Aggregating VMT within predefined geographic zones by cellular assignment:a non GPS-based approach to mileage-based road use charging.

Abstract

Currently, most of the costs associated with operating and maintaining the roadway infrastructure are paid for by revenue collected from the motor fuel use tax. As fuel efficiency and the use of alternative fuel vehicles increases, alternatives to this funding method are being considered that don’t use fuel consumption as a surrogate for road use. Many systems have been proposed which are capable of assessing mileage based user fees (MBUF) based on the vehicle miles traveled (VMT) aggregated within predetermined geographic areas, or travel zones, in which the VMT is generated. Most of the systems capable of this use GPS. However, GPS has issues with public perception, commonly associated with unwanted monitoring or tracking and thus an invasion of privacy. One method to mitigate these issues is to use a system that utilizes a cellular network based approach that can determine a vehicle’s current travel zone, but does not determine a vehicle’s position through the use of GPS. The approach proposed here is based on a knearest neighbors (KNN) machine learning algorithm focused on the boundary of such travel zones. This method has two main phases. In phase one, the training phase, data is collected near zone boundaries using a cellular modem and a GPS receiver. This hardware creates a database that pairs readings consisting of observable cell towers and the strengths with which they were received, with the travel zone in which the reading took place, as determined by the GPS receiver. Then in phase two, the operational phase, GPS is no longer needed as the system detects changes in the vehicle’s travel zone by comparing currently available cellular information with the database. This method, while capable of determining the travel zone, is incapable of determining a vehicle’s precise location, which better preserves both the user’s actual privacy and perceived privacy. The work described here focuses on the design and evaluation of algorithms and methods that when combined, would enable such a system. The primary experiment performed evaluates the accuracy of the KNN algorithm at sample boundaries in and around the commercial business district (CBD) of Minneapolis, Minnesota. The results show that with the training data available, the algorithm can correctly detect when a vehicle crosses a boundary to within ±2 city blocks, or roughly ±200 meters. A means for handling this relatively small ambiguous region between travel zones is also presented. The findings imply that a cellular-based VMT system may indeed be a feasible method to aggregate VMT by predetermined geographic travel zones.