This project continues research aimed at real-time detection and use of pedestrian traffic flow information to enhance adaptive traffic signal control in urban areas where pedestrian traffic is substantial and must be given appropriate attention and priority. Our recent work with Surtrac, a real-time adaptive signal control system for urban grid networks, has resulted in an extended intersection scheduling procedure that integrates sensed pedestrians and vehicles into aggregate multi-modal traffic flows and allocates green time on this integrated basis. The effectiveness of this procedure has been demonstrated in simulation under the assumption that requisite sensing is in place. In this project, we further this work in two ways. First, we consider the companion issue of adequately sensing pedestrian and vehicle traffic in an integrated manner. Through collaboration with Citilog, Inc., which provides a unique, software-based approach to pedestrian detection, we will investigate development of higher-level techniques for extracting pedestrian density and direction information in addition to presence data. We will similarly evaluate the potential for basic pedestrian detection via radar, using the Wavetronix equipment currently installed in portions of the Surtrac test bed network (in the Pittsburgh East End). The second thrust of the project will explore use of additional information (e.g., movement disabilities) that may be obtained through direct pedestrian-to-infrastructure communication in making adaptive signal control decisions that promote safe and efficient pedestrian movement. We will evaluate our technology results through pilot testing and where successful attempt to take steps to deploy on a continuing basis within the Surtrac test bed.
-
-
-
-
Name | Affiliation | Role | Position | |
---|---|---|---|---|
sfs@cs.cmu.edu | Smith, Stephen | Robotics Institute | PI | Faculty - Tenured |
Type | Name | Uploaded |
---|---|---|
Final Report | Smith_TSETFinalReport.pdf | May 7, 2018, 4:27 a.m. |
No match sources!
No partners!