I love "Design of Everyday Things" but this floor plan light switch design always irratated me.
Essentially you are sacrificing a high efficiency, low cognitive load task (that happens to have a poor first time experience) for a complex glossy UI that is intuitive but high cognitive load. (Suddenly you have to engage your visual system and import your full object recognition library where previously it was a finger twitch task.)
We've struggled with this a lot at Lockitron. My personal belief is repetive actions (locking/unlocking) should be mechanically efficient while complex actions (inviting guests) should be more visually intuitive.
Just think about how big shared spaces that would need such a light switch are typically used. Any one person will use the light switches only rarely (and frequently forget what the switches do between using them, so it’s not only about first time use). In such a situation an easily understandable light switch is preferable to a row of switches (especially since disturbing other people in that space by randomly switching on and off lights is not a pleasant experience for anyone).
Whenever a space is used by many different people such a design seems preferable to me. Also, these big shared spaces happen to be the ones that would require such a light switch in the first place.
Replacing every light switch with a floor plan switch? That would be a bad idea, sure, but there are applications that make a lot of sense.
The better argument, I think, is "Do we need that much optimization?"
Does the trial and error performed by a casual switch user cost us so much that spending the money on custom layout light panels makes sense? I think only in a very small set of applications.
To counterpoint your second paragraph: The existing wall switches are much more than a poor first time experience and require more cognition. When using the "efficient" vertical wall switch panel, you already have to import your full object recognition library without errors. You then have to call up your list of mappings between rooms and switch position on the panel so you can flip the correct one.
With the map, you only need to load your scheme of the office into active memory to select the appropriate switch. Further, the map provides error checking for your cognitive model, confirming / reinforcing the true layout (in a complex map, a user is bound to forget a room or take longer to recall it's location without the external map).
Anecdotally, I lived in my family's third house for a year and a half, and never completely memorized the functions of the 6 switch panel in our front hall.
Imagine an office, where different panels control different light banks depending on where they are in that large office space. The common user (average person in the office seldom turns on or off various things) will need to hunt for the right switch.
Yeah, but in a couple years new construction will probably just be putting LCD panels and voice control in instead of light switches anyway. So crude targeting lights on will just be done by voice and you can have a fancy map on the panel if you want since it is just an LCD anyway.
I loathe the day when this becomes true. It's an isomorphic situation to the decline in user experience that happened when automobile dashboard interfaces went from hard controls (tactile feedback, easy to grab, simple) to special-purpose touchscreens (unreliable, complex, ugly).
Cost will continue to be a factor for many years. A panel of switches may cost $25, While an integrated LCD touchscreen networked back to a control unit may cost factors more. Construction being the low-margin business it often is, I would be surprised if too many LCD panels and voice activation is common in the near future.
What I think is more likely are smarter motion/heat sensors combined with a scheduling system. During different times, certain banks of lights will turn on and off depending on the detection of people there. For conference rooms that may still need manual control lights for presentations, switches are fine because there's only one or two of them. No brainer.
I agree with you, especially because I rarely LOOK at the light-switch I'm flipping. Even at a friend's apartment, which of the switches I need to flip when entering the bathroom is a muscle memory task now.
More importantly, the floorplan lightswitch (at least from my reading of this short piece on it) doesn't have a nice way to handle multiple sets of lights interlaced with each-other on the same physical area. Experimentation with which switch on a boundary of two rooms that you need to flip is cheap.
Where experimentation takes longer is where the floorplan switch fails too. I immediately think to a gymnasium that had a huge bank of light-switches, for things like overhead fans, exhaust fans, lights that would take 15 minutes to warm up (don't want to accidentally shut those off!), secondary low lighting for when the main lighting is warming up and each of those would have up to 4 switches for different sections of the gym. The answer was text labels. Super easy, super simple. That's a pathological case, but business space has this problem, and multiple switches for single rooms is common enough in homes too.
It seems like the floorplan idea could be incorporated as a set of graphical map-based labeling on physical switches themselves. Anybody who knows the room relies on muscle memory, everyone else can use labels. And because the labels themselves aren't the active surface, they can be more specific that a touch sized button allows. Touching a kitchen on the map might be a lot of light at night, a switch underneath a picture of the kitchen that has the recessed lighting higligted wouldn't.
I don't like it either, but for a different reason. This is an example of narrow thinking: let's solve an immediate issue, rather than rethink the whole problem from the ground up.
How often do you need to control a single light that is in a different place than where you stand?
The problem is not with the light switch UI, it's with the whole concept of manually controlling every single light in the house. Why can't we have lighting that senses where we are and adapts accordingly? Why can't our lights go off when we lock the door?
Rather than get excited about a conceptually complex light switch, I'd like to see a home lighting system that doesn't require one in the first place.
To me this is one of those examples where you might re-think the whole problem from the ground up, then take a step back when realizing the absolute solution is not the perfect solution.
Take for example, the clap-it lights; while yes it requires less effort than walking over and flicking the switch, it's also extremely annoying to do. Likewise, I would get really annoyed when lights go on and off according to where I walk.
Of course I don't have a Nest either and yet it does fill a need for some people.
Has Goldee ever shown details on user input yet? I can only find concept photos with it lit us showing status, and vague mentions of gesture detection.
In fairness, I still haven't figured out what switches do what in the panel in my lab, but that's because there are multiple sets of switches in different areas of the lab that change thbe behavoir of the bank of switches I would use
> Why not arrange the switches on a floor plan of the space, so it would be easy to determine which switch worked which light?
Because our current binary model of "flip this switch, this light comes on" is dumb. We want lighting "scenarios", not control of individual bulbs/small clusters of bulbs.
What you should be solving for is these lighting scenarios, for example "I walk to walk from my bedroom to the kitchen and I want the pathway lit" (this is called pathway lighting). Or "I want ambient light in the living room appropriate for movie watching", or "I want the nightlights turned on."
Believe it or not this is already (kind of) solved with what's called Lighting Control Systems [0]. They're complicated, expensive, and I've only seen them on very, very expensive homes.
Supposedly the price is coming down, who knows. I think affordable consumer-grade is going to be an internet-of-things kind of solution.
My own approach to this from a few years ago is to have a system that tries to determine what you want to do and adjusts the lights accordingly, while allowing manual override for the times it gets it wrong:
That's awesome. Our simple system is just to have motion detector bulbs in the hallway and bathroom, which is where you usually want lights to turn on when there are people moving in those spaces.
I hate lights on motion (or sound) detectors. What if you want to stay there for while, maybe talking with a guest, or reading some paper? Or maybe you want to hook something on the wall?
I see what you mean and admit it has happened perhaps twice so far that the light went out when I needed it. But the times when walking in darkness at night or when coming home are much more often, so overall it has been a win. Not to mention saving electricity, as you never forget to turn off the lights.
Imagine you go in a hotel room, all this stuff is written in Chinese, you just woke in the night with an urgent natural need, and you just want a light on right now! So you have to read a sign with "go to bathroom"?
I think the idea of Don Norman is actually good for a big shared meeting room, were different people will have to change the lighting occasionally.
But for homes (and hotel rooms), just place your switches in a logical place and avoid too many lamps, that should be enough.
Or maybe this: have switches on main light at the doors of a room, and then each individual lamp has its switch on its cord. Wouldn't that be good enough?
Code in the various US cities I've lived in generally requires switches for a room be at a certain height and a certain distance from egress-ingress doors. I think it may have something to do with handicap accessibility, but generally there's a reason in most modern homes, you don't have to think about where the switch is.
For homes that's okay. Except overhead lights are not as nice as lamps, so turning on all the lamps you use with one switch is cool. Modern high efficiency LED bulbs can have odd lighting so having a few of these in lamps gives a much nicer effect.
And we want to encourage people to turn off lights when they're not being used, so a single switch is again useful. This might be saving tiny amounts of energy, but we want to move to a world where this kind of behaviour is built in.
<shameless self plug>
For example, my Android app, LampShade.io, has user-defined lighting 'moods' that work with the Philips Hue.
Currently these are primarily triggered via direct user interaction or through automation apps such as Tasker. I've also built a voice-controlled edition for Google Glass.
The app is still not as brainlessly convenient as a physical switch though. I don't think it can become that convenient without contextual inferring the state of the room/user, which may require more prevalent internet-of-things sensors to accomplish.
A thumbs-up for LampShade. It's far better than the horrible stock application for Hue lamps (and the recent update made for an even worse UX on iOS). I haven't even dug into the potential of Tasker integration and the like.
But you're right, as good as LampShade may be I still have to whip out the phone, fire up LampShade, swipe to the light or scene that I want, tap a button. Thing is, I don't know how to make it better, I don't think I'm wired to come up with the clever solution. I've considered NFC tags on the wall, but I don't think I'm getting that past the wife. (Plus she's an iOS girl, so NFC won't do her any good.)
Many LampShade.io power users take the NFC route. CNET TV's 'The Fix' even used LampShade.io with NFC in an episode. The issue with NFC is it only works when one's Android is unlocked...
I believe good lockscreen widgets will be more convenient that NFC, and my next big update will focus on this.
Automation, scheduled lighting, predictive lighting(not yet implemented, think Nest), etc are all very powerful, but I doubt they will ever be comprehensive enough for all daily home lighting interaction.
The only lighting system I've experienced in a very, very expensive home was kind of hilariously bad. One person turning on the whole house's lights at 3am, having to wave at a sensor every 5 minutes while reading - made me think dumb switches weren't such a bad idea.
I'm working in a recent building with no lightswitches. Since writing code doesn't involve a lot of movement, I have to wave at the sensor every so often when it goes dark. And the whole building was built to host IT companies, so it's a problem for nearly everyone.
circa 1999, GTE Internetworking (which had been BBN and would be part of Verizon later) moved to a new office building in Burlington, MA. There were several design failures there (the carpet patterns had moire effects, the main corridor was curved so that you could not see what was up ahead) and the biggest failure was the automatic motion-sensing light switches in every office.
They were designed to save money, because nobody would forget to turn off the lights. They did an excellent job of keeping people in the dark, partially because of the above sensor problem -- typing on a keyboard doesn't provide enough movement to trigger the sensors -- and partially because of their incredibly high failure rate.
At about $50 per sensor module installed, they had to keep an electrician on full-time replacing sensors. I heard that eventually they stopped relying on trouble tickets and just swept the building over about a two week cycle.
Meanwhile, people bought lights that they could control by themselves. Many of them were halogens running on 150 to 450W.
We managed to get hold of some kind of remote that's supposed to disable the sensor for two hours (though it's difficult to actually test, since there's no feedback at all when pressing it).
Since we're already using a Raspberry Pi for other purposes, I have the project of decoding the remote's protocol and driving an infrared led from one of the RPI's GPIO ports.
At Mozilla's mountain view office, the lights automatically turn off at night. To keep them on, you have to call a phone number and punch in what floor you are on. But of course this was annoying because you weren't always near a phone or aware of the number, so there is an intranet website that uses some VoIP service to phone the building controller to keep the lights on.
Still, it might be an interesting halfway point -- or a useful tool for designing such lighting scenarios -- to take into account where lights are on a floor plan. Currently, Hue lights (and WeMo switches) are the equivalent of lightswitches in a row, numbered and/or named and yet in any random order. It'd be a lot more interesting if I could control groups of switches as easily as one, without having to specify groups manually -- and the most logical way to group would be by proximity (knowing where walls are; and excluding type of lightbulb since that's often, though not always, identical). So it would make more sense to me if home automation actually cared about my home, but in the piece-by-piece approach that hasn't happened yet. I'll be writing my own software later this year to make up for it, but it'd be nice to see better, general purpose approaches, such as in the original article (with physical switches).
I've never really felt like the binary model is lacking flexibility for me. There may be some rare occasions where I want some dimming, but overall I think it works well
In all of the houses I've lived in here in the USA, its worse. Electricians literally don't care what order the switches are wired. The just grab wires and start adding switches. This results in the switch for the light you want to use often being farthest from the light on the panel. Mostly, its just random. It usually dives me nuts until I reorder them.
My very first "improvement" in my new house was "fixing" the order of light switches (before I had even moved into the house), so that they had some logical relationship to the physical layout of the space.
It's even worse when they put enough lights on a circuit that the first person to turn on a hair dryer without shutting most of them off blows the circuit breaker.
That's just incorrect wiring from my point of view; I don't know the US regulations, but you shouldn't mix lights and power sockets on the same circuit. Lights should have their own circuit breaker, of less amps than a socket circuit.
In the US, both the trade and the National Electric Code have a distinct preference to laying out branch circuits by room as much as possible. I lean towards that view as well, as it is a better logical organization for servicing and means you do not need to cut the lighting to half the house to work on a fixture.
I agree, but to my knowledge it's fully legal in most areas. I'm thankful I caught them putting my office on the same circuit and fixed it. One of my printers alone draws 13 amps while printing. I just wish I would've looked into the rest of the house when I found the office wiring.
If you had a known 13A load, why didn't you have a dedicated 15A branch circuit installed for it? Its uncommon but not prohibited in residential construction.
It's easier and cheaper for the electrician to run one power line and then distribute power via switches. Doesn't make it right, but most people put up with it.
I recently designed a laboratory and, inspired by having read Donald Norman's anecdote, tried to have the contractors match the switch layout to the position of lights on the ceiling. The result was an extremely ugly array of 2x3 separate panels that are mapped properly, but in the reverse order of what you would expect. I am now always turning on the light at the end of the lab when I want the one that is above me. I'm not sure it's any better than a linear array.
Was it more horizontally laid out or not? The original goal was to orient the switches so that, somewhat horizontally mapped to the same coordinates, you would see a clear marking of where you are and be able to turn on lights "in front" of you or "behind you". So if your finger started from the "You are here" dot, flipping the switch closest to it would turn on the light nearest you. Alternatively, perhaps you could put a symbol or distinguishing feature on the light switch you hit most often to indicate that it's not the one you want. E.g. a window, room feature, or label the door?
I've found two switches in a panel to be reasonably memorable, three in a row to be just over the border into confusion, and four or more right out.
Fortunately, all the places in my home that had more than two were places where you could control both interior and exterior circuits. My solution was to use black switches for the exterior lights and white for the interior. Like, a single panel might have a black and two whites, or a pair of each. I find it much more intuitive that way.
Just yesterday I was thinking about how to solve this exact problem, this is a great solution. However it takes a lot of effort to get switched like that made.
Plus, there are economies of scale: that same device can be used for just about every room / configuration, instead of having to create 100's of unique versions of the thing.
The hoped-for outcome for 3D printing is that it will bring the cost of a custom faceplate to maybe 5x times that of a generic faceplate instead of 100x. Of course the hardest part would be the software to turn the room layout into a 3D printed faceplate.
I question the need for custom smart light switch hardware like Goldee when low end Android tablets are significantly cheaper thanks to to the economics of scale.
I have to wonder why more than one commenter states a preference for lighting "scenarios" versus a visual representation of "what's where."
I'm one of those people who can't be bothered to fiddle with my Eclipse perspectives even though I'm sure I could summon twenty strong opinions here on why I should. Do you really think people (excluding single men under thirty left to their own devices to configure a dwelling) will compose "lighting scenarios" for their homes?
Having a map of light status and location in my entryway and on my mobile devices would be sweet. No labelling. Program it once and it's good until the walls move.
In particular, I don't think this has "high cognitive load." To me it looks like a toddler could get "touch the place and the thing happens there."
The design quickly fails when lights are not segmented out by location, but by type. For example, if you have accent lighting that goes around the top of the room, and then pathway lighting on the bottom, where does it show up on a 2D floorplan?
Are there really many locations in a conventional house layout which can afford to devote the floor space this requires? Having this stick out from the wall is a huge problem. If the house has enough space for that not to be an issue the owners probably already have an expensive lighting control system with labels by the push buttons indicating the type of lighting to use.
Look at the illustration and imagine how many times residents and guests walk into that thing. Maybe v2.0 flips down or something but that's a whole other problem.
I have a better design than this. It centers around the idea of having the light switch, itself, be near the zone of light that it operates. So instead of having a large panel of switches, we simply pair the switch with the light bulb. This means - revolutionary, I know - that the switch is itself in the same room/area as the light that it provides.
Also, the switches themselves have little lamps in them, so you can find them in the dark.
This is incredibly cool. I'm consulting for a company right now that does Power over Ethernet Lighting to enable fine grain control of lighting scenarios. I've definitely shared this link with the team. If anyone is interested in checking us out, www.igor-tech.com. My email is in my profile if you'd like to contact.
We don't have lcd touch screen switches yet, but it is certainly on the radar.
Now that we have light sources that can produce a small amount of light efficiently (LEDs) we don't really need that many lights built into the structure anyways. People can function in really really dim light. Just light up the whole building slightly all the time. You end up needing just the task lights and those are better controlled with switches at the place the light is used.
Essentially you are sacrificing a high efficiency, low cognitive load task (that happens to have a poor first time experience) for a complex glossy UI that is intuitive but high cognitive load. (Suddenly you have to engage your visual system and import your full object recognition library where previously it was a finger twitch task.)
We've struggled with this a lot at Lockitron. My personal belief is repetive actions (locking/unlocking) should be mechanically efficient while complex actions (inviting guests) should be more visually intuitive.
For another light switch concept see the Goldee.