Floating trash in rivers is a major source of pollution for larger bodies of water like oceans and lakes. Rain can wash litter into rivers and canals, where it makes its way down into sensitive ecosystems, hurting both animal life and making beaches and shore-lines unsightly. Several American cities have set goals to reduce or eliminate the waste found in their natural bodies of water. Achieving these goals, however, has proved elusive. Leveraging machine learning, this project will explore a cost-effective, manufacturable, and maintainable robotic solution, powered by the flow of the river, for monitoring and removing floating trash from waterways. The objective of this Phase I project is to conduct a feasibility study of a unique robotic system for monitoring and aggregating floating debris from rivers and canals. Specific objectives of the Phase I feasibility study include: Deeply understanding the needs of potential end-users, building a proof-of-concept prototype and testing it on the Jordan River (UT) and, developing a high-level system design for use in further prospective system development. The unique features of the proposed system allow it collect and aggregate waste from a wide range of waterway sizes, even when those rivers or canals are frequented by heavy boat traffic. These features allow the system to deployed in waterways that run alongside heavy urban developments, including locales that are infamous for waterway pollution.