Traffic monitoring is the subject of scientific publications and technical solutions, primarily due to the need to optimize and increase traffic safety. There are many systems that can determine direction, velocity, or make vehicle classifications. These can be, for example, traditional camera systems, laser gates, LIDAR scanning, pneumatic sensors, or induction loops. A non-traditional approach is the use of optical fibers as a traffic sensor. Optical fiber sensors use different technologies, one of them being sensors that evaluate the phase change of light (interferometers). The enormous sensitivity of these sensors combined with electromagnetic inertia are promising parameters for deploying system in areas that may be disturbed by magnetic fields or require special security requirements. The focus of this research is an optical fiber sensor based on the Michelson interferometer. The paper deals with the sensitivity of the measuring arm when changing its arrangement. Different configurations of the measuring fiber arrangement in relation to the objects to be monitored lead to the construction of a monitoring system with properties useful for direction monitoring and velocity measurement. This paper deals with the asymmetric configuration of the measuring arm of an optical fiber interferometer. Experiments in outdoor vehicular traffic conditions together with frequency spectrum analysis investigate the usability of the measurement system. The system itself can be extended to classify the measured objects by analyzing the amplitude frequency spectrum and using machine learning.