An algorithm that can accurately gauge heart rate by measuring tiny head movements in video data could ultimately help diagnose cardiac disease.
CAMBRIDGE, Mass.-An expedition this month in waters off the coast of
Washington State is using robotic submarines and buoys that talk to one
another to gather information that could ultimately aid the cleanup of
oil spills in this and similar areas. The scientists involved hope to
learn more about the tidal mixing of river and ocean waters.
The experiment, which will run from June 10 through July 5,
involves the transmission of brief pulses of sound. As a result, the
scientists have worked closely with marine biologists to ensure that
there is no potential for injury to marine mammals.They applied for and
received a permit for the experiment from the National Marine Fisheries
The researchers are from MIT, the Woods Hole Oceanographic
Institution (WHOI), the Institute of Ocean Sciences (IOS), and the
University of Victoria (UVIC), British Columbia. Professor Henrik
Schmidt, associate head of the Department of Ocean Engineering (OE),
leads the MIT and WHOI program; David Farmer leads the IOS team in
Surface slicks of foam, seaweed or other detritus in coastal waters
often signify a tidal front, in which two different water masses-for
example, river and ocean water-meet. Characterized by vigorous mixing,
the front stirs nutrient-rich water up into the surface layer,
attracting a variety of plants and animals. Fronts are therefore
important to the food chain.
But while scientists understand that the mixing along a front
supports an abundance of life, there are uncertainties about the way in
which the water mixes in the first place. They hope to solve this puzzle
by studying a tidal front at Haro Strait, a narrow channel between
Washington State and Vancouver Island, BC. To do so they are bringing
together for the first time autonomous underwater vehicles (AUVs), buoys
that talk to one another, and instruments designed to acoustically image
the small-scale structure of the front.
Understanding the mixing at the front may help protect Haro Strait
and similar environments. The experimental site itself is a very busy
waterway adjoining the straits of Juan de Fuca and Rosario, along which
numerous tankers carry Alaskan oil. If a tanker ruptures anywhere in
this area, emergency crews will have to act wisely to save marine life.
They can only do so if they understand how waters in the strait move and
mix the spill.
The five buoys, which are tethered to the ocean floor, are
clustered around the rough location of the front. Acoustic
instrumentation along the length of each tether sends and receives short
pulses of sound. This sound can be used to detect the front and to
communicate with the AUVs (which also send and receive sound).
Sound travels at different speeds through waters of different
temperature and salinity. So by tracking the speed of sound from one
tether to another, the scientists hope to map extreme changes in
salinity and temperature. This would indicate the rough location of the
This technique can't, however, give details about the front. Enter
the AUVs. Once the front is found (its position changes based on the
tides), its location will be communicated to the AUVs. The robots will
then move in to take detailed measurements of salinity and temperature
and otherwise characterize the area.
Imaging sonars are fitted both to the AUVs and a freely drifting
buoy. "The sonars will acquire images of turbulence effects on the
surface and beneath, providing valuable insight on the way in which
mixing takes place," Dr. Farmer said. A ship will make detailed maps of
the current and water properties throughout the strait; additional
moorings will provide measurements of tidal current and water properties
over a larger area.
Data from the AUVs and some of the moorings will be broadcast to
computers in the "nerve center" on shore, where researchers will review
it then send back instructions for where the subs should go next. "This
real-time access to the data allows us to adapt the experiment as we go
along," said Professor Schmidt.
Some of the data will be forwarded via the Internet to MIT and WHOI
for analysis by other researchers. "This is necessary for analyses that
require more computational power than is available in the nerve center,"
Professor Schmidt said. If successful, such use of the Internet could
allow future researchers to perform experiments from the office.
PROTECTING MARINE MAMMALS
The researchers built into the experiment a number of safeguards to
protect marine mammals. For example, mammals in the area will be
monitored closely before, during, and immediately after the experiment.
Those common to Haro Strait include harbor porpoises, Dall's porpoises,
killer whales, and harbor seals.
If the monitoring shows that the experiment is having a significant
impact on the animals-say, a radical change in travel behavior-an
independent oversight committee composed of marine mammal experts will
be alerted. The four members are from Marineworld and the University of
Washington; UVIC; the Department of Fisheries and Oceans, BC, and the
Whale Museum at Friday Harbor, Wash. This committee "has the authority
to shut us down," said Professor Schmidt.
The researchers have also adapted the acoustic system so as not to
startle the animals. "Rather than coming on suddenly, the sound will
start softly then ramp up," Professor Schmidt explained. "So even if an
animal happens by accident to be right next to the source, it won't be
The sounds from the experiment are similar to those made by the
underwater electronic sonar devices used by fishermen to locate schools
of fish. Professor Schmidt noted that "whale-watching boats are much
noisier than our experiment." The short pulses of sound-the longest will
last two seconds and be broadcast once a minute-will be transmitted for
about four hours a day during the three-week period of the experiment.
As standard procedure, the researchers applied to the National
Marine Fisheries Service for a permit for the experiment. The permit,
which was received June 7, details all monitoring and mitigation
Patrick J. Miller, a graduate student in the Department of Biology
and WHOI, is leading the monitoring effort. Mr. Miller is not one of the
scientists involved in the formal experiment. An expert on marine
mammals and sound, he heard about the experiment and recognized that the
instrumentation involved would also likely yield a wealth of data on the
habits of the animals themselves. As Professor Schmidt notes, "60-70
percent of the time the sound sources won't be on. We'll just be
listening." Mr. Miller and his advisor, Peter Tyack at WHOI, represent
the marine-mammal scientists. Mr. Miller wrote the application for the
MIT scientists involved in the experiment, in addition to Professor
Schmidt, are: Professor Arthur B. Baggeroer and graduate students Pierre
Elisseeff, John H. Kim, Jeffrey S. Willcox, and Yanwu Zhang, all of OE;
Principal Research Engineer James G. Bellingham, Research Engineer John
J. Leonard, Research Specialist Robert J. Grieve, and Postdoctoral
Fellow Bradley A. Moran, all of the MIT Sea Grant College Program, and
Max Deffenbaugh, a graduate student in the Department of Electrical
Engineering and Computer Science.
Scientists from IOS and UVIC, in addition to Dr. Farmer, include
Rich Pawlowicz and Ross Chapman. Scientists from WHOI are Principal
Investigators David Herold and Mark Johnson, and Matthew Grund.
The Haro Strait experiment is supported by the Office of Naval
Research. Mr. Miller is also supported by ONR, but through a different
grant. The development of the AUV technology has been funded by ONR, the
MIT Sea Grant College Program, the National Ocean and Atmospheric
Administration, and the National Science Foundation.