This paper proposes a new planning approach that authorizes an autonomous robot to reason about the inaccuracy of the world description and of its possible evolutions. We represent t h e uncertainty with the possibility theoryy this allows us to distinguish between two t ypes of non-determinism: a non-determinism from insuf-cient modeling and a non-determinism from uncertainty. Besides, we i n troduce perception actions as well as a model of the environment dynamics through \contingent e v ents". Finally , w e p r e s e n t an implemented experimental planner, based on Graphplan search paradigm. This planner is able to produce plans that are robust with respect to contingent e v ents, and whose goal-achieving ability i s e v aluated a pri-ori. The obtained plans can be conformant o r conditional depending on the context and the user requirements.