Autonomous operation of a small rigid hull inflatable boat (RHIB) is a complex problem that requires a robust network of sensors, controllers, processors, and actuators. Furthermore, autonomous navigation requires accurate state estimation, fusing and filtering data from an array of sensors to give the best possible estimates of attitude, position, and velocity. This thesis will address the hardware modifications and navigation state estimators used to configure the SeaFox Mk II RHIB for future autonomous operations. The study began with a RHIB capable of manual and remotecontrolled operation. The proprietary controllers and processors were replaced with an open architecture system that enabled an autonomous mode of operation and data collection from a suite of global positioning satellite receivers and inertial measurement units. Multiple navigation state estimators were designed using the extended Kalman filter and several variants of the unscented Kalman filter. Each filter was evaluated against simulated and actual sea trial data to determine its accuracy, robustness, and computational efficiency.