Freeway Safety Service Plan Designing and Optimization Modeling

Abstract

Traffic incidents on freeways cause a considerable loss of life and property. Therefore, some organizations provide freeway safety services to improve the roadway’s safety condition by assisting in detecting and clearing incidents. Incidents generally require timely assistance in removing debris and disabled cars from roads and transferring injured persons to medical care places. However, these fast reactions require more resources such as professional staff and service vehicles from the service providers. Limited funding sources constrain these operations. Thus, it is essential to design appropriate service plans for the service providers to offer in-time assistance to road users at reasonable costs. This research aims at providing appropriate freeway safety service plans based on historical traffic incident data and freeway road networks. First, clustering methods are applied to detect traffic incident hot spots. Second, with this hot spot knowledge integration, a standby service plan is formulated by configuring the whole freeway network and building a coverage model. Experiments are conducted to assess the performances of the designed plan by several proposed metrics. The data used in these experiments is from the patrol service provided by the Coordinated Highways Action Response Team (CHART) in Maryland in 2016. The proposed service plans are compared and evaluated by several metrics that are computed from evaluation experiments. The data used in the evaluation experiments are provided by the platform of PeMS in California from the incidents they assisted in 2017 due to the lack of vehicle dispatch timestamps in the CHART data. The evaluation experiment scenarios are built based on parts of road networks served by the California Highway Patrol (CHP). The experiment results indicate that the proposed model has the potential to improve upon the current operations by reducing the total response time and using the available service vehicles efficiently.