[ Page 4 of 5 ]

In lighting, a little intelligence goes a long way

Up to half the electricity used in commercial buildings is consumed by lighting, but control schemes that match lighting use to actual demand can significantly cut that figure. A wireless lighting control system under development at the University of California at Berkeley puts sensors and switches where the action is, on the theory that letting users, building managers, and even utility companies control the lights makes for greater efficiency.

The Berkeley researchers have assembled a prototype system of programmable wireless switches, each of which can control many individual light fixtures. The scheme uses wireless sensors developed at Berkeley that together form a “mesh network” of distributed switches. The fixtures controlled by such a network can be operated manually or automatically, in response to conditions in the immediate surroundings, predetermined schedules, or triggers like signals from utility companies. The scheme doesn’t rely on a single existing control protocol, such as the Digital Addressable Lighting Interface (DALI) or Building Automation Control network (BACnet), but is intended to be compatible with existing and new lighting equipment, according to the researchers.

This diagram shows the design of a flexible solar cell containing a layer of “quantum dots” that harvest sunlight, pioneered by the University of Toronto.

In spring 2004, a test of the Berkeley system in which users were given control of the lighting in their workspaces yielded a 40 percent drop in lighting energy use. The pilot installation was a small office with eight workstations and eight fixtures controlled by a pair of switches. “Our starting point is providing local control to occupants,” said Charlie Huizenga, a Berkeley research specialist and lecturer.

The test results highlight the inefficiencies of inflexible central control schemes for lighting, especially for open-plan offices. “One person near a window kept his light off because his space was nicely daylit. Another person kept the lights off when using her computer, but turned them on when reading and doing other paper-based tasks. Another person worked half-time, and was able to turn the light on and off as he came and went,” Huizenga said. A similar but larger-scale test program, involving roughly 40 controlled lights, is slated for this summer.

Because it doesn’t require rewiring, the low-cost wireless system developed for this study makes it more feasible to retrofit existing buildings with the technology. And the mesh network also makes it easier to provide precise lighting control in new buildings, says Huizenga. Drawing on sensor research conducted at Berkeley, the control scheme taps a variety of power sources. Where relay devices are part of light fixtures, they can draw regular A/C power, but remote switches and motion sensors can run on batteries. Huizenga said, “We are looking at powering them using solar cells, or scavenged vibrational energy.” Other researchers have developed push-button switches powered by piezoelectric elements (typically crystals that produce a voltage when they’re under compression or tension, or that cause compression or expansion when a voltage is applied).

Controlling the disparate parts of such schemes—by integrating motion sensing, daylight sensing, remote switches, and central switches—is a complex and expensive undertaking, which is why so few buildings use advanced lighting-control systems. But Berkeley researchers believe electricity prices will rise in the next several years, creating an incentive to owners and operators to adopt such measures to slash costs. The cost of mesh networks like the one studied here will also likely drop as the technologies are refined. Unlike earlier systems, the devices in Berkeley’s scheme can be installed in a matter of minutes, Huizenga noted. “Maintenance is also an important issue for affordability—the controllers will need to last 15 to 20 years, at least as long as a ballast,” he said. Components like those used in the Berkeley study will be on the market within a year or so, he predicts.

The study’s results are “very encouraging, and say a lot about how much commercial space is overlit in the U.S. and perhaps elsewhere,” said Stephen Conners, director of MIT’s analysis group for regional electricity alternatives. Ideally, he added, a wireless system’s interface will allow lights to be controlled individually or in groups—but does that mean that pranksters could hijack and change your lights, à la television’s The Office? Typical security measures like password-protection systems would eliminate this concern, Conners says.

Ted Smalley Bowen

ADVERTISEMENT