NEPTUNE Canada

News Details

New Funding Announced

Oct 29, 2009

CANARIE has awarded $980,000 in new funding to support some exciting new approaches to data collection and analysis. Data collection will be enhanced by installation of a "mini-observatory" at a nearby secondary school. Data analysis will be augmented by two new software-enabled approaches.

Extending the Observatory

Part of the CANARIE funding will help Vancouver Island's Brentwood College School set up and maintain a small near-shore sensor platform, which will feed realtime measurements into the combined NEPTUNE Canada/VENUS online database. Brentwood College School is located on the western side of Saanich Inlet, roughly opposite the VENUS Saanich Inlet location. Data collected from this site will be of great interest to researchers wishing to compare conditions across the inlet.

“We’re a school on the waterfront and we wanted to be more active in monitoring the environment that’s in our own backyard,” says David McCarthy, Brentwood’s director of studies. “We’ll also be doing our own projects in the estuary here in Mill Bay. We’re very excited about this collaboration with UVic.”

Schematic
Overhead view of the Brentwood College School campus in Mill Bay, BC. (Click to enlarge.)

Data Analysis by Software Agents

Our underwater video cameras and hydrophones promise to produce vast amounts of data, more than can be easily analyzed by small teams of scientists. To effectively sort through our mountains of rich-media data, we'll need help. Some help may be provided by new software that will attempt to automatically detect interesting features in our data stream. Part of the CANARIE award will go toward development and improvement of new detection software.

Crowdsourcing

Software agents will help us identify some things, but not everything. Sometimes, the human brain can analyze date much more quickly and easily than software. To this end, we hope to recruit the help of students and the general public. The new CANARIE funding will support development of software for analysis of video and acoustic data via crowdsourcing, an approach that allows non-experts to help interpret and filter data through the web.

Non-experts may not be able to identify specific species of fish, for example, the shortspine thornyhead, but they can certainly identify an orange fish in the video, marking it for further inspection by an expert. Similarly, a non-expert may not be able to tell the difference between songs of the Northern Resident Orcas and the Southern Resident Orcas, but they could mark unusual sound sequences that may be of interest to marine mammal specialists. In this way, the intelligence and energy of many individuals can be harnessed to advance scientific research. Analyses by humans will also be used to tune and verify detections made by software agents.

Schematic
Schematic overview of project framework for 'Data from the Deep, Judgements from the Crowds' project. (Click to enlarge.)

Related News Releases


Return to News Home