Monitoring border-crossing vehicles has been one of the focus areas of the Department of Homeland Security since its inception. U.S. has a total of 7,514 land miles of borders with Canada and Mexico, and more than 140 million vehicles enter the U.S. border every year. The vehicles then travel using a part of 4 million miles of available U.S. public roads. Monitoring and tracking such many vehicles is a huge challenge and requires automated non-intrusive computerized technology. This paper proposes a new way of tracking vehicle routes through vehicle sensors that exist in the U.S. transportation infrastructure. In most U.S. highways and local roads, inductive loop detectors (ILDs) are embedded in the pavement to monitor traffic conditions, and the number of installations is constantly increasing. This paper introduces a method that utilizes the inductive signatures of vehicles generated by ILDs for vehicle identification and tracking. Signal processing techniques of inductance signatures and the experimental results on a highway data are presented.