Abstract
The paper presents a new model for estimating individual vehicle speeds from conventional round single inductive loop sensors equipped with advanced high-speed scanning detector cards. The model presented improves upon existing models with final results of individual vehicle estimated speeds showing average errors of less than 2 mph and 4 mph, respectively, for two independent test datasets. Improvements to individual vehicle speed estimations can help to enhance analysis in traffic operations and planning fields which depend on accurate speed values such as traffic safety, where knowledge of speed variance is critical. The ability to estimate speed from a single loop, as opposed to double loop speed traps, reduces the need to install double loops and can assist in collecting data when information from double loops is not available.