Quantifying Position Accuracy of Multimodal Data from Global Positioning System-Enabled Cell Phones

Document Type

Journal Article

Publication Date


Subject Area

planning - signage/information, technology - passenger information, mode - mass transit, mode - pedestrian


Walking, Traveler information and communication systems, Transit, Telecommunications, Telecommunication systems, Site selection, Real time information, Real time data processing, Public transit, Placement (Location), Passenger information, Multimodal transportation, Multimodal systems, Mass transit, Location, Locating, Local transit, GPS, Global Positioning System, Data collection, Data acquisition, Cellular telephones, Cell phones


The emergence of cell phones with embedded Global Positioning System (GPS) chips provides opportunities to push personalized real-time travel information subject to an individual’s current location. One such application, a travel assistant device, notifies cognitively disabled public transportation users when it is time to request a stop and exit the vehicle. GPS-enabled cell phones must provide highly accurate real-time location data for this type of service. The components used in the data-collection process are identified, and a quantitative analysis of real-time GPS data obtained with a cell phone while walking, driving a vehicle, and riding public transportation is provided. The expectation was that position accuracy would suffer when the GPS signal was obstructed by a vehicle or the user’s clothing. The obtained data demonstrate the results of location fix attempts over different transportation modes in an urban environment. The highest percentage of GPS fixes (79.0%) was obtained by users walking with the cell phone open and unobstructed; walking also produced valid GPS data (i.e., location data estimated to be accurate within 30 m of the true position) 66.2% of the time. For bus trips, GPS and valid fix percentages were 71.7% and 66.1%, respectively, when the phone was held near the window; when the phone was placed in the traveler’s lap, these numbers fell to 51.3% and 27.8%, respectively. Car trips provided higher numbers: 77.7% and 71.6%, respectively. Location-based transportation applications are feasible using current technology, but predictive algorithms may be required to deliver highly accurate and timely location-aware services to cell phone users in highly obstructed environments.