The Soft Touch: How Might Lighting Controls Evolve?

February 1, 2016
Enlighted Inc.

Clifton Lemon is a well-known lighting and building guru with a keen eye for the future. He has been contributing a series of articles to our blog on the future of lighting and his thoughts about some of the implications of IoT. This is the third of 10 articles.

A recent article by New York lighting guru Beatrice Witzgall, The Death of the Light Switch, draws much fire in the comments section from people assuming they’re the last word in practicality. “What if I lose my phone, how will I turn off the lights?” is a common refrain when presented with new ideas in smart phone lighting control. It’s natural to assume that everything will be controlled from our smart phones of course, because so much increasingly is. Such carping entirely misses the point, which is that lighting controls must evolve, and there are many possibilities we’ve not even begun to consider yet.

As solid state lighting evolves it presents unprecedented new capabilities for electric light- luminaires can be smaller, smarter, more powerful, and more flexible than anything we ever conceived before. And the entire electrical infrastructure is also undergoing a deep transformation, driven by advancements and increasing complexity on many different levels simultaneously. In the midst of all this we are compelled to develop new ways of interacting with light and other building services so that we improve experience in buildings and save energy. Let us consider the humble light switch, which upon closer examination is not as simple and straightforward as we would like to believe. The graphic above, for instance, shows that there are already many dozens of variations on the simple device- few of them optimal, and none of them particularly well suited to the task of controlling solid state lighting.

For starters, the most common type of switch is off/on, and evolved to be so because when electrical connections are made, the less time spent in the transition phase of transferring power over the circuit the better, as the connection points are prone to corrosion. This physical fact resulted in our current understanding of electric light as something that is immediate (and actually rather jarring). Our casual language is filled with idiomatic expressions like “lights out” and “all lit up” that speak to this immediacy. This dramatic transformation in the delivery of light is an abrupt departure from everything that preceded it in human history: the gas lamp, for instance, had a key that turned so that graduation (dimming) was inherent in the experience. The earliest electrical light switches kept this design for decades until it was eventually replaced (see image below).


Today there are dozens of standard versions of off/on switches- as a mechanical device it’s hard to get much simpler, but there are illuminated, rocker, and push button versions. “On” can be indicated by “up,” “down,” “in,” or “out.” And of course in Australia, “down” means “on,” the opposite of Northern Hemisphere conventions. Maybe that’s because bathtub drain whirlpools go counterclockwise “down under.”

When asked, most humans indicate that all they want or expect to be able to do with lights is to turn them off and on and perhaps dim them. Of course if all Henry Ford had done was to ask people what they wanted, he’d never have built the Model T. Herein lies the difficulty in designing controls today – organizing choices and options that reflect the actual capabilities of new lighting and other technology. There is a massive cognitive disconnect between what manufacturers think people should be using and what they understand- absolutely no one thinks much yet about changing the color temperature of lighting, for instance, let alone how, when, or where to do it. Choice overload is a serious cognitive problem, and if lighting controls look like those things that are supposed to control TVs and other home media devices, we are doomed as a species. We already know how engineer-driven device complexity kills utility, and we can’t afford to repeat this approach.

A Design Brief for Lighting Controls

Whenever I start a design project, I want to make a design brief to clarify the basic requirements for what we’re making or building. It’s amazing how often projects get going without a clear idea of purpose, function, or other parameters. So I’d like to take a stab at an overarching general design brief for the new world of lighting controls. If a highly intelligent and effective design process ( say, perhaps similar to what drives Apple’s successful products ) were brought to bear on the problem of standardized lighting controls, here’s what I think the requirements for a device/interface might look like.

It must be understandable as a control- Its purpose must be inherent in its form. This is not an easy thing to define or narrow down. This may entail referring to earlier light switch forms purely for cognitive reasons.

It must handle multiple functions rather than a single one- Off/on options have the advantage of being, literally binary and pretty easy to figure out. Now that lighting is necessarily becoming more complex, other functions vying for attention include: graduation (dimming); color temperature; direction; memory (programmed scenes or dynamic routines); activation/override of responsive or adaptive switching. It’s completely unclear how many of these can or should be integrated into a single protocol or interface.

It must fit the body- One of the most elegant things about the iPhone to me is how it fits the hand perfectly, like a stone or a clay tablet. Small hand gestures that are natural, like swiping, flicking, or punching, all have different cognitive associations with certain actions. The haptic design of any control or input device today is crucial to its success.

It must make choices comprehensible – The balance between richness and complexity of effects or impacts and simplicity of settings is particularly challenging for the designer.

It must be interoperable with a broad range of new devices, systems, and infrastructure- This should also include HVAC, security, audio/visual, and communications.

More Questions than Answers

Making this attempt at a design brief only serves to illustrate the fact that I have the cart before the horse- a workable brief might be expected to come after careful behavioral research. The problem is that few companies and institutions that I know of are conducting the kind of research necessary to ask and begin to answer the crucial questions. Also, a large part of the design process will be to determine what problems with lighting controls are most important to solve in the first place. I’ll pose a series of some of the questions that may be said to represent “known unknowns” The right kind of investigation will seek to uncover “unknown unknowns,” of which there are certainly quite a few.

What gestures are optimal? The gestures we use on our pads, phones, and laptops are carefully thought out by UI designers. What feels better to control different aspects of light, heat, cooling, or ventilation? Swipe, tap, push, pull or slide? Gestures are not only physical, they’re representative in the graphic language used in many UI systems: icons represent mechanical buttons, sliders, or other things that trigger a memory of the old physical interface and signify its meaning. Does a swipe signify “on” better than a push or a pull? How long do we need to refer to outdated mechanical interfaces when designing new digital ones, where all we feel is the cool smooth surface of glass on a display?

Do we really need icons and graphic symbol systems? One point of view is that objects like buttons and switches should function with a minimum of explanation via labels or icons. While the international symbol system has proven to be a robust graphic language, how feasible is it to create a universal icon language to convey concepts like “warmer,” “dimmer,” “scene,” or “fade”?

Can we replace on/off with a brief gentle graduation? The abrupt on/off we’re so used to with incandescent switching was more due to the need to consider electrical and material properties than our optimal response to light. It’s a bit of a minor shock to the nervous system every time a totally dark room is suddenly blasted with bright light. Now that we can control light pretty much any way we want, why not incorporate a more nuanced graduation of light levels? Lighting designers understand the need to avoid glare and harsh contrast, but until now we have been limited to very basic controls. How will people react to a gentler delivery of light?

“Adaptive” environments or more individual control? Adaptive lighting that learns user preferences and behavior and delivers optimal light accordingly is in development. But is this the right thing to be designing? Some research indicates that occupants’ comfort levels increase significantly with only the perception of greater control of their environment- people will not necessarily use all the control options in a system but feel better simply knowing they’re there. This has prompted facilities managers to install dummy thermostats in buildings, and elevator designers to include dummy “close door” buttons. These do nothing, but make people feel better.

How quickly can people adapt to new control interfaces? This is a crucial known unknown: the assumption that people would always prefer actual buttons rather than digital ones killed Blackberry as Apple forged ahead with a more elegant solution that people simply got used to, relatively quickly. What would an equivalent story in lighting controls look like?

A New Paradigm for Research

Fortunately, intelligent lighting systems with sensors can help to create an entirely new research environment that doesn’t depend on carefully replicating real life situations- buildings with smart sensor systems can serve as constant experiments in a way, as behavioral, emotional, and experiential data from real occupants in real time will be constantly available. No more waiting around to crunch and analyze data from occupant surveys and energy bills. With the right sensor and data infrastructure in place, many versions of control devices and protocols can be tested and compared, allowing a building to evolve and improve over time as it adapts to users, rather than the other way around.

And the Future?

As lighting gets better, more efficient and higher quality, it gets more complex- all technology development follows a path to increasing complexity, just as evolution relentlessly produces life forms of increasing complexity, because technology is the key component of cultural evolution in our species. More complexity means more choices, and these must be managed cognitively in the design of devices, systems, and protocols. We’re used to extreme simplicity (off/on, dim) because for almost 100 years, lighting technology was curiously static compared to other technologies. Now it isn’t anymore, and it’s quickly becoming a vehicle for a vast, unimaginably complex data infrastructure. My sense is that when it comes to protocols and interfaces, we’ll be seeing an extended period of plurality in controls, with many competing schemes. As much as we might want broad standardization, we can’t expect global consensus on controls in a decentralizing power grid. At the same time, we will need better user interface in order to effectively use all this new technology. The way to do this is clear- pay attention to history, nature, our bodies, and our brains. This is something that’s pretty difficult for may technology companies to do.