The Navy’s role in the Global War on Terrorism (GWOT) requires reliable long-term intelligence, reconnaissance, and surveillance. For long time frame scenarios, the main challenge faced by methods of automated surveillance and threat identification is the significant power resources required by sensor systems. Developing power efficient, low cost, long term, smart sensing technologies will be very useful to address this challenge. IAI, in collaboration with University of Michigan, is developing a novel sensing system called Cognitive Ultra-low Power Sensor System (CUPSS). CUPPS is a multimodal sensing node with wireless fencing capability supported by ultra low power (pW-nW), and a compact Phoenix hardware platform that leverages a comprehensive sleep strategy using a unique power gating approach, a CPU with compact instruction set, a custom low leakage memory cell, adaptive leakage management in the data memory, and data memory compression. In addition, CUPSS has a cognitive sensor management framework for heterogeneous, sensor modalities; sensed events; and signal propagation that allows for smart sensor positioning and power management. It features hierarchical sensor placement and fusion architecture to efficiently reduce model complexity and uncertainty and to identify friendly encounters from threat intrusions. Finally, it has a sensor middleware supporting hardware and algorithm abstraction and software wizards to rapidly integrate new sensors, and easily update the sensor sleep/trigger hierarchy and rules by a non-technical human operator. A CUPSS architecture will be built and demonstrated for a wide variety of sensing modalities, including radio and vision-based sensors, seismic and acoustic sensors suitable for advanced laboratory and supervised field-testing.