Motivated by a recent optical-lattice experiment by Choi et al. [Science 352, 1547[Science 352, (2016], we discuss how domain-wall melting can be used to investigate many-body localization. First, by considering noninteracting fermion models, we demonstrate that experimentally accessible measures are sensitive to localization and can thus be used to detect the delocalization-localization transition, including divergences of characteristic length scales. Second, using extensive time-dependent density matrix renormalization group simulations, we study fermions with repulsive interactions on a chain and a two-leg ladder. The extracted critical disorder strengths agree well with the ones found in existing literature.Introduction. In pioneering works based on perturbation theory [1,2], it was shown that Anderson localization, i.e., perfectly insulating behavior even at finite temperatures, can persist in the presence of interactions. Subsequent theoretical studies on mostly onedimensional (1D) model systems have unveiled many fascinating properties of such a many-body localized (MBL) phase. The MBL phase is a dynamical phase of matter defined in terms of the properties of highly excited manybody eigenstates. It is characterized by an area-law entanglement scaling in all eigenstates [3][4][5], a logarithmic increase of entanglement in global quantum quenches [6][7][8], failure of the eigenstate thermalization hypothesis [9] and therefore, memory of initial conditions [10,11]. The phenomenology of MBL systems is connected to the existence of a complete set of commuting (quasi) local integrals of motion (so-called "l-bits") that are believed to exist in systems in which all many-body eigenstates are localized [8,12,13]. These l-bits can be thought of as quasiparticles with an infinite lifetime, in close analogy to a zero-temperature Fermi liquid [1,14]. Important open questions pertain to the nature of the MBL transition and the existence of an MBL phase in higher dimensions, for which there are only few results (see, e.g., [15,16]), mainly due to the fact that numerical simulations are extremely challenging in dimensions higher than one for the MBL problem.The phenomenology of the MBL phase has mostly been established for closed quantum systems. A sufficiently strong coupling of a disordered, interacting system to a bath is expected to lead to thermalization (see, e.g., [17,18]). Thus, the most promising candidate systems for the experimental investigation of MBL physics are quantum simulators such as ultracold quantum gases in optical lattices or ion traps. So far, the cleanest evidence for MBL in an experiment has been reported for an interacting Fermi gas in an optical lattice with quasiperiodicity, realizing the Aubry-André model [19,20]. Other quantum gas experiments used the same quasi-periodic lattices or laser speckles to investigate Anderson localization [21,22] and the effect of interactions [23], however,