BOSTON, Mass., Dec. 17 -- Fans of the television show Star Trek know that finding a colleague in a large spaceship is as simple as asking the computer to locate the individual's communication device. On-body sensors are staples of science fiction, integral to dramatic situations where individuals must be located, kept from harm or simply "beamed up."
Science fiction may soon meet science fact because of research being conducted by Christos Cassandras, a professor of manufacturing engineering and electrical and computer engineering at Boston University College of Engineering and a member of the College's Center for Information and Systems Engineering (CISE). Cassandras and a team of co-investigators from Boston University and researchers from the University of Massachusetts at Amherst aim to build an understanding of how to control and optimize sensor networks operating in harsh, unpredictable environments.
The team has been awarded a five-year, $2.5 million grant from the National Science Foundation (NSF) to investigate ways to improve how large networks of sensors can both better manage the data they gather and optimize their efficiency. The NSF grant is from its Division of Design, Manufacturing, and Industrial Innovation.
Applications of the team's results will hold promise for the complex, costly, even life-threatening situations that continually confront plant managers, emergency response teams, public utilities operators and quality-control technicians.
"What we hope to accomplish is to close the loop, to use the information collected by sensor networks to do something, to make the right decisions, to take the right actions through effective management of the information being gathered," says Cassandras. "We also want to ensure that we squeeze out as much efficient use of power as we can so that the sensors will be cheap to use and will last a long time."
The project will address both the basic research issues that concern sensor network information use and issues concerning their application in "real" situations. The researchers will draw on their areas of expertise to develop software that includes simulation tools and algorithms for control and optimization implementation and to work on hardware and laboratory activities.
Sensor systems can now be found in automobiles, simple household appliances and sophisticated industrial machinery. The sensor networks that Cassandras' team will focus on, however, consist of thousands or tens of thousands of tiny, information-gathering sensors that have diverse functions and characteristics but work collaboratively to monitor a variety of physical processes or systems.
Much like the application in the fictional circumstances of Star Trek, intelligent, collaborative sensor networks could be used to locate personnel in large complexes, offering information to authorities during fires or other calamities. They could also be used to monitor condition and maintenance needs of large machines used in manufacturing or processing plants. Manufacturing set-ups could use such networks for inventory maintenance or equipment tracking, informing supply decisions and production planning, while sensors for chemical and biological substances could be used to monitor the safety of workplace environments or food products. Sensor networks could also be used to ensure uninterrupted product function, for example, alerting you to call a service technician about a part in your washing machine that is about to need repair.
In addition to Cassandras, Boston University co-investigators include John Baillieul, professor and chair of Aerospace and Mechanical Engineering; David Castanon, professor of electrical and computer engineering; and Yannis Paschalidis, associate professor of manufacturing engineering. Collaborators at the University of Massachusetts include Abhi Deshmukh and Robert Gao, professors of mechanical and industrial engineering, and Weibo Gong, professor of electrical and computer engineering.
For more information, visit: www.bu.edu/systems