Long Term Goals. To effectively carry out the mine-reconnaissance and hydrographic experiments with multiple AUVs, they must be able to self-localize accurately in an unknown or partially known environment such that relevant targets and adverse oceanographic conditions can be spatio-temporally characterized precisely. We wish to address the implementation and enhancement issues associated with applying advanced processing algorithms to a low-cost computing platform to provide a synergistic navigation solution given the size, power and cost constraints of small AUVs. To refine the next generation ultra modular mini AUV platforms for current and future sensor development and deployment on a portable commercializible platform. This will reduce costs and foster the innovative development and spread of associated sensor systems and applications. BJECTIVES. The primary goal of this project is to continue development and refinement of our next generation mini ultra-modular AUVs and provide enhanced capabilities in navigation and positioning. This particular project is aimed at improvements to low cost navigation systems using COTS equipment and extended Kalman filtering with sensor bias learning and compensation. Other related work includes the study of autonomous fault detection and compensation, as well as sea state learning for use in motion planning and oceanographic survey. One important design characteristic of these miniaturized AUVs is their enhanced modularity and field-reconfigurability, as compared to the existing Ocean Explorer series. With this added capability, we are given an additional freedom to dynamically reconfigure the baseline of these mini AUVs for optimizing hydrodynamic loading, sensor interferences, and the stinger location for docking. We wish to further quantify the performance of the mini AUVs in relation to navigation, control, hovering, and communication, and compare them with those of the Ocean Explorers.