“…The role played by flat outputs for motion planning is thus clear: all system trajectories being parametrized by such an output, which, in addition, does not need to satisfy any differential equation, it suffices to construct an output curve by interpolation and deduce the desired state trajectory which, moreover, may be obtained without integrating the system equations, thus making this solution particularly simple and elegant. Our aim in this paper is therefore (1) to extend this property to fractional systems, a notion that will be called fractional flatness, (2) to characterize fractionally flat systems and fractionally flat outputs, (3) to provide an algorithm to compute such flat outputs, and (4) to show the usefulness of our approach on the motion planning example of a thermal system. Preliminary results on fractional flatness have been presented by some of the authors [45,44] without algebraic foundations and in the particular case where the matrix B of System (5) (see Section 3.1) is a 0-degree polynomial matrix, leading to a less precise characterization of the so-called defining matrices.…”