The integration of robotic technologies in surgical instrumentation has contributed to the further development of Minimally Invasive Surgery (MIS) aimed at reducing the patient trauma and hospitalisation costs. Recent advances in imaging and mechatronics promise to enable the execution of more complex interventions through a single incision or natural orifice access. The main requirement for such procedures is the ability to reach the operative target through complex routes and curved anatomical pathways whilst maintaining adequate stability for tissue manipulation. A mechatronically controlled device with a high degree of articulation can potentially fulfil these requirements. However, the incorporation of a high number of Degrees-of-Freedom (DoFs) increases the control complexity of the system. The purpose of this thesis is to investigate different methods for reducing the control dimensionality of a hyper-redundant snake-like articulated robot for MIS. The design of the robot is based on a modular, flexible access platform featuring serially connected rigid links and a hybrid micromotor-tendon actuation strategy to construct independently addressable universal joints. A path length compensation scheme for reducing the backlash at the joint is presented, together with experimental evaluation of the joint positioning accuracy when using our proposed kinematic control. The integration of an extra translational DoF along the joint axis allows the performance of a hybrid 'inchworm-snake' locomotive scheme for self-propulsion of the device. This 'front-drive back-following' approach is implemented by actuating only one module at a time in a serial fashion. Therefore, the operator only needs to steer the distal tip of the device while the body of the robot follows the desired trajectory autonomously. Once the distal tip of the hyper-redundant device has reached the target operative site, the body of the robot has to adapt its shape to the surrounding moving organs while keeping the end-effector stable. A DoF minimisation algorithm is designed to identify the minimum number of joints to be simultaneously actuated to ensure shape conformance whilst simplifying the control complexity of the system. Finally, optimal kinematic configurations of the platform are derived for performing two specific single incision procedures. The results, based on workspace requirements estimated through pre-operative imaging of the patients, demonstrate the suitability of the system for such procedures. Results from in vivo experiments on porcine models are also provided to show the potential clinical value of the device.