1: \begin{abstract}

2: 	Cubic spline interpolation on Euclidean space is a standard topic in numerical analysis, with countless applications in science and technology.

3: 	In several emerging fields, for example computer vision and quantum control, there is a growing need for spline interpolation on curved, non-Euclidean space.

4: 	The generalization of cubic splines to manifolds is not self-evident, with several distinct approaches.

5: 	One possibility is to mimic the acceleration minimizing property, which leads to \emph{Riemannian cubics}.

6: 	This, however, requires the solution of a coupled set of non-linear boundary value problems that cannot be integrated explicitly, even if formulae for geodesics are available.

7: 	Another possibility is to mimic De~Casteljau's algorithm, which leads to \emph{generalized Bézier curves}.

8: 	To construct $C^2$-splines from such curves is a complicated non-linear problem, until now lacking numerical methods.

9: 	Here we provide an iterative algorithm for $C^2$-splines on Riemannian symmetric spaces, and we prove convergence of linear order.

10: 	In terms of numerical tractability and computational efficiency, the new method surpasses those based on Riemannian cubics.

11: 	Each iteration is parallel, thus suitable for multi-core implementation.

12: 	We demonstrate the algorithm for three geometries of interest: the $n$-sphere, complex projective space, and the real Grassmannian.

13:   % The paper is accompanied by an open source software module, readily available for several geometries of interest in applications. 

14: 	% It is thereby readily applicable, for example, in computer vision tasks of shape recognition and in quantum control of qubits.

15: 

16:   \textbf{MSC-2010:} 41A15, 65D07, 65D05, 53C35, 53B20, 14M17

17: 

18:   \textbf{Keywords:} De~Casteljau, cubic spline, Riemannian symmetric space, Bézier curve

19: \end{abstract}

20: