KINGSTON, R.I. – January 22, 2015 – Three engineering professors at the University of Rhode Island have been awarded an $850,000 grant from the National Science Foundation to begin to develop the sensors and computer architecture for future “smart cities.” The system will incorporate a hierarchical layering of computer nodes for decision-making that is inspired by fundamental elements of the human nervous system.
According to Tao Wei, URI assistant professor of electrical engineering, the smart city concept is one in which all municipal infrastructure, including power grids, communication networks, water and wastewater systems, public transportation, health care and security, are linked by a computer architecture and sensor system for real-time monitoring and evaluation for response.
“We want to come up with the computer architecture that can manage to collect all of this data from all the individual components and use it for diagnostics applications,” said Wei, who is collaborating on the project with URI engineering professors Qing Yang and Haibo He. “If there were a fire somewhere, the system would immediately point out where it is and quickly evaluate its severity to determine how best to respond.”
The researchers call their project a reflex tree, borrowing from a concept from the human nervous system in which neuromuscular reactions and instinctive motions respond to urgent situations without requiring the direct intervention of the brain.
“When we touch something hot, we immediately pull our hand back without thinking about it,” explained Wei. “We have a controlled feedback loop in our arm, so we don’t necessarily have to use our brain in that situation. We would cause more damage if we waited for the signal to reach our brain and for our brain to tell us what to do.”
In a smart city, Wei envisions thousands of sensors or nodes throughout the city that report data to the next highest layer to respond to events.
“The nodes can talk with each other and decide whether it is a huge event that needs to be reported to the upper level nodes or whether they can respond without further assistance,” he said. “At the top of the structure is a supercomputer or cloud computer that can handle a huge amount of data. The beauty of the system is that we don’t necessarily have to bother the brain every time.”
According to a report by consultants Frost and Sullivan, the market for the development of smart cities is anticipated to reach $3.7 trillion by 2020. The first step, according to Wei, is to build a prototype, which the URI professors have already begun.
Wei has constructed a prototype of a municipal gas pipeline system in his URI laboratory with a fiber-optic network of thousands of sensors he created to collect data about the status of the pipeline. Associate Professor Haibo He is devising an artificial intelligence system for processing the data and decision making, while Professor Qing Yang is building the computer architecture, focusing primarily on the supercomputer at the top layer where the most complex tasks are performed.
“Right now we’re working on a four-layer system, but in the future it will have to have many more layers than that,” Wei said.
At the end of the four-year grant, Wei and his colleagues expect to have a working prototype that will be able to process data and respond to a variety of disturbances. Then they will apply the system to existing infrastructure, perhaps starting with just one building.
Pictured above: URI Assistant Professor Tao Wei and student Brian Chen set up a sensor system in their engineering laboratory as part of their research on a smart city network. Photo by Michael Salerno Photography