Task Order 5316
Traffic Operations Research


Evaluation of the Bay Area Incident
Response System (BAIRS)

Alex Skabardonis
California PATH

Objective:

The objective of this study is to evaluate the effectiveness of the BAIRS system in reducing incident response and clearance times, and quantify the incident delay savings and other benefits that are attributable to the BAIRS system. The study will also investigate and recommend ways for possible system enhancements and integration into other traffic management and operations systems and practices used by Caltrans.

Background

In response to traffic incidents on freeways and state routes, Caltrans dispatches highway maintenance personnel who work with law enforcement and other public safety individuals to control traffic, clean-up accidents, remove debris, clean-up hazardous spills, etc. However, the process of dispatching personnel has always been cumbersome and time consuming. To that end, it can take as much as 1.5 hours to get Caltrans personnel on the scene (including time to identify and contact the correct responding party, and then time for maintenance personnel to travel to the site). In studying the dispatch process and corresponding response times, the following issues were identified as needing to be addressed:

  1. Multiple calls are frequently required to secure appropriate response personnel
  2. Contact and availability information is not frequently updated, causing delays in contacting the appropriate party to dispatch
  3. For after-hours incidents, maintenance crews responsible for the area are dispatched, even though they often live much further away than other potential responders
  4. Limited information about incidents results in an inability to determine which tools and resources are needed to resolve the incident prior to arriving on-scene; thus, additional time was needed to get the necessary tools and resources on-scene

BAIRS is an incident tracking system that assists Caltrans staff identify and respond to traffic accidents/incidents in District 4 (San Francisco Bay Area). BAIRS creates a web-based incident log that is tied to Geographic Information System (GIS) software, which locates the specific coordinates of the incident and plots the incident on a map. Next, the closest and/or most capable maintenance crew can be identified and notified for response using the on-line contact and availability database. The responding supervisor can then use his/her wireless laptop to locate the nearest equipment and materials needed to clear the incident. Since BAIRS provides real-time communication and access to information, both the Dispatcher and the responding maintenance crew can be kept up-to-date with all pertinent information. The custom-built application replaces paper radio/telephone cards, post-mile books, call-out lists, regional directories, notification guidelines, paper maps, and paper telephone books, directors and contact sheets. BAIRS is expected to reduce the average traffic incident resolution time from 4 hours to 90 minutes or less.

Proposed Research

The objective of the proposed research is to evaluate the effectiveness of the BAIRS system, quantifying improved response times and estimating the incident-induced delay savings attributable to the BAIRS system.

District 4's BAIRS has been operational and collecting incident data since June 2003. The historic incident data will be reviewed determining appropriate incident categories, such as "during business hours" and "after hours" incidents, and categorizing incidents by cause and severity, e.g., hazardous spill, multiple vehicle collision, stalled vehicle, etc. Additionally, incident location data will be reviewed to determine if incident location is a relevant variable, categorizing on-ramp, off-ramp, shoulder, and in-lane incidents. Empirical incident response and clearance time probability distributions will be created for each incident type using BAIRS incident and response data, Caltrans and CHP incident logs, and any other available data sources.

Additionally, historic data will also be collected and reviewed to determine the average response and clearance times for incidents that were managed using traditional incident management techniques – either District 4 incidents prior to BAIRS implementation, or incidents outside District 4 (but within the State).

Incident-induced delays will be estimated via deterministic queuing models created and calibrated using Caltrans count data collected by PeMS and/or provided by Caltrans in conjunction with the above described incident data. One approach is to conduct detailed analysis of a carefully selected set of specific but representative incidents, using:

  • Actual measured freeway volumes at the time and location of each incident
  • Expected traffic volumes in the absence of the incident from non-incident days
  • Expected traffic volumes for longer incident duration (no BAIRS in place)


From these, recurrent vehicular delays, incident-induced delays and delay savings attributable to BAIRS will be estimated. Then from these and the incident (type and duration) probability distributions, the expected annual delays and delay savings will be estimated.

Another approach is to construct "typical", aggregate, or average freeway and traffic profiles, which are representative of typical Bay Area freeways covered by BAIRS. Expected delays and delay savings for these prototypical freeways will be estimated for the distribution of incident types and for the distribution of incident durations.

We also investigate the advantages/limitations of BAIRS compared to other systems implemented in other Districts, ways to utilize the information from the system for system planning and management, and possible integration with other systems (e.g., PeMS).