Motion planners take uncertain information about the environment as an input.\nThe environment information is often quite noisy and has a tendency to contain\nfalse positive object detection. State-of-the-art motion planners consider all\nobjects alike, thus producing overcautious behavior. In this paper we present a\nplanning approach that considers alternative maneuvers in a combined fashion\nand plans a motion that is formed by the probabilities of those alternatives.\nThe proposed planner can smoothly react to objects with low existence\nprobability while remaining collision-free in case their existence\nsubstantiates. In this way, it tolerates the faults arising from perception and\nprediction, thus reducing their impact on operational reliability.\n