Improving Safety Service Patrol Performance

About the report:

Safety Service Patrols (SSPs) provide motorists with assistance free of charge on most freeways and some key primary roads in Virginia. This research project is focused on developing a tool to help the Virginia Department of Transportation (VDOT) optimize SSP routes and schedules (hereafter called SSP-OPT). The computational tool, SSP-OPT, takes readily available data (e.g., corridor and segment lengths, turnaround points, average annual daily traffic) and outputs potential SSP configurations that meet the desired criteria and produce the best possible performance metrics for a given corridor. At a high level, the main components of the developed tool include capabilities to: a) generate alternative feasible SSP beat configurations for a corridor; b) predict incidents and SSP characteristics (e.g., incident frequency, SSP service time) for a given SSP beat configuration; c) estimate performance measures (e.g., SSP response time, number of incidents responded to); and d) identify and present the best SSP configuration(s) through visual aids that facilitate decision making.

To generate the incident data needed for the simulation-based SSP-OPT tool, a hierarchical negative binomial model and a hierarchical Weibull model are developed for incident frequencies and incident durations, respectively, based on the historical incident data. These models have been found to be effective in simulating the spatiotemporal distribution of incidents along highway corridors and for generating their attribute data (e.g., incident type, duration). The simulation program employs a discrete event-based approach and requires a few calibration parameters (e.g., SSP vehicle speed). After calibrating the model, the validation results show good agreement with field observations when applied to a sample SSP corridor from I-95. A user interface is created for the SSP-OPT tool in MS-Excel to facilitate data entry and visualization of the output metrics for a given corridor. The output includes the list of alternative feasible beat configurations and aggregated performance measures from multiple runs for each individual beat, as well as for each alternative beat configuration spanning the entire corridor. The proposed SSP optimization model could be applied to corridors with or without existing SSP service. The tool will help identify the best beat configurations to minimize SSP response times and maximize SSP response rates for a given number of SSP vehicles on a corridor. Implementing these optimal solutions in the field will result in travel time savings and improve highway safety since the SSP resources will be more efficiently utilized, thus reducing the impacts of incidents on traffic flow. 

Authors