If Boston Were Smart
Imagining intelligent traffic lights, parking spaces, buildings and appliances
By Leslie Friday (Videos by Joe Chan), BU Today
Professor Janusz Konrad (ECE) and Professor Venkatesh Saligrama (ECE, SE) have developed algorithmic security cameras that spot unusual activity. (Photo by Flickr contributor Miki Yoshihito.)
Professor Michael Caramanis (ME, SE) suggests that appliances connected to a home photovoltaic unit, like a solar panel, could be programmed to detect passing clouds and choose to cycle at a later time. (Photo by Flickr contributor Savannah Corps.)
Because the cost of electricity fluctuates throughout the day, depending on demand, smart meters that are currently available tell homeowners exactly how much energy they use and at what cost, encouraging them to delay energy-intensive activities until a time of day when demand and costs are low. Supported by a $2 million National Science Foundation grant, Professor Michael Caramanis (ME, SE), Professor John Baillieul (ME, SE) and two MIT faculty members are collaborating on a study of how these and larger-scale measures could result in a smarter electricity grid. In the United States, we lose about 8 percent of energy because it travels long distances between points of generation to use. Caramanis thinks the loss could be greatly reduced if we got our energy from closer and cleaner sources. A smarter grid could help us do that.
Security officers could sort through billions of hours of video footage and spot unusual events, such as someone attempting to enter a building in the middle of the night, using specially designed cameras with embedded algorithms. Professor Janusz Konrad (ECE) and Venkatesh Saligrama (ECE, SE) have developed the technology, supported by more than $800,000 in funding from the National Science Foundation, the Department of Homeland Security, and other agencies.
BU engineers have designed software that, once uploaded to a building’s HVAC system, would measure airflow room by room and revise it to meet minimum standards, decreasing energy costs while keeping occupants happy. The invention earned Associate Professor Michael Gevelber (ME, SE), Adjunct Research Professor Donald Wroblewski (ME) and ENG and School of Management students first prize and $20,000 in this year’s MIT Clean Energy Competition. The team plans to develop and market the software through its newly formed company, .
Smarter Traffic Lights
A smart traffic lighting system would mine GPS information from cars and smartphones and count the number of vehicles waiting at red lights. If there is no approaching traffic, it would switch lights from red to green. Professor Christos Cassandras (ECE, SE) is testing this system on a model mini-city in his lab.
Cassandras, working with research assistant Yanfeng Geng (PhD, SE’13), has developed the BU Smart Parking application, which can be downloaded to a smartphone from the iPhone App Store by searching “BU smartparking.” Drivers tell the app when and where they want to park, prioritizing price and location, and the app searches for available spaces, all of which are networked to the device. When the app identifies a spot that meets the search criteria, it tells the driver where to go. At the same time, a light installed above the spot turns from green to red. When the driver who made the reservations approaches, the light turns yellow. The catch? At the moment the system works only in BU’s 730 Commonwealth Avenue garage, but Cassandras hopes to expand it to private parking facilities throughout Boston.
The next-generation lightbulb could enhance sleep quality, send data like a Wi-Fi hotspot does, or help visitors navigate large buildings through a network of visible cues, while operating more efficiently. This technology is made possible by combining LEDs, sensors, and other control systems within a single hybrid bulb that needs 40 to 70 percent less energy than existing compact fluorescent lights or LED lightbulbs. It is being developed by Professor Thomas Little (ECE, SE), associate director of the Smart Lighting Engineering Research Center, working with researchers at the center under an $18.5 million National Science Foundation grant. Little is collaborating with colleagues from Rensselaer Polytechnic Institute and the University of New Mexico.
Refrigerators and hot water heaters are duty-cycle appliances, meaning they need to run only two to three times each hour. Caramanis thinks they could be designed to communicate with the electricity grid and run when electrical demand is lowest during that time period. Alternatively, if either of these appliances is connected to a home photovoltaic unit, it could be programmed to detect when a passing cloud blocks the sun and choose to cycle at a later time. Caramanis says this technology is mostly being tested in pilot settings. A New Jersey–based company called has installed temperature sensors and communication controllers that turn on and off the hot water heaters of thousands of consumers in relation to low or high energy costs in the Pennsylvania, New Jersey, and Maryland region.
Smarter Central Control
Imagine a network of sensors that would collect and send data to a centralized processor, which could order a garbage pickup or warn drivers of traffic jams. Cassandras, Professor Yannis Paschalidis (ECE, SE), codirector of the , and Professor Assaf Kfoury (CS), are testing a miniature version of this network in Cassandras’ lab, with help from a $1 million grant from the National Science Foundation.
Back to news headlines