Water sensor research is inspired by nature
June 25, 2014
When an oil pipeline burst near Kalamazoo in the summer of 2010, an estimated one million gallons of crude leaked into Talmadge Creek, a tributary of the Kalamazoo River. Contaminants snaked along in sinuous patterns, eventually winding through 35 miles of waterways during one of the costliest spills in the country’s history.
More than 30 households were evacuated and scores of others were warned about the quality of drinking water in the region.
What was needed was a network of cost-efficient environmental sensors that could have been deployed quickly and networked for data collection as the pollution meandered into residential and recreational areas.
That’s the project goal for a $500,000 National Science Foundation grant that began in October in the Department of Computer Science and Engineering. The research is a collaboration of Matt Mutka, professor and chair of computer science and engineering; Li Xiao, associate professor of computer science and engineering; and Ning Xi, University Distinguished Professor of electrical and computer engineering.
“We would like to demonstrate that we can create very inexpensive mobile sensors or small robots that can be quickly deployed, energy efficient and serve a large number of environmental monitoring applications,” Mutka said. “We need to detect the concentrations and directional flow of the contaminants so we can create a map of how it is distributed in the water.”
Mutka said when inexpensive sensors are dispersed in the water, investigators need to determine where they are on the surface. “The question is how do you orient them without expensive GPS units that consume too much battery power?”
The solution may already exist in nature, Mutka explained. “We have a model of a sensor that may work. This newest research will mimic insects, the way a water spider jumps and strides across the water.”
The idea is to develop a small, inexpensive sensor that floats upon the water as it collects samples of chemical or bacterial readings. The sensor will jump, enabling readings above the water to improve communications and localization.
“The sensors will be small and easy to place in the water by being tossed from shore or a boat. Once in the water, they will communicate with each other to form a multihop sensor network, through which the data samples are efficiently routed to the cyber infrastructure.
“We already have versions that work on land so we’re looking for a sensor that not only floats but can slap the water to help orient the network. It would show us how sensors place relative to each other without GPS and then how to manage large clusters of sensors.
“It will take a while to develop, but we expect it will be spherical shaped,” he added.