In a spacecraft’s Hazard Detection and Avoidance architecture, the Piloting function is tasked with selecting in real-time, during a descent onto a planetary surface, an adequate landing site which meets mission, safety and reachability requirements. These requirements are assessed by Hazard Mapping and Trajectory Planning functions, capable of producing for each candidate landing site a quantification of its quality according to multiple criteria. We developed a Piloting function, IPSIS, based on a dynamic multi-criteria decision making model, that uses a procedure based on fuzzy sets for normalizing and handling the uncertainty in this input data. For the identification of the best landing sites at each instant we use PSO-Tabu, a hybrid algorithm combining Particle Swarm Optimization and Tabu Search. This Site Selection process is aided by a Retargeting process that tracks during descent the ratings of identified high quality sites, enabling a more efficient exploration of the dynamic search space. We present here an experimental validation of this novel approach to landing site selection. Solutions of the highest quality are shown being consistently identified from evaluations of nonexhaustive subsets of visible alternatives, on an architecture seen achieving exceptional levels of performance in multiple hardware platforms.