Small Unit Precision Combat (SUPC) aims to extend current US military dominance to complex terrains, such as urban or mountainous areas. Currently, when forces engage in close-in combat in such terrains, US capabilities for precision strike and surveillance/reconnaissance are negated. SUPC contemplates creating a similar advantage in complex terrain through improved situation understanding generated by a combination of air and ground sensors and weapons available to highly trained small units in a manner that is available to these units as if they were controlled by them. However, the complexity of the surveillance/ reconnaissance and precision strike tasks in complex terrain means that these assets must embody a significant degree of autonomy. Such autonomy will be necessary to control the contemplated decentralized, multi-layered architecture that seamlessly coordinates national, operational and tactical assets. This talk will outline a concept of ground, low-air and high-air assets and the basic tactical and operational requirements of the system. It will point to specific challenges where persistent agent capabilities would be needed to support the SUPC concept: (1) loitering swarms of unmanned aerial vehicles that autonomously detect threats; (2) the control of remote nonlethal weapons to protect forces; and (3) intelligent indirect personal weapon systems that identify obstructed targets and select the ideal strike method from remote weapons. For the entire system concept the multi-layered management and control system requires a high level of autonomous allocation of assets across distributed operators and the ability to manage sustainment and replenishment.