Membrane molecules assemble in a liquid disordered phase, playing an essential role in compartmentalization and signaling of the cell. Predicting the timescale of membrane processes, such as permeation and rupture, is crucial for determining their biological functioning, but, unfortunately, most methods at the molecular scale are limited to thermodynamical free energy calculations where the kinetical information is simply absent. The central objective of DYMEMO is therefore to go far beyond these advanced free energy methods, and to develop new physics-based computational concepts to predict the timescale of membrane processes accurately, based on path sampling simulations that do not distort the dynamics. Two conceptual routes will be followed for developing new methods: (1) tuning memory for slow events and (2) non-equilibrium dynamics. The methods will be optimized to balance between accuracy and computational load, by assessing the role of memory effects. The treatment of the external membrane distortions, which are non-equilibrium processes, is here especially challenging. Moreover, the methods will be applied to highly relevant spontaneous processes (permeation, peptide translocation) as well as to the response of membranes to external distortions (nanoparticles, photoporation). DYMEMO will lead to a much-needed breakthrough in kinetics modeling and to a deeper biophysical insight in the ‘rules’ that govern membranes.