Earth Observation Programme/Science Missions and Satellites require, efficient and cost effective, Attitude Orbit & Control Systems (AOCS). Design of AOCS is subject to complex mission requirements, such as pointing, agile re-pointing under multi-axis attitude avoidance constraints. Furthermore, a mission involves multiple modes for AOCS and each AOCS mode require special attention when designed, through appropriate control laws. Rejection of external disturbances (gravity gradient, magnetic or solar pressure torque, etc.) as well as internal perturbations (slew manoeuvers, flex, slosh, step motors, thrust misalignments, MCI offsets, sensor misalignments and thermoelastics, etc…) and acceptable levels of transients while switching among modes impose stringent robustness requirements on AOCS. Henceforth, the success of a mission relies on adequate functional verification of AOCS control design to ensure the expected behaviour in presence various possible scenarios. Furthermore, it is of paramount importance to reveal worst-case behaviour of AOCS, if any, in advance during the design stage itself. The objective of the PhD research is the enhancement of the verification of AOCS subject to demanding robustness and performance requirements, safety and time constraints. The end goal is to improve the performance and robustness of future spacecraft across all mission phases, with process allowing early detailed analysis of the AOCS design. A part of this is expected to be realised with following research objectives: (i) significantly reduce the verification effort by employing enhanced modelling and analysis concepts that use more intelligent and efficient global and Bayesian optimization and sampling techniques. (ii) Perform mapping of the high fidelity engineering simulators into representative statistical/polynomial surrogate models, which can address stochastic as well as deterministic variations and hard nonlinear effects. These models should be derived in order to allow improved analysis, from a computational efficiency, use case genericity (or built-in modularity) and fidelity view point. (iii) Assess the probability profile of performance degradation in the presence of uncertainty and failures (iv) demonstrate the added value of the newly developed processes and MATLAB codes on the realistic ESA derived missions such as Euclid and Proba 3.