Motivated by opportunities in smart lighting, energy efficiency, and ubiquitous sensing, we present the design of polychromatic solid-state lighting controlled using a sensor network. We developed both a spectrally tunable light source and an interactive lighting testbed to study the effects of systems that adjust in response to changing environmental lighting conditions and users' requirements of color and intensity. Using both linear and nonlinear optimization, the setpoints of overdetermined systems (greater than three wavelengths) and the energy consumption of the network are adjusted according to the room's lighting conditions (e.g., lux and color temperature of multiple fixtures). Using these techniques, it is possible to maximize luminous efficacy or the color rendering index for a given intensity and color temperature. We detail the system modeling, design, optical calibration, and control theory required to modulate the luminous output and minimize wasted energy.
Control of two five-wavelength sources using the sensor node to update the equality constraints of the linear program. The sensor node is placed at different points on the table and when the button is pressed, it measures the contribution of both light sources and the ambient light.
Control of phosphor-based fixtures using the sensor node and camera system. In this video, the intensity and color temperature can be set using either the GUI or the sensor node. The illuminance gradient is on the lower left hand corner and the energy profile on the upper right hand corner. The user can use the camera system to select a portion of the video for illumination as well as the sensor node. Closed loop control is also demonstrated. Finally, a time lapse video showing the lighting network adapt to changing ambient light is shown.
