Center for Marine Robotics

In situ identification of marine environmental chemical pollution

Making measurements in situ rather than collecting samples for later analysis in a shipboard or shore-based laboratory fundamentally transforms oceanographic observation. WHOI’s long history of situ instrumentation development dates back at least 80 years to Athelstan Spilhaus’ invention of the bathythermograph for submarine detection. More recently, oceanographic sensing technologies have transitioned to autonomous operation, enabling observation and interpretation without human operators. These autonomous sensing technologies to a diversity of applications such as marine hydrocarbon detection.

Over the past decade WHOI researchers have begun fusing in-situ sensing with machine learning to create intelligent payload sensors that can be used to autonomously detect, classify, and map trace chemicals dissolved in the water column. This embedded intelligence is particularly effective when coupled with robotic vehicle guidance systems. In 2004 researchers in WHOI’s Deep Submergence Laboratory demonstrated the ability to use real-time data from a hydrocarbon sensor to dynamically re-task an AUV for hydrocarbon detection. In 2006 a WHOI scientist led spill response operations after hurricane Katrina using a newly developed mass spectrometer to locate petroleum leaks among damaged offshore oil infrastructure. Over the ensuing years these techniques have been successfully demonstrated with the U.S. Coast Guard for detecting heavy oil on the seafloor, identifying groundwater discharge and point-source pollution in coastal marine environments, tracking natural hydrocarbon seeps plumes at distances of several miles, and identifying unconventional military munitions.

In 2009, through a NASA ASTEP research grant, WHOI researchers built and demonstrated an embedded expert system that enabled a payload mass spectrometer operating on the Sentry AUV to autonomously discover and classify environmental states suitable for extreme life forms in the deep sea off the coast of California. Although this research program focused on extending autonomous robotics detecting life on other planets, the technologies developed were adapted to positively identify and track a subsea hydrocarbon plume emanating from the Macondo well in 2010.

These real-time chemical detection and mapping capabilities are limited to local spatial scales on the order of tens of kilometers. However, research is currently underway to develop a new generation of AUV-based chemical sensing systems that will enable persistent, low-cost monitoring of chemical pollution on basinto-global scales. When coupled with real-time analysis and decision-making, this technology will provide a level of situational awareness that enables immediate response to episodic or abrupt oceanographic events.

Principal Engineers

Richard Camilli
Associate Scientist, WHOI

Andone Lavery
Associate Scientist, WHOI

James Preisig
Associate Scientist, WHOI

Last updated: March 5, 2013