An Adaptive Ramp Metering System on I-880 in the San Francisco Bay Area Reduced Vehicle Hours of Delay by 16 Percent and Bottleneck Congestion by 18 Percent.
A Field-Based Evaluation Used an “On/Off” Approach to Compare Freeway Performance under Activated and Deactivated Adaptive Ramp Metering Operations.
Alameda County, California, United States
Santa Clara County
Evaluation of an Adaptive Ramp Metering System on I-880 in San Francisco Bay Area
Summary Information
Adaptive ramp metering (ARM) systems dynamically regulate vehicle entry onto freeways to manage traffic flow to reduce congestion and improve corridor efficiency. Though ARM systems have been tested through many simulation experiments, field-based evaluations remain limited. This study conducted a field-based evaluation of an ARM by integrating with multiple traffic data sources, such as detector and third-party data. The system was deployed along I-880 in the San Francisco Bay Area, a 45-mile-long corridor consisting of 136 on-ramps in both directions since 2022.
METHODOLOGY
An “on/off” evaluation approach was used to assess the performance of the ARM system. Under this method, the ARM system was temporarily deactivated to collect data representing the “off” condition and subsequently reactivated to collect data under the “on” condition. This approach enabled a more objective comparison by minimizing the influence of external factors such as seasonal variation and post-pandemic traffic recovery trends. ARM system was “on” between March 20 and April 14, 2023, then “off” between April 17 and 30, 2023, and then was turned “on” between May 1 and 26, 2023. To support the evaluation, multiple data sources were compiled, including the California Department of Transportation (Caltrans) Performance Measurement System (PeMS), the Advanced Traffic Management System (ATMS), and a third-party data provider. System performance of the “on” periods compared with the “off” period was assessed using key operational measures, including travel time, average speed, and vehicle hours of delay (VHD).
FINDINGS
- The field evaluation results indicated that the ARM system reduced travel times by four percent and increased average speeds by three percent during mid-weekdays.
- The system also reduced average delay by 16 percent and VHD by 16 percent.
- In addition, bottleneck congestion levels were reduced by 18 percent.
- Overall congestion levels were reduced by 26 percent in the March/April “on” period, and nine percent in the May “on” period.
