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.
175 CHAPTER 9. CONCLUSIONS AND RECOMMENDATIONS The following conclusions and recommendations have been developed in this research: 1. The forthcoming Highway Safety Manual (HSM) can serve as a valuable tool to assist highway agencies in estimating the safety performance of specific roads. The safety effectiveness of planned projects for specific roads can be estimated by applying the HSM methodologies to the conditions before and after the project. 2. The research presented in this report has developed and demonstrated a safety prediction methodology for urban and suburban arterials suitable for incorporation in the HSM. 3. The safety prediction methodology for urban and suburban arterials uses a âbuilding blockâ approach. Separate predictions of safety performance are made for the individual roadway segments and intersections that constitute a roadway or project of interest. Those separate predictions are summed over all of the segments and intersections that make up the roadway or project. For each individual roadway segment or intersection, predictions of safety performance are made for multiple-vehicle nondriveway collisions, single-vehicle collisions, driveway-related collisions, vehicle-pedestrian collisions, and vehicle-bicycle collisions and are then combined to estimate the safety performance of the roadway segment or intersection. 4. The safety prediction methodology incorporates base models that account for the effects of traffic volume levels on safety and accident modification factors (AMFs) that account for the safety effects of on-street parking, roadside fixed objects, shoulder width, and lighting for roadway segments, and left-turn lanes, left-turn signal phasing, right-turn lanes, right turn on red, red light cameras, and lighting for intersections. 5. The safety prediction methodology has been developed with data for roadway segments in Minnesota and Michigan and intersections in Minnesota and North Carolina. The methodology includes a calibration procedure to adjust the methodology to local conditions in each geographical area to which it is applied. A calibration procedure has been recommended, but further efforts are underway in NCHRP Project 17-36 to develop a common calibration procedure for all of the HSM prediction methodologies. 6. The safety prediction methodology for vehicle-pedestrian collisions presented in this report is preliminary. A more complete pedestrian safety prediction methodology is currently under development in Phase III of NCHRP Project 17-26 and will be incorporated in the HSM methodology when development and testing is complete.
176 7. The safety prediction methodology for roadway segments has been validated using data for arterial roadway segments in Washington. The validation results indicate that the HSM methodology is most accurate for sites where the predicted accident frequency is high. The HSM methodology provides less accurate predictions for sites with low predicted accident frequencies where factors not incorporated in the methodology, such as driver behavioral factors, may have a stronger relative influence on observed accident frequencies. 8. The safety prediction model for intersections has been validated using data for arterial intersections in Florida. The results of the validation study indicate that the intersection methodology is less accurate than the roadway segment methodology. However, this result may reflect limitations of the validation dataset rather than weakness of the safety prediction methodology. Additional validation with a different dataset would be desirable. 9. The safety prediction methodology presented in this report, with the planned improvements to the pedestrian safety prediction methodology, appears suitable for incorporation in the first edition of the HSM. Further validation of the intersection methodology may be desirable. The roadway segment and intersection methodologies in the draft of HSM Chapter 10 should be reviewed for consistency with the other methodologies planned for incorporation in the first HSM edition. The HSM will be maintained and updated over time. 10. The safety prediction models presented in this report should be incorporated into the first edition of the HSM as part of the HSM production work in NCHRP Project 17-36, with further modifications as needed. 11. The HSM needs to explain the limitations of the safety prediction methodology, including the methodologyâs focus on engineering factors. It should be clearly noted that the HSM prediction methodologies have not been developed to address site-by-site variations in driver behavioral factors. 12. Future research needs to improve the HSM methodology in future editions of the HSM include refining the current base models and AMFs and incorporating models or factors for the effect on safety of traffic speed on roadway segments and intersection approaches, multiple turn lanes at intersections, channelized right turns at intersections, sight distance at intersections, roundabouts, and school zones. 13. Further development of the factors for driveway safety would be desirable including the incorporation of factors to distinguish between the safety performance of driveways on divided arterials with and without median openings. 14. Future research is needed on the effect of bicycle facilities on vehicle-bicycle collisions, similar to the research that is currently underway on the effect of pedestrian facilities on vehicle-pedestrian collisions.
