It took less than two decades for plastics to become notable marine contaminants following their initial commercial introductions. In the ensuing five decades, global plastics production has increased 500% each year. Currently, 300 million metric tons of plastic are produced every year much of this ends up in the ocean. In the mid-1970s it was thought that most of this plastic derived from discarded trash by ocean-going vessels. Now, we know that plastic debris in the oceans is exported by nearly all the major rivers of the world, as well as from shing, aquaculture, petrochemical, and shipping industrial sources. These plastics are long-chain polymers that are extremely durable in the environment, which is in part, why they become troublesome pollution. One practical difficulty is that in addition to the massive geographic extent of the plastic debris, the distribution is patchy in 4 dimensions (i.e., length, width, depth, time). Our approach takes advantage of the unique spectral signatures of floating plastic that can be identified in freely available Sentinel-2 imagery from the European Space Agency (ESA). Recent analytical developments identified a floating plastic index (FPI; Biermann et al. 2020) that takes a ratio of red, red edge, NIR, and SWIR bands (4, 6, 8, and 11 respectively) to discriminate between floating plastic and other floating materials such as foam, wood and vegetation. Here we present a scaled approach to identify patches of floating plastic and generate coordinates to guide cleanup efforts. The initial area of interest is 100 km on a side and is passed through hierarchical loops that 1.) separate land from water; 2.) interrogate the image files for pixels with reflectance values consistent with FPI; and, 3.) generate pixel centroid coordinates for image pixels that contain suspected floating plastic pollution. To overcome the coarse spatial (20m, bands 6 and 11) and temporal (ca. 5 days) resolution, and dynamic nearshore environment, remotely piloted aerial systems (RPAS) navigate to pixel centroids to confirm plastic accumulations. Once identified and located, updated coordinates guide navigation via autopilot in autonomous sweeper drone vessels. These sweeper drones are equipped with stereo bow-mounted hyperspectral cameras that scan the proximal waters to identify precise locations of floating plastic, plot a navigation solution, and engage the drone’s sweeper apparatus to collect the plastic. The approach integrates, distributed processing of satellite imagery in the cloud, RPAS verification of plastic locations, communication with navigation software and hardware of cleanup vessels to plot navigation solutions at the 10m scale. Finally, onboard hyperspectral cameras collect and process stereo imagery to identify precise targets which are layered on the navigation software and hardware system to direct cleanup actions. The approach is comprised of five hierarchically integrated modules that include: Coarse Location Map