Research submarines have traditionally been the creation of governments or large institutions. The best of them are craft like Alvin, which has served the oceanographers of Woods Hole Oceanographic Institution since 1965. And navy agencies around the world have a proud record of submarine innovation. But private sub-builders have their own long tradition as well. The original bathysphere of the 1930s was built by visionary William Beebe and wealthy adventurer Otis Barton. And the Trieste, which reached the sea's deepest point in the Marianas Trench in 1960, was acquired by the U.S. Navy from its builder, Jacques Piccard.

Today, Hawkes Ocean Technologies continues that tradition with its underwater flyers: small winged submarines that maneuver in the water like airplanes. Company founder Graham Hawkes is now pushing development of his next-generation Deep Flight II, which will revisit the Marianas Trench and place the entire ocean at our disposal. Meanwhile, you can earn an underwater pilot's license on the two-person Deep Flight Aviator shown here.

Image courtesy Hawkes Ocean Technologies (created with Autodesk Inventor software).

As I've said before in this series, deep-sea exploration is a lot like the space program. And if sexy vehicles like Deep Flight are the Apollo programs of marine science, it's true again that unmanned vehicles deliver huge payoffs at modest prices in both space and the sea.

Robot submarines, without the need to support a human pilot, can be quite small. They run on batteries or cable power and may use a long fiber-optic tether to relay data to a human operator. They range from the blocky ROPOS (Remotely Operated Platform for Ocean Science), used as part of the NeMO Project (see Part 1), to the sleek, rocket-shaped REMUS units (Remote Environmental Monitoring UnitS) being deployed by Woods Hole.

These craft come in two kinds. Remotely operated vehicles or ROVs are controlled by a human at the other end of a tether. These can do a lot, but the units and their tethers require constant attention and are limited to short missions. ROVs have industrial cousins: cable-powered probes that are used to inspect large pipelines in factories and nuclear power plants.

The real cutting edge, just as with the space program, is with autonomous underwater vehicles or AUVs, which use their own computer brains to patrol a study area, take and record data, and watch for unusual events. These can stay down for months at a time and travel for tens of kilometers without a tether. Some are designed to return to a seafloor docking station like the GEOSTAR platform (described in Part 1) to upload their data. Others communicate via acoustic modem for truly hands-off operation. A master AUV page has links to 32 different institutions working in this field.

But even dumb buoys�your basic floats�are capable of landmark achievements. I'm thinking of the Argo program, which has begun populating the world ocean with 3,000 identical untethered floats. They spend most of their time drifting at a level 2 kilometers below the waves. Every 10 days they rise to the surface, taking a string of water measurements along the way, and make radio contact with a satellite. The satellite transmits the data, plus the position of the float, to a data center where it's put on the Web.

When deployment is finished, the condition of the upper ocean�currents, salinity, temperature�will be as well known as the atmosphere is today. The harvest of information from Argo will make long-range weather forecasts more reliable. And truly well-grounded climate modeling, based on Argo data, may gain us the insight into global warming that we don't have today. Keep an eye on this program.