In-situ measurements are essential for monitoring, understanding, and predicting marine phenomena. Monitoring and sampling missions often require observations of phenomena with spatial and temporal dynamics that span different scales (e.g. seconds to months, meters to kilometers, etc). A combination of different vehicles, fixed nodes, advanced payload sensors, and advanced control algorithms are usually required for success. Basic research needs to focus on the individual elements composing such systems but also, more importantly, on strategies to ensure all components function together smoothly as a whole. The challenges presented by coastal environments, such as low depths and commercial vehicle traffic, increase the likelihood of collisions with oceanographic monitoring hardware and consequently the environmental geometry becomes an important constraint. In this study our group presents the progress and recent achievements of a distributed heterogenous autonomous sensor network that combines underwater, surface, and aerial robotic vehicles along with advanced sensor payloads, planing algorithms and learning principles to successfully operate across the scales and constraints found in coastal environments.@en