GPS-enabled mobile phones as sensors in the determination of the components of individual trip chains withouth user input
Gómez-Torres, Nelson R.
AdvisorValdés-Díaz, Didier M.
CollegeCollege of Engineering
DepartmentDepartment of Civil Engineering
MetadataShow full item record
Dynamic Traffic Assignment (DTA) is used to maximize the network’s capacity utilization. Research for DTA is active, but the data to apply it in real world situations is still under development. The solution proposed herein is the use of GPS-enbled mobile phones (also called cell phones) as sensors (probes) to determine origins, destinations and modal components of trips without user input. The market penetration of mobile phones makes it the perfect instrument to detect locations. The Mode Identification Algorithm (MIDA) is the algorithm designed to identify origins and destinations of the following modes: pedestrians, motorized vehicles and heavy rail. MIDA does not identify trips on buses and it can only identify heavy rail trips on single route networks. MIDA is divided in two components, Identification of Heavy Rail (IDHR) under the limitation previously described and Identification of Stop, Walk and Motorized vehicles (IDSWAM). The IDHR component relied heavily on positioning, while the IDSWAM component is a fuzzy algorithm that relied on speed, direction and consistency. The algorithm was tested with data gathered by the researcher and by five (5) volunteers. MIDA was designed and tested in its identification of several modes, and it showed to be capable of identifying those modes. Testing MIDA with data from volunteers showed an error of less than 6% in the identification of modes over time period of the tests. Most of the errors come from transferences between modes, due to the fact that most of the transferences are done walking. Yet, one of the most important technological challenges for the use of mobile phones as sensors is battery life of the phones; however manufactures are constantly trying to improve battery life in their devices. In terms of public perception it appears that (at least in Puerto Rico) people are prepared to provide the location data as long as some policies are adopted. Those policies include that the purpose of the study must be clearly stated, the results of the study should be readily available, their personal data must be protected from inappropriate use, and some incentives should be adopted.