This document presents the methods and algorithms for the development of robust and resilient timetables as developed in WP3. Chapter 1 presents the general ON-TIME WP3 timetabling framework consisting of three integrated modules: a microscopic model, a macroscopic model, and a finetuning model, as well as the I/O data transformations from standardized RailML input files. Moreover, the performance measures and objectives are defined that are taken into account in the timetabling approach explicitly: infrastructure occupation, stability, feasibility, robustness, resilience, transport volume, journey time, connectivity and energy consumption. Chapter 2 presents the ON-TIME WP3 algorithmic framework and defines the three hierarchical levels of network modelling: the microscopic (detection-section level), mesoscopic (block level) and macroscopic (open-track level) networks, which are used by different modules and can be converted into each other enabling a consistent data structure at different levels of detail. Chapter 3 presents the microscopic model which consists of a suit of microscopic and mesoscopic (block level) algorithms that compute basic timetable elements such as running times, blocking times, and minimum headway times, as well as infrastructure occupation, stability, conflict detection, and bandwidths of train path envelopes. The results of these computations are used in the macroscopic and fine-tuning models. The microscopic model also takes care of the aggregation of processes for the macroscopic data and acts as the kernel for the other two models in an iterative process. The microscopic model feeds aggregated data to the macroscopic model that computes a feasible timetable at macroscopic level which is then transformed back into a microscopic timetable which is checked on conflicts, capacity consumption, and stability. If one of these performance measures is not yet satisfied, new aggregated data is computed and the macroscopic model is called again. Otherwise, energy-efficient speed profiles are computed and bandwidths are determined for optimization by the fine-tuning model. Chapter 4 considers the macroscopic model which consists of an optimization model and a stochastic delay propagation model with re-optimization for robustness evaluation. The optimization model finds a macroscopic timetable that minimizes a cost function consisting of running, dwell, and transfer times, and costs for cancelling trains, cancelled connections, and periodicity (where relevant). A heuristic algorithm generates multiple timetables that are analysed on robustness using a stochastic model which selects the most robust timetable as final output. Chapter 5 considers the fine-tuning model which consists of a dynamic programming model that finds the optimal allocation of running time supplements and dwell times within a corridor for local trains taking into account stochastic dwell times at the intermediate stops and energy-efficient speed profiles between the stops. The objective is minimizing the expected energy consumption and expected delays at the intermediate and target stations. Finally, Chapter 6 gives conclusions and classifies the developed timetabling approach at Timetabling Design Level 3 and 4. The approach incorporates stability, feasibility and robustness explicitly in the computation of a high-quality timetable using an integrated approach of deterministic, stochastic, macroscopic and microscopic models. It is based on standardized RailML files, where the RailML Timetable scheme has been extended with microscopic time-distance and energy-efficient speed profile information for punctual running. Timetabling Design Level 4 is obtained regarding the timetable resilience to ad-hoc freight path requests.