1: \begin{abstract}

2: We study a Markov process with two components: the first component evolves according to one

3: of finitely many underlying Markovian dynamics, with a choice of dynamics that changes

4: at the jump times of the second component. The second component is discrete and its jump rates 

5: may depend on the position of the whole process. 

6: %This process can model for example a spatial motion in an inhomogeneous random environment. 

7: Under regularity assumptions on the jump rates and Wasserstein contraction conditions for the underlying  dynamics, we provide a concrete criterion for the convergence to equilibrium in terms of Wasserstein distance. The proof is based on a coupling argument and a weak form of the Harris Theorem. In particular, we obtain exponential ergodicity in situations which do not verify any hypoellipticity assumption,

8: but are not uniformly contracting either. We also obtain a bound in total variation distance under a suitable regularising assumption. Some examples are given to illustrate our result, including a class of piecewise deterministic Markov processes.

9: \end{abstract}

10: