In the midst of this energy crisis, how do you know if you're
wasting power? Try asking your house.
smart dust mote.
That's the idea behind the smart energy technology in development at Berkeley's Center for Information Technology Research in the Interest of Society (CITRIS). Instrumenting buildings with a network of tiny and inexpensive electronic sensors could save the state as much as $7 to $8 billion a year in energy costs while keeping consumers' utility bills in check, estimate UC Berkeley engineers and energy and environmental experts.
"One of the main goals of CITRIS is learning how to manage communal resources -- from time to people to roads," says Professor Jan Rabaey, who with colleagues from the College of Engineering and the Center for the Built Environment, wrote the CITRIS "Smart Energy Distribution and Consumption" white paper that details how information technology developed at UC Berkeley can help quash California's energy crisis.
To demonstrate the CITRIS approach to the energy crisis, researchers and graduate students installed fifty matchbox-size "Smartdust Motes" throughout Cory Hall, which houses the Department of Electrical Engineering and Computer Sciences. Outfitted with wireless radio transceivers and their own "TinyOS" operating system, the Motes cost $100 each but could easily be produced for less than $1 in the near future. Comparatively, bulky energy monitors commercially available today can cost nearly $1000 each to deploy, requiring walls to be ripped out in order to install the sensors and run conduit.
Fueled by batteries or solar power and hidden in office corners, conference rooms, and along hallways, the Smartdust Motes keep a constant vigil on light and temperature conditions. Similar devices coupled to electrical circuits in breaker boxes monitor power consumption. The readings hop from one Mote to another, ultimately landing at a central Web site for storage and data mining.
"The SensorWeb will provide huge reams of data about what's actually happening at any moment," explains EECS Associate Professor Kris Pister, a key researcher on the project. "This is important information because people have no idea where electric power is actually being burned in their homes or offices."
For instance, imagine clicking on the energy Web page for your home and seeing a list of every appliance and lamp you have plugged in and how much each device costs to operate. Watching a virtual cash register ka-ching every minute your air conditioner blows makes it much easier to settle for a slightly warmer home.
Kris Pister with his smart dust motes. Peg
"We're not going to solve the energy problem tomorrow, but these technologies could have a gigantic impact just by making people aware of what they're doing," says Rabaey, who predicts that the preliminary benefits of the smart energy technology could be seen within a year. Full realization of all its potentials could be accomplished within the next decade, he adds.
Once the buildings have a bit of brains, the next technological step is to evolve today's passive sensors into more active Motes. In recent experiments, the research team simulated Stage I, II, and III power alerts to graphically demonstrate how next-generation Motes could smartly switch off certain devices during peak power demand.
For example, during a Stage I alert simulation, a user is able to turn on two lamps in a room with no problem. However if a third light is switched on, an alarm beeps for several seconds. The user then has to turn off the third lamp or else the Mote at the circuit breaker, wirelessly notified of the alert, automatically interrupts power to all of the lamps in the room.
"Wouldn't people rather spend several hours a day with their power reduced by twenty percent instead of having a one-in-five chance that they'll be cut off entirely as part of a rolling black out?" Pister says. "Then critical equipment could continue to operate."
According to other CITRIS research, real-time pricing of electricity combined with power-aware buildings and homes may be one key to limiting peak demand. By receiving real-time pricing information, a smart refrigerator could know to fire its compressor only during off-peak periods when power prices are low.
"It's just a matter of closing the feedback loop," Rabaey says.
Smartdust Motes will be sprinkled throughout other campus buildings as well to improve the efficiency of the University's energy consumption. The researchers also plan to tackle the campus steam tunnels. According to Pister, only 40 percent of the steam generated by the boilers loops back through the system as water ready for reheating while the rest is wasted through leaks. Under the researchers' plan, the pre-existing valve sensors will be augmented with temperature and pressure-sensitive Motes to identify problem spots.
"Most importantly, we'll implement the tunnels with wireless communication to bring the data to a central location for it to be studied," Pister says. "After all, you can't fix the energy problem until you know what the problem is."
White Paper: "Smart Energy Distribution and Consumption:" www.citris.berkeley.edu/SmartEnergy/SmartEnergy.html
Smart Dust: robotics.eecs.berkeley.edu./~pister/SmartDust/
TinyOS: An operating system for Networked Sensors: tinyos.millennium.berkeley.edu
Berkeley Wireless Research Center: bwrc.eecs.berkeley.edu
Center for the Built Environment: www.cbe.berkeley.edu