In the biological world, social insects such as ants and bees use a volatile substance called pheromone for their foraging or homing tasks. This study deals with how to utilize the concept of the chemical pheromone as an artificial potential field for robotic purposes. This paper first models a pheromone-based potential field, which is constructed through the interaction between mobile robot and RFID tags. The emphasis in the modeling of the system is on the possibility of the practical implementable ideas. The stability analysis of the pheromone potential field is carried out with the aim of implementing the model on a real robotic system. The comprehensive analysis on stability provides the criteria for how the parameters are to be set for the proper potential field, and has also led to a new filter design scheme called pheromone filter. The designed filter satisfies both the stability and accuracy of the field, and facilitates a more straightforward and practical implementation for building and shaping the potential field. The effectiveness of the proposed algorithm is validated through both computer simulation and real experiment.