Traditionally, Radioactive Particle Tracking (RPT) relies on nuclear mathematical models to triangulate the position of a radioactive tracer, a process that involves significant computational resources and manual calibration. This work presents a novel methodology leveraging artificial intelligence and collaborative robotics to overcome these limitations. In this approach, a collaborative robot is employed to maneuver a radioactive tracer, generating a precise and extensive dataset that correlates physical positions with radiation levels measured by surrounding detectors. This dataset is then used to train an Artificial Neural Network (ANN) to reconstruct the particle positions. Results demonstrate that this approach provides superior accuracy in position reconstruction and flow field prediction, compared to traditional methods. The proposed technique offers significant advantages, including reduced computational complexity, faster and more accurate data acquisition, and the ability to handle complex geometries and flow conditions. These improvements make RPT method a promising tool for industrial applications.