Everything I do at UH Manoa started with a coastal monitoring project here. I built and still expand and maintain a network of sensors to monitor the water condition around the coasts of the Hawaiian islands.
Most of these sensors measure water level, but there are some that measure dissolved oxygen, temperature, salinity, turbidity, pH etc. Most of them run on solar power with battery backup. They provide real-time telemetry. The water level monitors are practically maintenance-free.
A critical design requirement for these sensors is that they have to run autonomously without servicing for years - otherwise the size of the network would be capped by the amount of manpower available, and people's time is very expensive. Some have survived storms and floods up and down the east coast of mainland US, while some are configured to operate under dense canopy along streams in Hawaii. Data links are either ISM band (with internet gateway within a few miles) or cellular (anywhere in the US), depending on the project requirements and resources available.
(No this is not how these sensors are typically deployed; this one is just sitting on a rock for the photo because I like the reflection of the tree top on the solar panel.)
Solar-powered real-time water level monitor, from design to deployment.
As is common of these rapid R&D projects, by the time they are documented and published the design is already way out of date. These are very early prototypes. I'm always on the lookout for new technologies and capabilities in embedded systems (e.g. lower power sleep mode, more powerful software support, more connectivity), so I always have a bunch of new designs using different platforms being evaluated and developed at different stages along the pipeline.