Developing a Guide for Quantitative Approaches to Systemic Safety Analysis (2020)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

31 Agency Survey of Practice This section summarizes the current state of practice of systemic safety analysis. Early in the research project a web-based survey was conducted that targeted state, county, and local safety engineers to learn about the experiences of highway agencies implementing quantitative approaches to systemic safety. As a follow-up to the online survey, the research team conducted telephone interviews with several agencies to gather more detailed information regarding each agency’s experience implementing quantitative approaches to systemic safety analysis. The results of the survey are summarized below. A web-based survey was administered to all 50 U.S. state highway agencies and approximately 300 county and local highway agencies to learn about their experiences in implementing quantitative approaches to systemic safety. Seventy-six agencies responded to the survey, and only one response was provided by each agency. Responses to categorical questions are summarized by showing both the percentage of the responses and the frequency/number of responses shown in parentheses; and for those questions where it makes sense, the categorical responses are ordered from the highest response to the lowest response. For those questions that asked agencies to further explain an issue, verbatim responses are provided in bullet form. Not all respondents answered all of the questions. Survey Results: 1. What type of agency do you represent (please check one)? State DOT 57.9% (44) County agency 25.0% (19) City or municipal agency 14.5% (11) Metropolitan planning organization 2.6% (2) Federal agency 0.0% (0) Other (please specify) 0.0% (0) Totals 100.0% (76) 2. What is your current position for the agency that you work for (please check one)? Safety engineer 49.3% (37) Traffic engineer 17.3% (13) Highway designer 2.7% (2) Planner 5.3% (4) Other (please specify) 25.3% (19)

32 3. What types of safety analyses does your agency perform in deciding what site-specific safety improvements should be made (check all that apply)? Crash-Data-Based Analyses—methods that are intended to identify high-crash locations by analyzing site-specific crash data (e.g., finding the locations with the highest observed crash frequencies or rates) 94.6% (70) Systematic or Systemwide Improvements—methods that involve making specific infrastructure improvements for all similar sites on a road network based on established policies, guidelines, or design criteria (e.g., adding shoulder rumble strips at all locations that meet certain established criteria) 71.6% (53) Systemic Safety Analysis—methods that use risk-based or predictive models to compare locations and select those with the greatest potential for safety improvement (e.g., FHWA Systemic Safety Analysis Tool, AASHTOWare Safety Analyst software, usRAP Tools or ViDA software) 43.2% (32) Other (please specify) 6.8% (3) Other Responses:  FHWA focus state roadway departure (RD) analysis that was crash threshold and inventory risk base.  Corridor based analysis, using rates.  Transit.  Alternatives comparisons. 4. If you checked YES to Systemic Safety Analysis in Question 3, which systemic analysis tools does your agency use (check all that apply): Yes No Not Familiar With This Tool FHWA Systemic Safety Analysis Tool 36.1% (13) 50% (18) 13.9% (5) AASHTOWare Safety Analyst software 24.3% (9) 59.5% (22) 16.2% (6) usRAP Tools or ViDA software 20% (7) 51.4% (8) 28.6% (10)

33 5. If you checked YES to Systemic Safety Analysis in Question 3, the following safety analysis tools may have some systemic safety analysis features. Does your agency use any of the following tools (check all that apply): Yes No Not Familiar With This Tool Highway Safety Manual Part C 68.4% (26) 21.1% (8) 10.5% (4) Plan4Safety 0% (0) 36.1% (13) 63.9% (23) CARE Correct 0% (0) 34.3% (12) 65.7% (23) DiEXSys Vision Zero Suite 8.3% (3) 41.7% (15) 50% (18) FHWA Pedestrian and Bicycle Safety Crash Analysis Tool (PBCAT) 8.1% (3) 62.2% (23) 29.7% (11) 6. If you checked YES to Systemic Safety Analysis in Question 3, are there any other systemic safety analysis tools, not included in Questions 4 and 5, that your agency uses? Yes 50.0% (19) No 50.0% (19) Other Responses:  We have developed an internal risk-based approach similar to the FHWA Systemic Tool but not quite the same.  Internal safety analysis.  In-house software.  Internally developed tools.  Tools created in-house to analyze crash data.  In-house procedures.  Part C spreadsheets.  Interactive Highway Safety Design Model (IHSDM).  We used FHWA RD focus state crash density threshold and inventory risk analysis. Also, use Part B intersection and segment SPF EB potential for safety improvement (PSI) and are embarking on developing for RD and possible other intersection characteristics.  Hierarchical Bayesian model.  Total risk factor weight analysis.  We use traditional network screening method using the crash rates of different collision types throughout the state, prioritize, and treat.  Excel analysis on frequency.  Graphical information system (GIS), Access, Excel tools for systemic implementation.

34  MS Excel, mobility (by County Road Administration Board), and information provided by State DOT.  Excel and GIS.  We utilize crash and road inventory data to determine risk factors and select locations for improvements. We do not use specific tools.  Manual analysis. 7. If you checked YES to Systemic Safety Analysis in Question 3, how long has your agency been using the systemic safety analysis tool(s) identified in your response to Question 4, 5, or 6? Less than 1 year 8.8% (3) 1 to 5 years 76.5% (26) More than 5 years 14.7% (5) 8. If you checked YES to Systemic Safety Analysis in Question 3, please describe the primary steps or process your agency follows in implementing systemic safety analysis (please include a link to online documents/reports if available): Responses:  We have developed a series of safety plans that highlight locations for safety consideration and a suggested treatment based on a risk assessment. Plans are available upon request - 87 counties and 8 districts (2 versions of district plans).  Shared FHWA RD route segments by countermeasure in XLS and KMZ map files. Annually publish PSI for intersections (signalized and stop controlled for spot and systemic) and segments (more spot by overlay with RD). Used RD wet and curve crash data for friction testing data collection for resurfacing and high friction surface treatment (HFST) scoping (presently testing to develop new continuous friction measurement testing for systemic pavement treatment).  We develop internal lists annually that involve a systemic analysis. Our Districts are required to evaluate the lists and include the strategies to implement in their District Safety Plans.  Does it align with SHSP and RD emphasis area? Use "Crash Trees" to show breakdown of fatal and serious-injury crashes by facility type to prioritize projects.  Determine 1. Crash total 2. Crash over representation 3. Crash under representation.  Identify target crash type, contributing factors, then sites.  Identify crash contributing factors through an evaluation of crash data (serious injuries and fatalities), select specific countermeasure, and develop list of locations with similar contributing factors.  Review statewide rural crash data. Identify crash types with largest percentage. Using crash data reported by officer, pear down road characteristics to find most common contributing factors. For example, single-vehicle crashes, curves, downhill.

35  Network screening for specific crash types that fit specific countermeasures (e.g., RD Plan, Intersection Plan), Modified usRAP (draft study status).  We are currently early on in the process, having implemented systemic safety projects for centerline rumble strips and horizontal curve signs. Currently, we look at a variety of contributing factors and find candidate locations which have high amounts of contributing factors and then include them in our projects.  Our agency has developed risk-based tools as needed.  Our systemic treatments generally are limited to installing centerline rumble stripes, usually in response to a request.  Annual in-house review of countywide crash data using GIS.  Manual risk-based analysis to identify priority locations for systemic treatments.  Working on intersections in Safety Analyst. Building inventory as we go and analyzing even with the limited locations we have.  We do not have a unique systemic process. The primary steps followed are integrated with our overall HSIP project process.  Provide safety features at locations whenever possible because we do not know where the next driver is going to be impaired, distracted, or fatigued.  Analyze crash data, look for potential projects with mitigation strategies, do an HSM analysis, prepare crash reduction factors, conduct economic and cost /benefit analysis, and coordinate with local entities.  Network screening, diagnosis, countermeasures, prioritize, implement.  We use a combination of roadway data and crash to prioritize locations for treatment.  One high-risk rural road (HRRR) project.  Looking at crash data.  Identify locations that show a benefit-cost ratio greater than 1.0 for the countermeasures chosen.  Intersections and or segments are determined through AASHTOware Safety Analyst then they are mapped (GIS) for visual understanding. There is not an official online document that details that process.  Most use fitted negative binomial models for target crash types/facilities. Implementations vary.

36 9. If you checked YES to Systemic Safety Analysis in Question 3, what advantages has your agency derived from implementing systemic safety analysis (check all that apply):  Being able to program projects further into the future as projects are based on risk factors (i.e., roadway characteristics) that do not change as frequently from year to year as crash history 59.4% (19)  Observed reduction or trend in the number of severe target crash types, especially on focused facilities where improvements were made 56.3% (18)  Safety funds are more equally distributed either regionally or across jurisdictions compared to programming safety improvements based solely on traditional crash-data based method 43.8% (14)  Being able to estimate the safety effectiveness of individual countermeasures implemented as part of the systemic safety analysis program 40.6% (13)  Other (please specify) 37.5% (12) Other Reponses:  We now have projects into the future 6 yr construction program for systemic countermeasures.  We recently completed the centerline rumble strip project on state roads. We will evaluate in a year and also in three-year period.  Implemented policy based on safety improvement benefits.  There has been an enhanced understanding of what can be done with enhancing safety on a systemic basis versus spot treatments.  Using contributing factors makes it easier to explain and justify why we want to install signs on low-volume, town maintained roads.  Better communication to the public (we are being proactive and not waiting for something to happen). The public often talks in terms of risk and this approach correlates to that.  Redirected funds from less efficient to more efficient programs.  Really just getting started; hard to tell.  Too early to tell.  Not quantified.

37 10. If you checked YES to Systemic Safety Analysis in Question 3, what are some of the lessons your agency has learned in implementing systemic safety analysis: Responses:  Observationally, it appears that the more locations you can touch, the better the performance in terms of fatal and serious-injury crash reduction.  There is a benefit in helping identify design policies/guidelines that were not developed with safety as a primary consideration (i.e. raised medians, shoulder widths, clear zones, etc.). Takes additional time to educate Region/District staff on the justification of spending funds on a project that may not have a crash history.  Coordination with municipalities, especially rural locations that are not used to seeing signs, delineators, rumble strips, etc. is very important. Also, the general public does not yet grasp the meaning/benefits of systemic improvements. More education is necessary.  To make public aware of the project in advance.  Sometimes it is difficult to justify to management the value of systemic applications unless a thorough state-specific analysis can be provided, which is often difficult if there are insufficient sites to model.  It helps convincing the upper management to develop policies for systemic treatments; in our case we were able to develop policies for rumble strips and cable median barriers.  Need for outreach to explain how this is a data-driven approach.  The initial round of requests for satellite offices to submit for systemic efforts was met with confusion. An example of a situation in which systemic type fixes were applied and worked should have been sent with the initial request to do systemic safety.  Public response can be very positive or very negative.  Repetitive education is necessary because it is not intuitive to people to put safety measures where little to no crashes may have occurred. Must continuously explain risk concept.  The process and approach continues to evolve as more data become available for analysis.  Different programs require different methods.  Probably fewer contributing selection factors, checking overlaps as to not over-represent a specific factor. Data - use what you have and build from there. Check trends with injury crashes especially as one dives further into the data and is using fewer/fewer fatal/serious crashes to base decisions.  Even without perfect data on locations, it is possible to do systemic analysis, but it requires more individual work/investigation for engineers.  We used process for the first time last year. We need better ways of identifying the problem curves in order to make this approach more robust.  We are currently creating an Electronic Linear Referencing system on our off system streets to utilize pattern recognition on all road segments.

38  Take care in initial database development segmentation. Make appropriate determination of data collection methods and precision needed to achieve goals of the program.  Need for improved road inventory information.  Having a complete road inventory and crash data helps.  The need for good roadway and crash data.  Having national crash research helps implement new countermeasures.  It is difficult.  We want to make intersections and Safety Analyst work; but it is going to take a lot of work collecting all the data.  Currently involved in a rural road safety project, but it has not been implemented yet.  We are decentralized, so not sure Districts used detailed data in their selection. Sometimes they like to generate their own analysis.  Many of our safety countermeasures rely on other projects in the STIP and timing of implementation is often dictated by this.  Too early to tell. 11. If you checked YES to Systemic Safety Analysis in Question 3, do you quantify the benefits of implementing a systemic safety analysis program/policy: Yes 54.8% (17) No 45.2% (14) Other Responses:  usRAP calculates an overall and individual project safety benefit (cost and serious-injury crash reduction).  We use b/c analysis, but are also working on developing a severe crash reduction from each safety project to be determined.  Quantify the cost/benefit.  Cost/benefit.  Typically done through performance reporting.  Plan to perform before-and-after analysis.  In terms of reduction in crash cost.  Have performed EB analysis of certain systemic applications to justify and further the use of such countermeasures.  Wilcoxson signed-rank and EB before-after.  Cable median barrier installation was evaluated for effectiveness; Databases and processes are being developed to evaluate effectiveness of rumble strip and HSIP low- cost improvement systemic applications.

39  Naive before-after analysis.  We are in the process of implementing a statewide systemic lighting retrofit project for bicyclists and pedestrians. As part of this project, we will be doing an evaluation to quantity the benefits.  We are trying to get an effort together to better track the locations where implemented. We are doing better for some countermeasures than others.  We have not done this yet, but I believe we would at least do a before-and-after analysis of each site that is being treated.  We have a very general time-of-return form that is filled out.  Using CMFs and HSM crash costs for different severity levels.  We have not for anything other than policy-based implementation yet because too early to tell.  Insufficient manpower to evaluate most programs. 12. If you checked YES to Systemic Safety Analysis in Question 3, does your agency track the implementation of the safety improvements: Yes 87.9% (29) No 12.1% (4) Other Responses:  We track what HSIP funds are spent on systemic projects.  All the details of HSIP projects are tracked. For spot improvement we have the lat/long of the location; but for systemic, our project cost/schedule/scope management system location description is various.  Tracking is accomplished at all phases of the project, from planning through evaluation.  This is part of the annual HSIP.  We are evaluating the benefits of installing systemic improvements by project and improvement type. Currently, naive before-after 3-year studies are being conducted. We are looking to perform additional before-after studies based on other methods as well.  Maintain database of funded systemic projects.  We keep track of all the projects we have put out to gauge their effectiveness.  Systemic sites are selected for treatment by Central Office. This is easier to maintain locations with improvements.  Majority are part of HSIP, thus required to track and report.  We can track the systemic via specific pay items.  Track the type of safety improvements.

40  Identified improvements are added to Capital Improvements Projects or Safety Improvement Project list depending on estimated cost of implementation. As these projects are funded and installed, they are removed from the respective lists.  Tracked through spreadsheets.  Some treatments are evaluated using Illinois Center for Transportation research; others are in the process of statistical evaluation (CMF development); while others are observed by the Districts to track effectiveness and adjust as needed.  Some improved locations are tracked.  We report performance of our efforts.  GIS tracking.  Letting, construction, monitoring in the future years.  We have created a statewide project to implement this type of project on local, town maintained roads. Our agency implements the improvements through a statewide contract. Right now, we are using the regular design and advertising process, but we are working on an on-call contractor approach and hope to have this in place by 2018.  We are too early on to have performed any Before/After Studies, but we anticipate doing so and re-evaluating these projects once they have been constructed for a long enough time.  Tracking is performed during the annual countywide crash data review.  Implementation has only been in place a few years; but for our target crash type, it appears we are noticing a decline in crashes on order of about 30 percent (all crashes).  The systemic efforts have not made it far enough in the post time frame to perform an analysis.  Not yet but will be. We just finished the first major systemic program. 13. What types of challenges has your agency encountered while performing systemic safety analyses or perhaps are the primary reasons your agency has not performed systemic safety analyses (check all that apply):  Lack of quality roadway data (e.g., volumes, number of lanes, lane widths, intersection attributes, etc.) 53.2% (33)  Lack of software to support quantitative systemic safety analysis 45.2% (28)  Lack of knowledge on how to evaluate the benefits of a systemic safety analysis approach for programming safety improvements 37.1% (23)  Convincing management that systemic safety analysis can complement traditional crash-data based methods for programming safety improvements 32.3% (20)  Concerns over maintenance of safety improvements after installation 33.9% (21)

41  Lack of guidance on how to conduct quantitative systemic safety analysis 30.6% (19)  Lack of knowledge to determine appropriate proportion of funding to support traditional crash-data based methods and systemic safety analysis for programming safety improvements 25.8% (16)  Lack of quality crash data (type, severity, location, etc.) 21.0% (13)  Potential legal concerns over interpretation of systemic safety analysis results 9.7% (6)  Other (please specify) 30.6% (19) Other Responses:  We have considerable work to complete before we can move from a crash-based program.  Ability to tie roadway data to crash data to understand crash potential.  Quality data on the location and installation dates of systemic treatments when implemented as part of a large districtwide or statewide project.  The roadway data comment above is specifically focused to intersection locations. We currently do not have an intersection database that includes intersection attributes.  Lack of roadside data for roadside systemic improvements.  We are currently in the process of preparing data in the format which Safety Analyst uses.  Lack of personnel trained to perform systemic safety analysis.  Lack of money for these tools and staff capacity. We only have had a transportation planner for the first time in 10 years.  Lack of time for devoting attention to systemic safety needs compared to specific issues, ongoing projects, signal operations, etc.  Lack of available staff time for analyses.  Lack of staffing resources to complete thorough assessments.  The effort required to manipulate/develop the data for input into a systemic safety analysis program is too great for our relatively small work group.  The systemic safety management approach requires more resources/staffing, so proactive Districts and local agencies are adopting systemic safety management approaches.  We have done crash-data analysis to identify crash contributing factors for head-on collisions (Highway Corridor Priority, volume, posted speed), and we are currently looking at key contributing factors for went off road crashes. The head-on crash findings have guided systemic installation of centerline rumble strips.

42  Right-of-way issues. Our FHWA Division Office requires survey to ensure all work (even things like centerline rumble strips) is done within public layout so low-cost systemic projects are extremely difficult.  Availability of matching funds.  Convincing general practitioners that systemic application contains a great safety benefit.  Preference for selecting locations based on other than safety concerns. 14. Please state any recommendations on how the methodologies or tools for systemic safety analysis could be improved (e.g., do you have any concerns related to the use of the FHWA Systemic Safety Analysis Tool, AASHTOWare Safety Analyst software, or usRAP Tools/ViDA software)? Responses:  A key to our successful use of systemic tools is our robust statewide roadway asset inventory. We recommend this be an initial focus for all states.  Analysis tools must limit data needed to Fundamental Data Elements (FDEs) in the Model Inventory Roadway Elements (MIRE).  Learning opportunities on how to use.  More training on tools and their availability.  Additional case studies and examples of peer sharing methodology, approach, and outcome.  Quantifying the "importance" of each data element to know what has the greatest effect on the analysis (i.e. how accurate does the ADT need to be before it truly affects the analysis).  Workarounds on data quality limitations (poor volume data on lower order roadways where estimates are far off from actual counts).  We experimented with usRAP and concluded a road safety audit would provide similar results.  We are looking into SPF enhancements to support systemic for RD, intersections, and bikes and pedestrians.  Tools are difficult and time consuming to link to state's data.  Safety Analyst is a very data intensive tool, and there is a concern that data collected may not be 100 percent accurate and the need to maintain/update this data could be an overwhelming task.  Safety Analyst requires data which is not available for all the roads. Safety Analyst software requires large investment in time & effort to build the database. We have hired the local University to do the data build and after 2 years, they are just about to be able to use the software.  AASHTOWare Safety Analyst can be used to improve systemic safety analysis.

43  AASHTO is too expensive for local agencies so less expensive software. I am not very familiar with usRAP. I believe the FHWA tool is great!  AASHTOWare Safety Analyst is challenging to update the data.  Have not used them at this time. Checked out using Safety Analyst but the effort to get the data in a format to use it was time prohibited.  Justifying cost of AASHTOWare can be difficult.  We have not used any of the software. It will be much easier to use one of them. We are in the process to acquire Safety Analyst though.  We use DiExSys.  While I understand the HSM is somewhat technical, it is complex for new users. We are conducting our own state training to show easier use of the HSM. We are also looking at using a tool called Numetric. We are currently waiting on our IT division to approve the use of the new tool.  HSM Methodology Part C.  State to state mandatory training for city/municipal staff.  Start looking at project planning and process software connectivity.  Locations with lower volumes or limited observed crashes are difficult to prioritize funding.  Not applicable.  None at this time. 15. Please state any type of systemic safety analysis tool that does not currently exist that your agency would like to have available: Responses:  Low or no cost implementation computer application of the FHWA toolkit. Should allow for basic to in-depth analyses. This would help agencies with engineering staff; however, many locals do not have engineering staff. In that case the respective state or FHWA needs to develop a program to get safety plans done for high priority local municipalities. Funding is helpful. Plans are great but funding for implementation is better.  Better road inventories, faster turnaround on crash reports to database.  A more direct link to crash databases and roadway design for fast outcomes in design or recommendation.  Efforts should be on helping states use the tools that exist--not creating new ones.  We plan to look into Safety Analyst and right now we are waiting to get our 2015 crash data using eCrash which is a transition from paper-based to e-form reports.  Vision 0 platform.

44  Our agency is partnering with our local university to develop crash-type-specific SPFs that will likely help us in future safety analysis.  Numetric. We also created a spreadsheet tool to help with smaller scale site-to-site evaluations. 16. What type of guidance would be beneficial to your agency in implementing systemic safety analysis (check all that apply):  Training on the use of each or selected tool 63.0% (34)  Training on selecting the most appropriate tool for specific applications 59.3% (32)  Written guidance document 57.4% (31)  Other (please specify) 20.4% (11) Other Responses:  FHWA to encourage state DOT's executive support for applying systemic applications.  Maybe not training, but the guidance on each tool would be helpful.  Case study applications; peer exchange to share practices.  Peer exchange on best practices is very good.  Trainings on how to work within the limitations.  Improved methods/standards for evaluating safety benefits in combination with capacity/delay benefits, uncertainty of results, etc.  The key issue is data and collecting it and then in the case of Safety Analyst or ViDA, being able to set up the tool.  Communicating the benefits to the public as part of planning outreach.  Our federal-funding local agency project procedures and paper-work requirements significantly hinder progress on a simple HRRRP project having only pavement marking, isolated signing, some tree removal, and utility pole relocation!  Practitioner systemic project development and benefits.