Developing inexpensive crash countermeasures for Louisiana local roads.

PDFTech ReportFHWA/LA.13/50310-5SS30000240Traffic accidentsCountermeasuresSafety programsRisk assessmentBenefit cost analysisHighway safetyAnnual average daily trafficBayes' theoremMethodologyLouisianaLouisiana. Dept. of Transportation and DevelopmentSchneider, HelmutSun, XiaoduanKelle, PeterLouisiana State University (Baton Rouge, La.)United States. Federal Highway AdministrationUS Transportation CollectionAlthough 40% of all crashes in Louisiana are on local roads, local road safety improvement programs have not receivedthe attention needed to reduce crashes. Local road crash countermeasures are an important part of the overall efforts toreduce crashes and their severity in Louisiana. The efforts to develop a local road safety program are hampered by the lackof appropriate risk assessment that enables local agencies to reduce crashes using low cost countermeasures. This paperprovides a methodology that can be used by local agencies to deploy countermeasures based on a risk assessment andoptimization to meet a fixed budget. First, a statistical model is presented to assess the risk of local road segments takinginto account AADT and geometric features of the road segment or intersection. Secondly, low cost countermeasures arerecommended for individual road segments and costs for improvements are assessed. Thirdly, a score which allows theranking of road projects is developed for each road segment. This score incorporates the risk associated with the observednumber of crashes, the benefits of improvements, and the total cost of a project. Finally, guidelines for a local road safetyimprovement program are presented to allow local agencies to institute procedures for a systematic system-wide roadimprovement methodology. The deliverables include an Excel application that uses OLAP to obtain a ranking ofcandidates for road improvements. This application makes use of crash data, engineering features, and AADT to computeempirical Bayes (EB) estimates and tail probabilities for each road segment and intersection. Road segments andintersections with a tail probability below 5% are selected as candidates for countermeasures. These candidates areevaluated using Google Earth, countermeasures are suggested, and costs and benefits of the countermeasures are obtainedusing published information. The resulting road improvement projects are then ranked using multi criteria DEA includingcosts, benefits and crash risks