The article studies construction of a local positioning system using ultrasonic distance sensors and efficiency of some smoothing filters. The system was implemented with the usage of sensors US-015 and microcontroller Arduino Uno. The following methods were used for smoothing: a moving average method, a median filter and a Savitzky-Golay filter. The result of work is an indoor positioning system, which allows to figure out object's position on a flat surface to the nearest 2.5 centimeters. During the experiment metrics, which show average and maximum differences between the calculated and the actual trajectories, were computed for each smoothing method. Experimental results are given in the article.