Localizing and Tracking an object, with both mobile or static sensors, is a complex but essential task when we want to use moving robots in a structured and uncertain environment. We propose in this paper a swarm approach for addressing this issue using an interaction paradigm inspired from Physics. The combination of different interactions results in a self-organised process that builds patterns interpreted as solutions of the problem. After a description of the model, this paper analyses the properties and the performance of our device through various experiments. In particular, a comparison with the standard Kalman filtering method demonstrates the relevancy of our approach.