Building operators have long recognized that the users of a space tend to take a "set it and forget it" approach to their individual lighting needs. That includes simply turning on the overhead office light, and leaving it on for the balance of the day. As a result, in the commercial office environment, automated occupancy sensors have long-since replaced the simple on/off switch. While an occupancy sensor may have solved part it, the rest of the problem remains unsolved: how much light does the user really want or need?
Many individuals may make adjustments themselves by turning the overhead light on or off, but more often than not, this introduces the problem of the lights being left on when the space is unoccupied. In all cases, current lighting installations fail to address the real need, which is to deliver a user-defined amount of light only when the space is in use and maintaining that specific light level as other ambient light sources change.
Opening The Door To Energy Savings And Greater User Comfort
The solution is cognitive lighting, which yields lighting systems that can "think" and respond on their own to maintain the users' preferred and productive lighting level in response to changes in ambient daylight. While some leading-edge lighting installations have attempted to respond to, or "harvest" daylight across larger spaces, the advent of more flexibly controllable luminaires has opened the door to more granular control, which results in substantial energy savings while enhancing user comfort and productivity.
There are a number of important considerations in the design and implementation of a daylight-responsive lighting system, but they can be greatly simplified when the cognitive lighting design is first addressed at the luminaire or lamp level, rather than attempting to instigate daylight responsiveness at the building level. In its most basic sense, if you design a luminaire or replacement lamp to correctly respond to, and compensate for, the ambient daylight that it senses, the building-level task of lighting management has been simplified by orders of magnitude.
Three basic ingredients are needed to implement a cognitive lighting design:
- A dimmable light engine, i.e., a dimmable fluorescent or LED solution in which proper consideration has been given to optical design, driver circuitry, and thermal management.
- Precision ambient light sensors specifically optimized for daylight harvesting
- A microcontroller-based intelligent lighting controller (ILC) that serves as the light-processing decision engine
Creating A User Experience From User Perspectives
To implement an architecture that results in a rich user experience, we need to first define how the system behaves from the user's perspective as well as also defining how it should not behave. Current occupancy-based systems are simply automated on/off switches that pop full-on when presence is detected and turn full off when no presence is detected after a pre-set period.
Incumbent systems, especially the most common fluorescent-based implementations, often need to set long time-out periods in order to avoid on/off cycle times that can be detrimental to the lamps themselves. The common lifetime rating of a lamp actually assumes that switching does not exceed eight times per 24 hours. An on /off switching frequency beyond that daily rate can result in a dramatic reduction of luminaire life.
The best the user gets in the majority of cases is "all on" or "all off", which is fully contrary to an optimal user experience. Automated dimming would appear to provide a partial solution to both lamp life and user preference issues. In more expensive fluorescent systems that include dimming ballasts, users can dim to a level that then becomes a default pre-set, but user behavior will generally "default" to leaving it set at the highest level needed for a particular task.
What the user really wants is the light set to a particular level in the space and the luminaires to maintain that constant level as long as the space is occupied. Other than when they are actually setting the desired level, they will not want to see adjustments happening. In a sense, the "dimmer" should not be perceived as controlling the light output (power input) for the luminaires it controls, but rather as an overall ambient level controller that sets the brightness of the space.
Overview Of A Sensor-Based Closed Loop Daylight-Responsive Control Structure
With the dimming control function defined as the "target ambient level setting", the responding luminaire then has a straightforward task to accomplish: hold the illumination level of the room constant. The process of accomplishing that depends upon utilizing a high quality ambient light sensor that can measure and integrate the lux level within the space and then adjust the luminaire's output to maintain the target value.
It's worth a brief discussion on the nature of the lux value that the sensor will detect. The maximum output of a specific luminaire or replacement lamp is specified in lumens, which is a measure of the response of the human eye to photonic energy. The human eye responds differently to differing frequencies of light. For instance, x number of photons in the green region of the spectrum will seem brighter to the viewer than those in the blue region. Modern white LEDs typically produce a spectrum of light that is roughly matched to that of the human eye, with relatively more blue and less green photons, to create what we call white. While lumens are about the total amount of visible light emitted from any source, "lux" is the luminous flux hitting or passing through a specific space.
Imagine our reference source being a luminaire that we'll call a flashlight, a.k.a. a torch, emitting 100 total lumens out of its lens in a typical cone pattern. If that precisely designed lens is shining on one square meter of a big flat target, it is delivering 100 lm/m2 or 100 lux of illuminance. Move it further away along the cone shaped pattern such that the beam is now illuminating a two-square-meter area, and you'll have those same 100 lumens, but they'll be spread out across twice as big an area, giving us a measure of 50 lux. For the most part, given the same mix of frequencies/quality of light, lux are what we really care about in our lit environment. Foot-candles (fc) are another common measure that has been in use for many decades, and which follow the same basic methodology with an approximate conversion of 1 fc to 10.764 lux.
The task of the Cognitive Lighting-based luminaire is to measure the lux value that is reflected back at it from the target area for a particular set point. That task is accomplished by the ambient light sensor (ALS), which should be specifically designed to respond to daylighting scenarios. It is important to note here that the ALS is tasked with perceiving the ambient lux value in the same way that the user of the space would, which requires that it operate in a phototopic mode, meaning that it assigns a relative value to the photons reflecting back towards it and thereby accurately approximating the ambient brightness that a human would see. It is also important that the ALS is only responding to a realistic average lux value, and not to cyclic 60-Hz peaks and valleys, such as might be generated by an older fluorescent fixture in the nearby hallway.
It should be noted that since the measurements are made on reflected lux values, there could be variations introduced based upon differing reflectivity of the surfaces in a room. A shiny conference table, for instance, would reflect back substantially more light than the carpeted floor. For practical deployments, where the ceiling height is rarely as low as eight feet, and more likely 10 to 12 feet in a modern building, reflective scattering helps average things out, properly allowing the more major effect to be whether the room is dominated by table tops, or by open floor space, which themselves determine the overall ambient effect of any lighting in that space.
As an example, in a Cognitive Lighting system, each luminaire acts independently based upon the target illumination level selected by the user. As daylight enters the room, each luminaire will operate in a closed loop mode, sensing the lux value and making subtle dimming or brightening adjustments to maintain the target level. It's important to note that occupancy sensing remains a critical function, and while illustrated as a wall-mount unit in this case (presumably co-located with the illumination level control), they could just as easily be integrated into the individual luminaires.
Architecturally, system will have a straight-forward implementation that is greatly assisted by the intelligence and communications capabilities built-in to state of the art components. For example, the ambient light sensor is positioned to allow it a clear field of view of the target space, and is connected to the ILC via an industry-standard serial bus, such as I2C. The ILC serves as the decision engine, accepting data from the ALS, as well as inputs from the occupancy sensor and illumination level control, which would typically make use of a standard 0-10V control signals. The ILC drives an output 0-10V signal to any dimming ballast or LED light engine that incorporates 0-10V input dimming controls.
In terms of specific functionality, the ILC initially compares the ambient lux as detected by the ALS to the user selected target level of illumination, indicated by the dimmer setting, to determine the appropriate 0 to10V output to drive the dimming ballast or LED driver. The dimmer can be replaced by a preset potentiometer in designs where direct user adjustment is not needed, such as a warehouse space for example.
As sunlight enters the room the ALS senses the increase in ambient light in the room, reducing the 0-10V signal proportionally to maintain a constant ambient lux as determined by the user-programed target. The light engine dimming is performed on a non-linear scale optimized to provide smooth dimming, that would be a more visually pleasing experience to the human eye, enabled through programmable ramp times and adjustable lux goal targets.
In the daylight-harvesting mode the ILC should additionally minimize abrupt changes to the lighting that could result from short-term variance in the ambient lighting conditions. The use of rapid fluctuation timers used to filter the ambient environmental noise can accomplish this by allowing the ILC to re-sample the environment and if the same level change is detected in sequential sampling, only then would the 0-10V output be adjusted in the direction indicated by the change that was sensed.
When the occupancy input is tripped it sets a preset timer in the ILC Once the timer expires the ILC will shut the lights off, assuring that daylight control only operates when the space is occupied.
The ROI of energy savings alone creates a compelling value-add opportunity for properly implemented daylight-responsive lighting that also enhances the user experience. We're currently at a market inflection point where intelligent lighting is being recognized as both necessary and practical, which suggests that manufacturers need to begin to address those needs by placing designs on the drawing-board now if they want to capture their share of that market. As we've seen, the key to a fully-capable daylight responsive lighting system hinges almost entirely upon the quality of the sensors and a well-structured programming set that looks at the functionality from the user's perspective, rather than from the perspective of the controls, and the tools to accomplish it are available now.
ABOUT THE AUTHOR
Sajol Ghoshal is Director, Sensor Driven Lighting Business Unit, ams AG.