State of The Practice for Shoulder and Center Line Rumble Strip Implementation on Non-Freeway Facilities

CHAPTER 5. INFORMATION GAP ANALYSIS AND ACTION PLAN

PURPOSE OF THE GAP ANALYSIS

The literature review and current practice review revealed several topics for which there is insufficient information. Additionally, research has been conducted for several areas related to rumble strips, but the state of knowledge has not converged on a consensus. The purpose of this section is to summarize the identified gaps, while the action plan identifies tasks and objectives necessary to fill in the knowledge gaps. The following is a list of identified gaps by topic area:

• Rumble strip design: The project team identified the following gaps for rumble strip design:
  
  
  • Departments have used a wide variety of designs; however, there is little evidence supporting the use of one design over another. Most evidence suggests that larger rumble strip dimensions (i.e., length and depth) increase the change in sound level over environmental noise inside and outside the vehicle; however, no studies have identified a safety differential based on rumble strip dimensions. Research has not converged on the best rumble strip design given constraints, including paved shoulder width, external noise pollution, and bicycle traversability.
    
    
  • The optimal offset for SRSs on non-freeway facilities has not been identified. NCHRP Report 641 noted that there is conclusive evidence that SRSs placed closer to the edge line are more effective on freeways.⁽⁸⁾ However, non-freeways facilities have more considerations, including accounting for the trade-off between safety and the potential for incidental contacts. The optimal placement for the highest safety benefits with the lowest potential for incidental contacts is unknown.
    
    
  • NCHRP Report 641 indicated that rumble strips with smaller dimensions appear to be adequate for generating noise and vibration for passenger cars, while larger dimensions may be necessary for alerting drivers in heavy trucks.⁽⁸⁾ Policies and standards that were reviewed did not discuss the potential for larger dimension rumble strips on roadways with heavy truck traffic. Additionally, policies did not discuss when a larger dimension rumble strip should be used if it was needed for trucks (i.e., a truck volume warrant).
    
    
• Noise and vibration: The project team identified the following gaps for noise and vibration:
  
  
  • Departments have made assumptions on the minimum level of sound necessary to alert a drowsy or inattentive driver; however, no research has been conducted to determine whether these assumptions are valid.
    
    
  • Due to the complexity of issues related to noise level, the literature review did not identify a standard method for assessing noise internal or external to the vehicle. The issue is complex due to the varying levels of ambient noise, nature of noise generated from rumble strips, and impacts of vehicle and tire characteristics on noise generation. For internal noise, most studies used very similar equipment, but there were differences in the test speed and location of the sound level meter.
    
    
  • Many departments struggle with handling noise impacts, defining locations where external noise will result in complaints, and managing the trade-off in providing gaps near residences and resulting offsets in terms of safety effectiveness. There is little consensus on how far rumble strips should be located away from residences, and residence density that may be required for gaps in the rumble strip pattern should be considered.
    
    
  • Departments have been using two basic principles for mitigating noise. The first is to discontinue rumble strips in areas near residences or in areas where they will be inadvertently struck (i.e., intersections and driveways). The second is to use rumble strip patterns that reduce external noise through reduction in rumble strip length or depth or by using an alternative design such as sinusoidal. The safety impact of this decision is unclear. Discontinuing rumble strips results in the greatest reduction in external noise but has the potential for the greatest reduction in rumble strip effectiveness. Reducing the noise level generated by rumble strips will still create external noise but has the potential to provide a continuous benefit for preventing roadway departure crashes.
    
    
  • Several departments noted that different patterns for CLRSs and SRSs are specified to differentiate between the two. However, this is not a consistent practice, and no studies have shown that drivers understand this to be the case for those States that do so.
    
    
• Bicyclist, motorcyclist, and pedestrian impacts: The project team identified the following gaps for bicyclist, motorcyclist, and pedestrian impacts:
  
  
  • As with noise mitigation, there have been two general principles for altering rumble strip patterns for bicyclists. The first is to create gaps in the pattern for bicyclists to cross, and the second is to use an alternative pattern that is more bicyclist friendly. It has not been substantiated how these changes impact rumble strip safety effectiveness.
    
    
  • There is little research examining adequate shoulder width from the rumble strips to the outside of the shoulder. None of the studies recommended a minimum shoulder width; however, the AASHTO guide did recommend a minimum shoulder width of 4 ft.⁽²⁶⁾
    
    
  • No research studies to have studied the impacts of rumble strips on pedestrian or bicyclist safety. No CMFs have been developed for vehicle-pedestrian or vehicle-bicycle crashes, and the same is true for vehicle-motorcycle crashes.
    
    
• Pavement condition impacts: The project team identified the following gap for pavement condition impacts:
  
  
  • Most assessments of pavement condition are anecdotal in nature. There is little quantitative research identifying the impacts of rumble strips on pavements, particularly longitudinal joints, yet departments still often struggle to implement rumble strips in these locations. Additionally, further research is needed to more fully understand the impact of pavement depth, especially for thin overlays and chip seals.
    
    
• Pavement marking visibility: The project team identified the following gap for pavement marking visibility:
  
  
  • Research studies have not conclusively determined the impact of ELRSs and profiled thermoplastic pavement markings on wet nighttime pavement marking visibility. Smadi and Hawkins noted that very few States have measured wet retroreflectivity of rumble stripes.⁽⁹⁾ Further research is needed to conclusively assess pavement marking visibility, particularly rumble stripes.
    
    
• Safety effectiveness: The project team identified the following gaps for safety effectiveness:
  
  
  • While the aggregate benefits of CLRSs and SRSs/ELRSs have been clearly demonstrated for standard installations of 12- and 16-inch rumble strips, departments are using alternative designs for reducing impacts on other roadway users and for reducing external noise. Research studies have not established whether there are trade-offs in safety benefits for narrower rumble strips, designs with gaps for bicyclists, or shallower “bicycle-friendly” rumble strips. Additional insights would also allow decisionmakers to consider shallower rumble strips for thin overlays if research shows no loss of safety effectiveness.
    
    
  • The safety effectiveness of profiled thermoplastic pavement markings or reflective pavement markers as rumble strips should be established to determine how the benefits compare to milled ELRSs.
    
    
  • Few of the safety studies reported the dimensions of the evaluated rumble strips. In the case of Torbic et al., the researchers noted that all designs were considered together because it was difficult to verify the dimensions of the strips being studied.⁽⁸⁾ Future work should further examine the relationship between rumble strip design and safety effectiveness.
    
    
• Development of an implementation tool: The project team identified the following gaps for the development of an implementation tool:
  
  
  • Departments observed different needs and uses for a rumble strip tool to help with installation decisionmaking. Departments that have set design standards and clear policies for rumble strips seem more disinclined to the development of the rumble strip tool, whereas departments that were still developing or looking at research were more apt to accept the idea of a tool. Some departments, such as FHWA’s EFL, struggle with justifying the installation of rumble strips and have little to no rumble strips in their area. Alternatively, LaDOTD has clear policies and little pushback on rumble strip installation, which leads them to find little use for a tool that will help inform decisionmaking with a lack of hard evidence to back up decisionmaking. Although there are vast differences between the departments surveyed, there is a common theme that better information would help inform decisionmaking. Additionally, the interview departments indicated that there is rarely an instance where a blanket solution exists. Rumble strip installation typically involves the consideration of several users and the use of crash history and may require committee recommendations or vetting with outside groups, such as bicyclist groups. All departments indicated that they would be willing to consider a rumble strip tool, but successful departments’ needs differ from those of struggling departments.
    
    
  • Additionally, departments identified the following needs for a rumble strip implementation tool:
    
    
    • The ability to make optimized recommendations based on roadway and roadside characteristics.
      
      
    • The ability to exist as a warehouse or inventory for maintaining rumble strip data and mapping capabilities. This would allow the department to know where rumble strips are already in place and to conduct effectiveness evaluations more easily.
      
      
    • The ability to incorporate SPFs and CMFs to rate the effectiveness of potential installations. More positive evidence increases buy-in.
      
      
    • The ability to recommend alternatives to rumble strips for situations where rumble strip installation cannot be accommodated.

ACTION PLAN

The action plan serves as a list of objectives that are grouped under overarching goals. Ordering of goals and objectives are not sequential in nature; in most cases, projects can be executed in any order. However, some objectives should be ordered. For example, bicyclist- and noise-related performance measures should be established before researching the ideal design dimensions for further implementation. The following list provides overarching goals and objectives or potential future project objectives necessary to fulfill the goals:

• Goal 1—Establish safety effects of rumble strips: This goal includes the following objectives:
  
  
  • Objective 1: Identify safety impacts of rumble strip dimensions and determine whether decreasing dimensions reduces the safety effectiveness or whether crash reductions are retained. This research can use the many miles of rumble strips that are already in place that have not been evaluated. The findings of this research are important to determine whether larger rumble strips are necessary to achieve reductions or whether smaller rumble strips will suffice. This is important for research related to special considerations (i.e., noise and bicyclists) and for interactions with roadway geometry (i.e., narrow rumble strips on narrow shoulders).
    
    
  • Objective 2: Identify safety impacts of noise-reducing rumble strips. Departments have been using shallower rumble strips as well as sinusoidal rumble strips to reduce external noise pollution. No research studies to date have identified the safety effects of shallower rumble strips or sinusoidal rumble strips as alternatives to more traditional design dimensions.
    
    
  • Objective 3: Identify safety impacts of rumble strips on bicyclists and identify crash-based optimal design of bicycle gaps if available (perhaps gaps can even be larger than current?). It is unclear how rumble strip placement and rumble strip bicycle gaps affect vehicle-bicycle crashes. Additionally, the safety effectiveness of bicyclist-friendly rumble strips is unclear.
    
    
  • Objective 4: Identify relationship between rumble strip safety effectiveness and roadway geometry. As with rumble strip placement (i.e., offset), future research should determine whether effectiveness varies by shoulder width and should determine whether there is an interaction between rumble strip dimensions, placement, and shoulder width. Additionally, future research should identify whether there is a differential effect on horizontal curves and horizontal tangents.
    
    
• Goal 2—Identify performance measures for noise for trade-off analysis: This goal includes the following objectives:
  
  
  • Objective 1: Develop a standardized testing method for external noise (i.e., vehicle speed and location). A standardized method may help departments develop external noise models and identify rumble strip design criteria based on proximity to residences. Noise models will need to consider the issues related to ambient noise level, characteristics of noise generated by rumble strips, and complexities associated with vehicle and tire characteristics and their impact on noise generation.
    
    
  • Objective 2: Smadi and Hawkins noted that only two departments have specifications for audible and palpable warnings.⁽⁹⁾ Research should consider how powerful these warnings need to be to alert drowsy or distracted drivers at a minimum and how they can best be measured. Additionally, research should determine whether rumble strips can produce an optimal alert and whether a different pattern for the center line and shoulders can be readily apparent to distracted or drowsy drivers.
    
    
  • Objective 3: Quantify noise threshold and/or frequency of strikes for nearby residences. Defining these standards may help departments more successfully install the optimal design near residences. The optimal design should maximize safety, reduce noise as much as possible, and result in a level of incidental contacts that may be tolerable by nearby residents. Developing models for incidental contacts may help optimize rumble strip offset based on the roadway’s geometric characteristics. In combination with assessing the safety impacts of rumble strip offset and shoulder width, an optimal location may be found.
    
    
• Goal 3—Identify performance measures for bicyclists for trade-off analysis: This goal includes the following objectives:
  
  
  • Objective 1: Develop performance measures for bicyclists for rumble strip trade-off analysis. Crash reduction and noise level changes can be quantified, but it is difficult to determine how rumble strips (or lack thereof) impact bicyclist activity or safety. Future research should develop performance measures that can objectively be weighed against other factors, such as rumble strip installation or shoulder widening with rumble strip installation for alternatives analyses.
    
    
  • Objective 2: Quantify high versus low or medium bicycle usage and methods for data collection. Departments often use bicycle usage as a factor for making the decision on whether or not to install rumble strips on narrow shoulders. However, there are no guidelines for determining how much activity would preclude a narrow shoulder for installation. Quantifying use may also help with alternatives analysis for higher cost solutions, such as combined shoulder widening and rumble strip installation.
    
    
• Goal 4—Establish effects of pavement condition and depth on deterioration: This goal includes the following objectives:
  
  
  • Objective 1: Quantify performance measures for pavement condition. Many departments specified that the pavement must be in good condition or it will not be resurfaced within the next few years for rumble strips to be placed. However, few States had specific measures of “good condition,” and it is unclear whether rumble strips have a negative impact on already poor pavement that is scheduled for resurfacing.
    
    
  • Objective 2: Assess the impacts of rumble strips on pavement deterioration. This objective was identified in NCHRP Report 641 and by Smadi and Hawkins.^(8,9) Smadi and Hawkins note that several departments have electronic databases with locations of rumble strips. This information can be used with pavement condition data to model the relationship between pavement deterioration and rumble strip presence.
    
    
  • Objective 3: Identify the maximum rumble strip depth by depth of the pavement top layer. Departments differ on the minimum depth of the top layer of pavement for rumble strip installation. Further guidance may be useful for considering the rumble strip dimensions that may be used by the pavement depth existing on the roadway.
    
    
• Goal 5—Develop an optimization tool based on goals 1–4 and existing rumble strip implementation policies: This goal includes the following objectives:
  
  
  • Objective 1: Identify preferable platform for housing the optimization tool. A Web-based tool may be easier to update with new information, but a spreadsheet tool may be easier for users to navigate.
    
    
  • Objective 2: Develop decision logic for navigating user inputs and providing recommended rumble strip or alternative treatments. The logic should be based on best practices used by successful departments with performance measures and optimization algorithms developed based on research in goals 1–4.
    
    
• Goal 6—Identify impacts of current in-vehicle technologies and future in-vehicle technologies on rumble strip needs: While not directly addressed in this research, future in-vehicle technologies may impact the need for or effectiveness of rumble strips. With lane-keeping technology, the vehicle may alert the driver when the driver is leaving the travel lane. The effectiveness of this technology will depend on market penetration and the reliability of the assistive technology. This may eliminate the need for rumble strips altogether or may result in more conservative application for noise sensitive or recreational areas (i.e., areas with bicyclist activity).
  
  
• Goal 7—Identify the impacts of rumble strips on driver behavior: Researchers can use the second Strategic Highway Research Program Naturalistic Driving Study data to identify how the presence of SRSs, CLRSs, and their combination impact in-vehicle behaviors. Additionally, the data can be used to look at variations in rumble strip dimension and roadway geometry to further identify optimal rumble strip designs by roadway geometry.