Invited talk. Motion Autonomy and Safety issues in Autonomous Vehicles are strongly dependent upon the capabilities and performances of Embedded Perception and Decision-making systems. This talk presents how it is possible to address these important issues using Bayesian and Machine Learning approaches. The talk will be illustrated using results obtained in the scope of several R&D projects with IRT (French Technological Research Institute) Nanoelec and with industrial companies such as Toyota or Renault. The talk will be illustrated by experimental results obtained using the Inria & IRT Nanoelec Renault Zoe car.