Two years in the making, AQUA is a mechanical creature that can walk off a beach, swim below the surface and find its own way around obstacles employing technology developed at York’s Centre for Vision Research. The vision sensors are key to AQUA’s ability to manoeuvre through and study underwater environments creating three-dimensional models of things it sees such as coral reefs, ships’ hulls and marine topology.

The three-year joint project includes researchers Gregory Dudek and Martin Buehler from McGill University and Evangelos Milios from Dalhousie University and is supported by the federally funded Network of Centres of Excellence Institute for Robotics and Intelligent Systems (IRIS). The project also enjoys the support of the Canadian Space Agency and Brampton-based MD Robotics.

"The primary reason for AQUA is to build intelligent systems," says Dr. Jenkin, a professor in York’s Department of Computer Science. "The underwater environment is as complex as you can get, more so than outer space, so it allows us to develop manoeuvring capabilities that space researchers are interested in."

With its trinocular vision system, AQUA can capture video data of the world around it, recognize familiar places and remember the shape of objects it is looking at. "While these visual tasks seem simple to humans," said Dr. Jenkin, "they represent the cutting edge for robot systems and are, in fact, beyond the reach of other robots such as the recent crop of mars exploration vehicles." The robot also has locator abilities that will relay position information to mark the site of what it finds or for mapping reference data.

AQUA has many applications of interest to earth-based researchers. The autonomous robot can chart changes in a coral reef over a period of time, inspect ships’ hulls and oil rigs, or map the ocean floor.

The project has already attracted interest from several potential users and testing will continue in June at a freshwater lake in Ontario and again next January at McGill’s Bellairs Marine Research Institute in St. James, Barbados.

The York team includes researchers Jim Zacher, of the Centre for Vision Research, and graduate students Arlene Ripsman and Andrew Hogue from the Department of Computer Science.