Roads are one of the most important transportation routes in the world, yet the way we build roads has remained the same for decades. The road system’s structure is virtually unchanged, and there is little use beyond the primary function of load transfer. However, the road could be an essential data source for different applications. This paper presents an algorithm for detecting and tracking vehicles passing over the road surface using real-time load data. We detect individual tires based on local pressure maxima on the surface and track them using a multiple-target tracker. Our algorithm subsequently identifies individual vehicles performing pattern matching with the tracked wheels. We tested the algorithm in the Cyber-Physical Mobility Lab because there is yet to be a system for real-world testing, and cyber-physical labs are more flexible and less expensive than real-world testing. In our test run, we achieved a vehicle detection accuracy and recall of 100% and a localization accuracy of a few centimeters.