The study is on the designing a robot assisted endonasal endoscopic surgical system; NeuRoboScope, the pituitary tumor resection surgery system. This system comprises a passive and an active arm. The passive arm positions the active arm in the surgery zone while the active arm assists the surgeon by positioning the endoscope during the surgery. The focus of this thesis is the mechanical and control safety features that can be implemented in the system. The safety enhancement methods of robot assisted minimally invasive surgery systems are investigated. Among the seventeen robot assisted endoscope holders, sixteen of them have been implemented in pituitary tumor and sinus surgeries. Safety is the main criterion that advances the progress of these systems and places them in operation rooms. Accordingly, two optimization procedures have been applied during the design of the NeuRoboScope system that have a direct effect on the suggested safety features. A novel optimization technique is proposed by employing a redundancy resolution method. The most suitable fixing point of the passive arm and its first link length is optimized to achieve the maximum manipulability with restrictions imposed by a modified condition number index and impedance of the passive arm. The active arm's partial gravity compensation is studied. Three spiral springs are used as counter-springs as the most compact and lightweight partial gravity compensation method. Particle swarm optimization method is employed for the optimization of the design parameters: spiral spring stiffnesses and preload angles. Consequently, at least 66% of actuator loads are compensated.