Guidance Augmentation Using VPS for Transit Applications
Principal Investigator:
Craig Shankwitz, Director, Intelligent Vehicles Lab, Mechanical Engineering
Project Summary:
High-accuracy, dual-frequency, carrier-phase differential Global Positioning System (DGPS) units can provide kinematic positioning estimates with accuracies of a few centimeters under favorable conditions. High-accuracy DGPS serves as the basis of the various driver-assistive systems developed by the Intelligent Vehicles Lab. Under FTA sponsorship, a DGPS-based lane-assist system was pilot tested by 16 trained Metro Transit bus drivers in mixed traffic on bus-only shoulders in 2002. Results of the pilot testing showed that the lane-assist system improved lane-keeping capability by approximately 75 percent compared to shoulder operation without assistance. However, driver stress did increase somewhat, indicating that more than a four-hour training session is needed to familiarize a driver with the system. The DGPS system described above was tested on a suburban corridor. Urban areas offer significant obstacles to implementation of such systems, including restricted views of satellites, multipath reflection of satellite signals, and cellular network holes that interfere with DGPS correction. This project addresses these issues with urban DGPS by fusing the position information provided by a vehicle positioning system (VPS) with data from scanning laser sensors. In a previous ITS Institute project, VPS was developed to provide lane-level positioning. This provides the information needed to determine the lane of travel of a bus, as well as the length of the path traveled from a known reference. Also developed (albeit to a lesser degree) was a technique that uses laser scanners to identify the presence and location of curbs with respect to a moving vehicle. The goal of this project is to demonstrate an integrated urban lane-assist system for dedicated bus lanes, integrating information from VPS, laser scanners, and a map representation describing the optimal distance from bus to curb as a function of distance traveled along the lane.
Sponsors:
- ITS Institute (RITA)
Project Details:
- Start date: 08/2006
- Project Status: Completed
- Research Area: Transportation Safety and Traffic Flow