rumble strips through passing zones from 1/2 to 3/8 inch milling to reduce the impact on passing maneuvers and still provide a warning system for drowsy and inattentive drivers who begin to drift across the centerline. If the 3/8 inch milling is used, test sections of both depths will be initially installed and an evaluation of the impacts of both depths made before wide use of the 3/8 inch depth.
Centerline rumble strips will be limited to those pavements 22 feet or greater in width.

Advanced Curve Marking Systems

Only apply on hot mix surfaces less than four years old.
Install on curves where drivers have to reduce speed by about 10 mph or more, horizontal curves superimposed on vertical curves, and curves that pose an unusual driving situation.

Tree Removal

Consider focusing on rural areas.
Receive input from the DOT State Environmentalist on avoiding any endangered species, removal conditions, and any other environmental requirements.
If trees are off the right of way, discuss the hazard with the property owner and endeavor to obtain written permission to remove the trees. Replacements in a safe location and/or storage of the wood may be offered to the property owner as appropriate.
Avoid sensitive areas where the trees have substantial aesthetic value.
In sensitive areas where trees cannot be removed and pose a night safety concern, install tree delineation on an experimental basis.
Include stump removal, minor regrading, sodding of the affected area, and removal of any other adjacent fixed objects.

Guard (Guide) Rail Upgrade

Limit to obsolete guard (guide) rail (strong post cable), replacement of blunt end sections and turned down ends, connecting unattached approach rail to bridge rail, and protection of unprotected bridge parapets.

Cost Effectiveness

Improvements deployed on a systematic basis have to be cost effective. A benefit/cost (B/C) analysis is used to make the determination. However, unlike a conventional analysis, the B/C is given or set, and the answer one seeks is the number of targeted crashes per unit length needed to make the improvement cost effective. The formula is as follows:

T = (Annual Cost)(B/C)/(Effectiveness)(Average. Crash Cost)

Where:

T = Threshold level of targeted crashes per unit length to consider the strategy to be cost effective.

Annual (Ann.) Cost = The annual cost of the improvement. For construction improvements, ann. cost is the construction cost annualized over the expected life of the improvement. For an education or enforcement improvement, ann. cost is the annual cost of a full year of enforcement or education.

B/C = The set benefit/cost ratio used to determine the threshold level of targeted crashes per unit length. It is usually between 1.0 and 2.0. In this case a value of 2.0 is used.

Effectiveness (Eff.) = The estimated effectiveness of the strategy in reducing targeted crashes.

Average (Avg.) Crash Cost = The average cost of targeted crashes.

A five-year crash history (1999 to 2003) is used in the analysis to provide more information on crash occurrences. The results of the B/C analysis for each of the low cost improvement types are provided in Table 10. Because of the low unit cost, it doesn’t take many crashes to justify an improvement.

Table 10. B/C Analysis Results – Threshold Crash Levels

Type of Improvement	Target Crashes	Unit Length	Construction Cost ($)	Expected Life (Yrs)	Average Crash Cost	Estimated Effectiveness	T** (Five Year Threshold Crash Level)
Edge/Shoulder Rumble Strips	ROR	3,000	9,000	10	61,000	.2	0.7
Centerline Rumble Strips	Head-On	15,000	37,500	10	122,000	.2	1.5
Advance Curve Warning	ROR Curve	1,000	4,000	4	66,000	.15	1.0
Tree Removal	Tree	1,000	40,000	10	84,000	1.0	0.5
Guard (Guide) Rail Upgrade	Guard (Guide) Rail	1,000	20,000	10	42,000	***	1.1

** T (five year) is the threshold level of target crashes needed over a five-year period to achieve a B/C ratio of 2.0.
*** Guard (guide) rail crashes will remain constant, but the severity of crash will decrease from $ 42,000 to $24,000 or a savings of $18,000 per crash.

The crash data system is now searched to find locations above the threshold and to estimate the cost and impact of applying the low cost improvements at the sites. Two examples are provided. The first example assumes that a State can locate crashes on the State highway system by location on a given route. The second example assumes that a State’s data system can locate crashes on a route but not to a specific location on that route.

Example 1 – Data Systems Where State Highway Crashes Can Be Located to Points on a Route

In this first example, where a State can locate crashes by specific location, a summary of costs and crash reduction impact by strategy is provided in Table 11. A higher threshold of crashes per unit length is selected above. This is calculated in Table 10 and is shown in Column 2 of Table 11. Because of the large number of independent variables that impact when and where crashes occur, a higher threshold has been established for each category to better insure that, if no actions are taken, a crash problem may exist in the future considering regression to the mean.

The crash data system is used to identify the statewide number of clusters of targeted crashes per unit length that equal or exceed this threshold (Column 3 of Table 11).
The total number of targeted crashes that occur in these clusters is provided in Column 4.
An estimate of the number of cluster sites that can be improved is provided in Column 5 of Table 11. Not all of the crash clusters can be improved using low cost improvements for a variety of reasons. As an example, one of the ROR clusters may have a stone and chip surface and not be suitable for milled rumble strips. This estimate will be refined in the next stage of development during site visits at each location to determine the applicability of a designated low cost improvement.
The total estimated cost of Column 5 improvements is provided in Column 6 and is the product of the estimated number of improvements times the average unit cost per improvement.
Column 8 estimates the number of annual targeted crashes reduced. This is based on the proportional estimate of targeted crashes that occur at all of the sites estimated to be improved multiplied by the estimated effectiveness of the improvement type {Column8=Column4x(Column5/Column3)(Eff.)/5}.
The estimated number of annual serious injuries and deaths reduced (Column 9) is computed by multiplying the annual number of targeted crashes reduced (Column 8) by the appropriate ‘deaths and serious injuries per 100 crash’ value found in Table 9.

Table 11. Cost and Impact of Cost-Effective Improvements – Low-Cost Improvements (Crash Data Systems That Can Locate Crashes By Specific Location on a Route)

Category	Five Year Threshold Crash Level Selected	Number of Statewide Crash Clusters	Number of Targeted Five Year Crashes in the Clusters	Estimated Number of Improvements	Construction Costs ($ Million)	Type of Crash Reduced	Annual Targeted Crash Reduction	Annual Serious Injury and Fatality Reduction
Edge/ shoulder rumble strips	4 ROR in 3,000’ (rural only)	2,650 (3,000’ sections)	25,800	1,650 (3,000’ sections)	15	ROR	640	52
Centerline Rumble Strips	3 HO in 15,000’ (rural only)	250 (15,000’ sections)	2400	200 (15,000’ sections)	7.5	Head-On	77	29
Advance Curve Markings	2 curve ROR in 1,000’	3,000 curves	24,600	2,400 curves	10	ROR	590	48
Tree Removal	3 hit tree in 1,000’ (rural only)	500 (1,000’ sections)	2500	400 (1,000’ sections)	16	Tree	400	40
Guard (Guide) Rail Upgrade	3 guard (guide) rail hits in 1,000’	250 (guard (guide) rail sections)	1400	200 (guard (guide) rail sections)	40	Guard (Guide) Rail Severity	0	4*
Totals					88.5		1707	173

* The number of guard (guide) rail strikes remains the same but the severity decreases from 4.4 to 2.6 serious injuries and deaths per 100 strikes.

Example 2 – Data Systems Where State Highway Crashes Can Only Be Located to a Route)

In the second example, where a State can only locate crashes by route and not by location on a route, a somewhat different approach is taken to make best use of the data available. In this case a minimum targeted crash level by route is established for each strategy. In addition, it is desirable to know the length of each route that exceeds the threshold such that a density of targeted crashes per unit length can be calculated and compared to a minimum density of targeted crashes per unit length. The density component can be dropped if a State does not have the route lengths available for automatic calculation of densities or doesn’t have the human resources to make the manual calculations. (It is estimated that no more than one to two person weeks of technician work effort would be needed for manual calculations if the route lengths are known and tabulated.)

Table 12. Cost and Impact of Cost-Effective Improvements – Low-Cost Improvements (Crash Data That Can Only Locate Crashes By Route)

Category	Five Year Threshold Route Crash Level Selected	Number of Routes that `Exceed the Threshold Clusters	Number of Targeted Five Year Crashes on the Exceeded Routes	Estimated Number of Route s Improved	Construction Costs ($ Million)	Type of Crash Reduced	Annual Targeted Crash Reduction	Annual Serious Injury and Fatality Reduction
Edge/ Shoulder Rumble Strips	8 ROR per route and a minimum density of 1ROR/ 5000’ (rural only)	1,200 routes	23900 ROR	800 routes	15	ROR	640	52
Centerline Rumble Strips	4 HO per route and a minimum density of 1 HO/10,000’ (rural only)	600 routes	2870 HO	200 routes	7.5	Head-On	77	29
Advance Curve Markings	3 curve ROR per route and a minimum density of 1 curve ROR/ 5000’	1,500 routes	29350 curve	,1000 routes	10	ROR	590	48
Tree Removal	4 hit tree per route and a minimum density of 1 tree/5000’ (rural only)	500 routes	2500 tree	400 routes	16	Tree	400	40
Guard (Guide) Rail Upgrade	5 guard (guide) rail hits per route and a minimum density of 2 guard (guide) rail crashes/5000’	250 routes	1850 guard (guide) rail	150 routes	40	Guard (Guide) Rail Severity	0	4*
Totals					88.5		1707	173

The threshold number of targeted crashes and minimum targeted crash density used to identify potential routes for low cost improvements is shown in Column 2 of Table 12.
The numbers of routes that equal or exceed this threshold, as determined through a computer run of the crash data system, are indicated in Column 3.
Column 4 indicates the five-year number of targeted crashes that occurred on the routes that exceeded the threshold.
The estimated number of routes that exceed the threshold and can be improved is provided in Column 5. A small, randomly selected, sample of routes (approximately 15 to 25) that exceed the threshold can be site-reviewed with input from local police and the county maintenance supervisor to determine the appropriateness and extent to which designated low-cost improvements can be placed on the selected routes. This information can be used to develop the overall estimate of routes that exceed the threshold and can be improved (Column 5).
Once the plan is approved, routes that exceed the threshold can be further assessed by: (1) a review of the actual crash reports (if readily available) to get further insights of targeted crashes along the route, (2) discussions with local police and county maintenance manager to get their input on the frequency and location of targeted crashes along the route, and (3) a site review to determine the applicability of the improvements.
The total estimated cost of Column 5 improvements is provided in Column 6 and is the product of the estimated number of improvements times the average unit cost per improvement.
Column 8 estimates the number of annual targeted crashes reduced. This is based upon the proportional estimate of targeted crashes that occur at all of the sites estimated to be improved multiplied by the estimated effectiveness of the improvement type {Column8=Column4x (Column5/Column3)(Eff.)/5}.
The estimated number of annual serious injuries and deaths reduced (Column 9) is computed by multiplying the annual number of targeted crashes reduced (Column 8) by the appropriate ‘deaths and serious injuries per 100 crash’ value found in Table 9.

Targeted Medium- to High-Cost Improvements at Higher Crash Locations

Strategies that are medium- or high-costs may also be considered for system deployment. However because of the higher costs, they are only considered at high crash locations. A review of the medium to high cost strategies in NCHRP Report 500 coupled with the crash analysis yields the following crash types for further analysis:

Hit Ditch and/or Rollover – Re-grade and flatten to traversable slopes.
High Crash Curves – Most of these curves should have been considered for the High Hazard program but for some reason (extremely high costs, right of way issues, or environmental concerns) the improvement cannot be made. In addition, existing signing and markings at these curves are in excellent condition. This initiative will first determine if specific physical improvements are both feasible and cost effective. Where they cannot be implemented cost effectively, infrastructure-only ITS smart curve warning systems are recommended.
Shoulder Widening – The roadway database contains information on shoulder width and type; sections with minimal or no shoulders and high ROR crashes can be identified. The improvements will consist of shoulder widening to at least six feet paved and rumble strips.
Slippery Surfaces – Sections that have a high number and proportion of wet pavement crashes will be skid tested to determine if they are slippery. If the skid number is below 30, they will be scheduled for improvement. If it is between 30 and 34 they will be further evaluated to determine if improvements are justified.

The threshold crash levels for the above improvements to reach a B/C ratio of 2.0 are as follows:

Hit Ditch and/or Rollovers – Minimum 25 ROR crashes in five years in 1,000 feet.
High Crash Curves – Minimum 25 ROR curve crashes in five years in 1,000 feet.
Shoulder Widening – Minimum 75 ROR crashes in five years in 3,000 feet.
Slippery Surfaces – Minimum 50 wet pavement crashes in 3,000 feet and a wet/total crash rate of at least 0.28 (one standard deviation above the mean).

Application of the criteria to the crash data system yields the following results.

Table 13. Impacts of System Deployment for Medium- and High-Cost Strategies

Category	Threshold Crash Level (Five Years)	Number of Statewide Clusters at or Above Threshold	Number of Targeted Five Year Crashes in the Clusters	Estimated Number of Improvements	Construction Costs ($ Million)	Type of Crash Reduced	Annual Number of Crash Types Reduced	Annual Reduction in Serious Injuries and Deaths
Re-Grade/ Flatten Ditches	25 ROR (hit ditch) in 1,000’	35	1620	28	8.4	ROR-Hit ditch	260	21
Dynamic Curve Warning Systems	25 ROR on a curve in 1,000’	40	1,250	32	3.2	ROR	60	5
Shoulder Widening	75 ROR in 3,000’	25	3,100	21	5.1	ROR	180	15
High Friction Surfaces for Slippery Pavement	50 wet pavement crashes in 3,000’ and a wet/total ratio >0.28	45	2,650	36	7.2	Wet ROR and others	120 (ROR)	10
Totals				117	23.9		620	51

If a State can only locate crashes by route, the same process specified in Table 11 can be used for medium and high cost improvements. In this case the threshold levels of targeted crashes by route suggested are as follows:

Category	Route Threshold (five years)
Re-grade/flatten ditches	40 ROR crashes
Curves	40 curve crashes
Shoulder widening	100 ROR crashes
High friction surface	75 wet pavements crashes and a wet/total ratio >0.28

Local Road Lane Departure Safety Initiatives

Almost one third of all run–off-the-road and head-on crashes occur on local roads. Unfortunately, the data for the local road system is not near as accurate and complete as that for the State highway system. The three major deficiencies are as follows:

There is no location referencing system, so crashes can be identified by a named route, but not to a specific location on the route.
There is no unique system of named routes for local roads. As such, misspellings and multiple names and spellings of the same route can array crashes for a given local road among a variety of alternatives.
There are no roadway or traffic data to complement the crash data.

From a separate non-computerized file, it has been ascertained that the average length of a rural local road is five miles and an urban local road is three miles.

The crash data system will be used as an initial tool to identify potential municipalities and routes that have concentrations of targeted crashes. As an initial step, a minor manual cleanup of local crash data was made to eliminate misspellings, multiple spellings (e.g., First Ave., First Avenue, 1st Ave.) and multiple names of local roads in which crashes where reported. This effort took between one to two person months to complete.

Since there are fewer crashes on local roads and they occur over a much larger road system, the level of concentration will be much less than that found on the State system. Three general types of lane departure strategies will be pursued on local roads as follows:

Innovative Signing and Marking for Curves. Approximately 70% of ROR crashes on local roads occur on curves. In addition, signing on local roads is considerably below the standard found on State highways. This initiative proposes to upgrade municipal-wide curve signing and marking in those municipalities that have 10 or more local road ROR or HO crashes on curves over a five-year period. This initiative will also be coupled with an unsignalized intersection signing and marking upgrade, since these types of crashes comprise approximately 25% of all local road crashes.
Selective Tree Removal on those rural local roads that have had six or more tree crashes in the past five years. (Trees are the predominant fixed object struck on local roads associated with death and serious injuries.)
Selective ROR Treatments on those local roads that have a high number of ROR/HO crashes (at least 10 over the past five years).

The initial identification of potential municipalities and local roads with sections that meet the threshold levels is as follows.

Table 14. Local Roads That Meet Threshold Levels

Category	Threshold Crash Level (Five Years)	Number of Statewide Clusters at or Above Threshold	Number of Targeted Five Year Crashes in the Clusters	Estimated Number of Improvements	Construction Costs ($ Million)	Type of Crash Reduced	Annual Number of Crashes Reduced	Annual Reduction in Serious Injuries and Deaths
Municipal-Wide Innovative Curve Sign and Marking Systems	10 or more ROR or HO on curves per municipality	560 munici-palities	10000	450 munici-palities	1.8	ROR and HO on curves	240 ROR	18
Tree Removal	Six or more hit tree per local road	235 rural local roads	1,650	186 rural local roads	5.4	Tree crashes in rural areas	260	26
Rural Local Roads with High ROR/HO Crashes	10 or more ROR or HO per local rural road	270 rural local roads	3,200	215 rural local roads	8.2	ROR and HO crashes on high ROR/HO local roads	160	12
Total				851	15.4		660	56

If the plan is approved, local technical assistance program (LTAP) engineers and local officials, using the crash data local road list above as a base, will manually assess each route. This assessment will determine if there are additional reportable crashes for the route and if there are sections within the route that meet or exceed the target crash thresholds per unit length established for system deployment on the State system. LTAP engineers will also query municipal officials to determine if they can document crash data at other locations on other routes that meet or exceed the threshold levels.

Those municipalities that have a section(s) meeting the threshold levels for a given strategy established on the State system will be asked if they are interested in applying improvements at the site. If so, a crash analysis will be performed, cost effective improvements will be identified, and cost estimates will be prepared. Federal aid funding at the 80:20 rate will be made available and the municipality must agree to assume the match.
The process for transferring funds to local governments used in the Transportation Enhancement Program is recommended for use in this safety initiative with appropriate modifications.
Only municipalities that have crash problems that meet the threshold levels and are interested in implementing improvements at these sites are eligible for this initiative.

If the crash data system for local roads does not include any information more specific than the municipality where a crash occurred, a similar process to that for the first local road strategy, municipal wide signing and marking for curves, can be applied to the latter two strategies. In these cases the suggested threshold levels of 10 tree crashes in rural areas per municipality for the tree strategy and 25 ROR/HO crashes in rural areas per municipality for the ‘local roads with high numbers of ROR/HO crashes’ are used to identify municipalities that have numbers of these crashes that exceed the threshold. For purposes of this plan, an estimate of the number of improvements can be made by selecting a small sample of municipalities that exceed the threshold levels in each of the categories. If the actual crash reports for the crashes that meet and exceed the threshold criteria are readily available for this small sample, they are a valuable resource and can be used to identify portions of local roads that have concentrations of tree or ROR/HO crashes. The municipal manager and local police can also be contacted to gain first-hand knowledge of targeted high-crash locations and the municipality’s interest in addressing them. This information can be used to estimate the number of municipalities and locations that may be improved. Once the plan is approved, this process would be extended to the remaining municipalities that exceed the thresholds.

Corridor Enforcement-Education-Engineering Initiative for Lane Departure Crashes

Within the State, there exist approximately 40 longer sections of highway where an inordinately high number of run-off-the-road severe crashes have occurred. These sections average between three and five miles in length, although a few approach 10 miles in length. Information from the crash data system has been used to identify the corridors. Two approaches to identifying corridors are provided. First, lane departure crashes involving serious injury or fatalities over the past five years have been plotted on statewide maps using geographic information system (GIS) technology. Sections of routes that have concentrations of these severe crashes become readily apparent once the maps were available. An alternative, if GIS maps are not readily available, is to develop a computer generated listing of lane departure crashes that involve serious injury or death that have occurred over the past five years by route and stationing sequentially. The listing will not be overwhelming large, and an insightful analysis and review can identify sections that have concentrations of severe lane departure crashes. The result of either analysis yields a set of 40 corridors, which have between a low of 30 to a high of 85 lane departure serious injuries and deaths over the 1999-2003 time period.

Forty sections have been identified from the crash data. Forty was selected because it approaches the upper limit of what can be effectively addressed, given human and financial restraints. This information was shared with the State Police to get their input on problem corridors. While no sections were deleted, the State Police identified five additional priority sections that, from their perspective, have combinations of high frequencies of severe lane departure crashes, very severe crashes determined from their firsthand crash investigations, and heightened frequencies of serious traffic law violations such as DUI, speeding, aggressive driving, and unbuckled drivers and occupants. These forty-five sections form the basis for the initiative. The sections are predominantly high-volume, high-speed, rural, free access arterials.

A meeting was then held with the key partners to gain consensus in participating in the initiative and establishing a game plan. Those partners included the Transportation Department, Governors Safety Representative, State Police, Chief of Police Association (representing local police), and Department of Health-EMS Division. The ‘team’ agreed to the following goals for the corridors:

Try to keep the driver on the road, primarily through the application of shoulder or edge rumble strips. Coordinate this effort with that proposed in the low cost initiative.
Attempt to address any overrepresentation of location or type of fixed object involved in these crashes.
If there is an overrepresentation of these crashes involving a specific illegal action (e.g., speeding, other forms of aggressive driving, alcohol involvement), attempt to initiate a targeted corridor specific, combined enforcement and education initiative to reduce the frequency of the illegal actions identified on the corridor.
If the death/serious injury and death crash ratio is well above the average for all the corridors or there are a high number of deaths in the corridor, apply targeted EMS system enhancements to the corridor to improve survivability. (e.g., ,What can be done to get the EMS unit to the scene quicker? What can be done to improve the capability of the arriving EMS team to sustain life? What can be done to get the victim to the proper hospital quicker?)

Once consensus was reached and coordinating/participating concerns resolved, a detailed crash analysis of each of the sections was undertaken. A good analysis is critical to a final successful outcome. Key questions asked in the analysis are as follows:

Are specific types of fixed objects overrepresented in the crashes? If so, which ones and are they throughout the corridor or limited to certain portions of the corridor?
Are rollovers a problem in portions of the section? If so, what is the cross section in the areas and is it conducive to rollovers?
Are there concentrations of these crashes on select curves or other areas? If so, where?
Is the level of illegal driving actions reported in the crash data for a given corridor overrepresented? (i.e., Does the corridor have high levels or proportions of speed related, reckless driving, illegal passing, or alcohol involvement?) If so, what are the characteristics of these crashes in terms of time of day, day of week, driver age and sex, and type of vehicle?
What percentage of the serious injuries and deaths involve an unbuckled driver or occupant? What are the total number of unbuckled serious injuries and deaths? Does the corridor have an overrepresented number of unbuckled deaths and serious injuries? If so, what are the characteristics of these unbuckled severe crashes in terms of time of day, day of week, and age and sex of unbuckled victim?
What are the numbers of lane departure deaths that have occurred over the past five years for each corridor, and which corridors are highest?
What is the ratio of deaths/deaths and serious injury crashes for each corridor and which corridors have the higher ratios?

From an engineering perspective, the following information may also be gathered:

Are there any planned roadway improvements that may impact the problem? What are they and when is it estimated they will be implemented?
Are there any known issues (e.g., pavement, bicycle, noise) associated with the application of rumble strips on the corridor? If so, what are they?
What are the possibilities of adding special signing to complement corridor targeted enforcement and education initiatives on those corridors in which safer driving behavior is being pursued?
Are there other opportunities to reduce lane departure deaths and severe injuries from an engineering perspective?

From an enforcement perspective, the following information may also be gathered for corridors that exhibit a high number or proportion of severe crashes involving any one of the specific illegal actions identified:

Is there firsthand knowledge of the key characteristics of specific illegal driver actions resulting in the severe crashes on identified corridors that can supplement the information gathered from the crash reports?
Do the illegal actions reported from the crash reports make sense in comparison with enforcement patrol experience and observations? Are there other serious concerns?
Are there major sections within the corridor where enforcement is not practical because there is no safe pull-off areas?
If a high rate of alcohol involvement exists in the corridor, are there drinking establishments within close proximity that may be contributing to the problem?
Do the times and days of frequent severe crashes align with the enforcement patrols level and tactics that serve the corridor?
Are there other opportunities to reduce the level of specific illegal actions identified on the corridor?

From an EMS perspective, for corridors that either have a high number of deaths or a high proportion of death/serious injury or death crashes, the following information may be gathered:

Are there significant difficulties in locating severe crashes, particularly at night, which have resulted in substantive loss of time in arriving at the scene? What are the root causes of those difficulties? Can they be addressed? Are there other opportunities that can improve the arrival time?
What is the level of EMS units training and capability that frequently serve the corridor?
Can air evacuation be provided on the corridor? If not, are there clearance problems that could be addressed to provide air evacuation?
Are there other opportunities to improve survivability once a severe crash occurs on the corridor?

A combined meeting and field view is then held for each of the corridors involving team agencies and their local representatives that serve the corridor. The purpose of this meeting is to review the problems, identify activities that can help others perform safety functions more effectively (e.g., maintenance clear out and stabilize pull-off areas for enforcement, reference markers that EMS units can use to locate night crashes, clear areas for air evacuation, special signing to supplement enforcement), discuss potential solutions, and reach consensus on a set of recommendations to apply to the corridor that will significantly reduce the potential of future lane departure deaths and serious injuries.

A summary of the results is shown in Table 15.

Table 15. Corridor Safety

Type of Concern	Initial Concerns from the Crash Analysis (Number of Corridors)	Corridors Recommended for Improvements After Field Views
Infrastructure	45	30
Driver – Speed	12	9
Driver – Aggressive Driving	15	12
Driver – Alcohol Involvement	15	12
Driver – Occupant Unbelted	20	16
Post Crash Survivability	13	10

The infrastructure improvements consist primarily of shoulder/edge rumble strips, appropriate sign and marking enhancements, limited shoulder improvements, fixed object removal or protection, limited guard (guide) rail enhancements, and other improvements that will assist other agencies perform their safety work on the corridor more effectively. The estimated cost of these improvements on the 30 corridors is $3.0 million.

The driver-oriented approach consists of a set of enforcement/education/engineering initiatives on each of the identified corridors listed above. There are a combined 32 corridors that have at least one driver-oriented safe behavior initiative. It is assumed that at least 30 person hours of dedicated, highly visible, and active enforcement (including checkpoints) per corridor per week is needed. The education component cannot be a conventional area-wide or targeted population approach because a large portion of the drivers come from and go to destinations beyond the corridor. Fixed or variable signs along the corridor that reach all of the affected motorists will be used. These signs will: (1) alert drivers of the specific problem (aggressive driving, high DUI crash area, etc.), (2) advise drivers what to do about it (slow down-save a life, don’t tailgate, DUI-you cant afford it, and buckle up), and (3) advise drivers of the consequences of illegal actions (targeted enforcement zone). Follow-up radio and newspaper articles on the effort, highlighting information on citations issued in the corridor and checkpoints held will also be developed and issued. The cost of the enforcement effort is 30 hours per week times an average $50 per enforcement hour times 52 weeks equals $76,000 per corridor each year. The engineering/education component, including signs, is estimated to cost $24,000 per corridor. Thus the implementation of this initiative on the 32 corridors is estimated to be $3.2 million.

The EMS enhancements consists primarily of technical/infrastructure upgrades to improve the time and reliability of EMS units arriving at the scene, increased capability of servicing EMS units to sustain life, and facilitation for air evacuation of life threatening injuries. The estimated costs of these improvements are $1.0 million.
The average effectiveness of this combined improvement is estimated to be 20%.

Table 16. Corridor Safety Improvements – Estimated Reductions and Costs

Number of Corridors Identified	Number of Corridors Recommended for Improvement	Total Number of Five Year Serious Injury and Death Crashes on Corridors to be Improved	Annual Estimated Reduction of Serious Injuries and Deaths on Improved Corridors	Construction Cost	Enforcement Education Costs (Annual)	EMS Costs (Annual)
45	30	1,150	46	3.0	4.2	1.0

Statewide Initiatives to Improve Safe Driving Behavior

Since 62% of all serious injuries and deaths are from lane departure crashes, it is anticipated that statewide action plans to reduce drinking and driving, (stronger DUI laws and more sobriety checkpoints) increase safety belt usage, decrease aggressive driving, promote young driver safety, and improve rural EMS will also help reduce the statewide number of lane departure serious injuries and deaths. A short discussion of these initiatives and their impact on lane departure follows.

Reduced Drinking and Driving

The State currently has a 0.08 law and conducts a limited number of sobriety checkpoints. This initiative calls for much stronger penalties for BAC levels above 0.15, second and multiple DUI offenses, and drivers with suspended/revoked licenses who continue to drive. In addition, the sobriety checkpoint program will be greatly expanded from approximately 50 to 1,000 checkpoints conducted annually. These efforts will be complemented by a major statewide marketing and education campaign.

Increase Safety Belt Usage

The 2003 State safety belt use rate is 75%. The proposed action plan calls for enactment of a primary safety belt law and a significant expansion of the ‘Click It or Ticket’ campaign to four additional periods of the year.

Aggressive Driving

The primary thrust of this initiative is to reduce the levels of excessive speeding on rural interstates and rural major and minor arterials.

Young Drivers

The primary initiative in this area is to legislate a graduated driver licensing system for new drivers. The proposed legislation will be based upon the model law and include provisions that require parent or guardian driving supervision for the first six months of permit, an 11:00 PM curfew for provisional licensees, increased penalties for traffic violations during the provision phase, and only one peer passenger during the provisional phase.

Rural Emergency Medical Services (EMS)

The rural EMS action plan will dramatically reduce the amount of time from when a crash occurs to the arrival time of a patient to the hospital. A much more rapid detection system using cell phones, global positioning satellite (GPS) systems, and other technologies and a more time efficient evacuation system to the hospital (primarily helicopter evacuations) are the primary components of the plan. This initiative is expected to save a considerable number of lives in remote rural areas. It will have no or minimal impact on the combined serious injury and fatal number since nearly all of the people who live (who would have otherwise perished without the EMS enhancements) will suffer a serious injury.

Other Actions

Pavement Edge Wedge – As part of all 1R and 3R projects which do not have paved shoulders, the sloping 45 degree pavement wedge will be applied to all pavement edges that butt up to stabilized shoulders or do not have any shoulders. This is considered an experimental strategy and will proceed through an evaluation phase prior to full deployment.
Snow or Ice Covered Pavements – Five-year ROR and head-on crashes that involve skidding on snow- or ice-covered pavements have been analyzed to determine sections of highway where concentrations of these crashes occur. The criteria for these locations is a combination of 10 or more snow/ice crashes in a five year period in 5,000 feet (about 15 times the average) and a snow or ice/total crash ratio of 0.18 or greater (one standard deviation above the mean). Statewide, 246 sites have been identified that meet or exceed these criteria and have been GIS mapped and given to maintenance personnel for enhanced snow plowing operations during the winter months.
Maintenance of Clear Zones – If clear zones are not intermittently and systematically maintained, they may be compromised by vegetative growth over time. A maintenance plan will be developed to prevent compromises to the clear zones by vegetative growth. Initially an inventory of clear zones on the interstates and rural major and minor arterials will be compiled. Routes on which the clear zone is already maintained by mowing operations will be noted. For those routes not maintained by mowing, a cycle will be established to periodically inspect each clear zone and remove any vegetative growth that may grow beyond four inches in diameter before the next inspection. As an initial starting point, an inspection cycle of three years and any growth exceeding one and one half inches in diameter will be removed. The criteria may be adjusted, as the growth rate of vegetation is better known.
Urban Landscape Policy – An internal team of landscape engineers, safety engineers, highway designers, and environmentalists has been formed to develop a urban landscape policy that addresses the aesthetic, environmental, and safety needs of urban highways. The team has been charged with giving particular emphasis to establishing guidance for tree placements in urban settings that is both aesthetically pleasing and safe for motorists. Recognizing the relationship between speed and severity when trees are struck (from the crash data, for tree crashes there are two fatalities per 1000 tree crashes on 25mph streets, seven fatalities per 1,000 tree crashes on 40mph highways, and 15 fatalities on 55mph highways) traffic calming will be incorporated into the policy.

Unfortunately, no estimates of the effectiveness of these initiatives can be made at this time.

Summary of Proposed Initiatives

The previous analyses indicate that a significant reduction in lane departure deaths and serious injuries is achievable through the widespread deployment of a number of cost effective strategies. A compilation of candidate programs needed to achieve the goal indicates that an estimated funding level well beyond the existing levels (Table 17) is needed to prevent nearly 300 more annual lane departure deaths and serious injuries.

Table 17. Tabulation of Costs and Impacts by Category

Category	Construction Cost ($ Million)	Enforcement, Education and EMS Costs (Annual $ Million)	Annual Lane Departure Serious Injuries and Deaths Reduced
Low-Cost System Deployment	88.5		173
Medium/High-Cost System Deployment	23.9		51
Local Roads Safety	15.4		56
Corridor Safety	3.0	5.2	46
Total	130.8		326

Key Strategic Direction Issues

Successful implementation of this initiative requires upper management leadership guidance in three major areas: funding, broad policy actions, and key next steps. A brief discussion of the issues in these areas follows.

Funding

If a substantive reduction in lane departure deaths and serious injuries is to be achieved, funds above existing levels are needed to deploy cost effective strategies as defined in the plan. This plan estimates that approximately $25-30 million annually will be needed on the State highway system, $4 million annually for local government safety improvements, and $5 million for corridor driver safety for each of the next four years to fund the level of cost effective improvements needed to prevent over 300 lane departure deaths per year over the next five years. The numbers may fluctuate somewhat after the next phase of development is completed, when site visits of the cluster locations occur, and more precise information on the site conditions, applicability of the strategy, and costs are known. At this point in time, however, it is safe to conclude that some level of additional funding will be needed if we are to realize a substantive reduction of lane departure deaths and serious injuries. Considering the competing needs of other priorities, and recognizing the potential for some additional safety funding in the upcoming federal reauthorization, a level of additional funding will need to be made available so that this important initiative can continue.

Broad Policy Actions

Four areas where policy actions are needed to effectively implement the plan are as follows.

Systematic Deployment of Cost Effective Improvements – Traditionally the safety improvement program has concentrated on addressing the highest crash locations in the State. Each year, between 50 and 70 safety improvements have been implemented, mainly at intersections and curves. For the most part, significant reductions of crashes have occurred at these sites. However, because of the relatively small number of improvements, the reduction in statewide numbers of deaths and serious injuries is almost negligible. Because this effort effectively addresses a number of high crash locations, it needs to continue. However, if we are to lower the statewide numbers of deaths and injuries, an additional effort of systematically applying low-cost, cost-effective improvements at a large number of locations with documented crash histories needs to be pursued. Carefully administered, systematic deployment of key low-cost improvements can produce results of such a magnitude that the statewide numbers of targeted deaths and serious injuries can be reduced. Many of these low-cost improvements have not been used extensively in the State. It is proposed that our approach to systematic deployment proceed both cautiously and aggressively to expeditiously resolve design and application issues, minimizing the potential for noticeable controversy while insuring program cost-effectiveness. Once we become comfortable with a specific low-cost improvement and have resolved implementation issues, it is proposed that we launch County- or District-wide contracts to implement designated improvements at large numbers of locations that can be favorably impacted by the improvements.
Center/Shoulder/Edge Rumble Strips and Hazardous Tree Removal – Wide use of these two key strategies, which are rarely applied in the State, are both very important in attaining the goal and potentially controversial. Minimizing controversial actions/decisions while implementing the improvements as extensively and cost-effectively as possible will be the keys to successful implementation.

With rumble strips, it is important that we are sensitive to potential noise issues of residents and riding activity level and issues of bicyclists. It is suggested that we assemble a team including the State Pedestrian/Bike Coordinator, a designer, maintenance manager, and pavement design engineer to resolve design and maintenance issues and establish tentative parameters for application on the routes that have a ROR crash problem. Approximately five to 10 rumble strip sections should be installed on agreeable non-controversial route sections that have a ROR crash problem during the 2005 construction season to get some experience with the strategy. We can then expand implementation, learning from the experiences on these test sections.
With tree removal it is important that we develop and institute a process that only targets hazardous trees at locations that are not controversial. The process also needs to be able to identify vulnerable trees that are of value environmentally, to the community, or to adjacent residents. It also needs to offer alternate improvements to removal, such as tree delineation, that preserve these environmental and community values while addressing the crash problem. The safety engineer, environmental engineer, a designer, and a maintenance manager need to work as a team to develop the process.
Inclusion of Low-Cost Safety Improvements in the 1R and 3R Program Projects – Many of these low-cost improvements need to be considered for inclusion in 1R and 3R design projects, particularly if a crash problem exists within the limits of work. This is particularly important because smooth pavements resulting from the 1R and 3R improvements promote increased speeds and more severe lane departure crashes. It is proposed that in approximately 12 to 18 months after the Department has obtained more experience with most of the low-cost improvements, that the 1R and 3R design process be revisited to determine under what circumstances low-cost safety improvements mentioned in this report should be incorporated into the contracts.
Technical and Financial Assistance to Local Governments to Implement Safety Improvements at Locations with Crash Histories – Approximately one third of all serious injuries and deaths occur off the State highway system. Realistically, safety on local roads needs to improve if statewide reductions are to occur. Past experience indicates that improvements will not be implemented extensively on local roads unless federal funds and technical transfer assistance are extended to the municipality. In the past, funding to municipalities has been limited to specific programs that have targeted funds, such as the Transportation Enhancement Program. It is proposed that (1) federal funds be extended to municipalities to implement targeted, cost-effective, safety improvements at local road locations that have a crash history, (2) the Transportation Improvement Program process be adapted to provide federal funds to municipalities for this safety initiative, and (3) additional resources to the LTAP program be extended to provide the necessary technical assistance to successfully implement this component

Next Steps

Once upper management provides direction on the above funding and policy issues, implementation of the above strategies will commence in accordance with the guidance received. It is envisioned that an implementation plan with milestones and a schedule will be developed for each of the strategies.

	Skip to content

Home \| About \| Subject Index \| Events \| Publications/Multimedia \| Links \| Newsletters \| Contact