I was not even aware that there was such a thing as this lighting fixture. I wish I was still ignorant of its existence.
I can help.
[Puts on MiB sunglasses]
Look into the flashy thing. On 8 seconds, off 2 seconds, on eight seconds, off 2 seconds, on 8 seconds…
You missed a cycle, which did a total brain wipe.
Oops.
I watched this video and laughed my ass off, and shared it with my colleagues. It is unintentionally hilarious. However, there are some interesting lessons to take from it. Yes, obviously this is a weird and convoluted sequence to go through, but it saves having an extra hardware button on the bulb, and it’s probably something that most users won’t ever have to do.
If you look at the two different reset sequences, you see that the older sequence takes less time, since you only turn the bulb on and off for two seconds most times. My guess is that it was hard for people to follow the directions so quickly, so they extended the amount of time the bulb remained on.
I expect that the reason the reset sequence is so long is to make it very very unlikely to happen by chance. Sure, 5-off,8-on,2-off,8-on,2-off,on is pretty unlikely, but if you’re planning to sell 10 million of these things, some of your users are going to run them in areas with flaky power. Double it! A 20-second longer factory reset sequence is much much better than a bulb with a 2% higher chance of factory resetting itself during a lightning storm.
The video is actually very well designed. The different steps in the video take the amount of time you’re supposed to follow them, so you can just play the video and follow along. A video is way easier to follow than the same instructions in a book.
Anyway: hilarious, but there are a lot of careful design decisions that went into this, even though the end result is… not good.
I have Phillips Hue lights in three rooms, and I like that with one touch of my phone, or a request for Alexa I can have the lights move from Watch TV to Reading to Dim Evening, Christmas, or any combo I want. The master bedroom is one of these so I can get the lights to change to a very soft red/white blend and all I have to do is say “Alexa turn lights to Sexy Time.” I’ve had them set up for several months… I’ll let you know how they work when I finally get a chance to try them out 
.
Oh yeah… my reset is to press the large button on the bridge for 5 seconds or so until the blue lights start to flash.
Sure, a hardware button might cost them ten cents more, and that adds up… but having a video like this highlighting the absurd reset sequence also adds up. How many people who are considering smart lights saw this video and decided to get another brand? That ten cents they saved doesn’t help much, if the entire hundred-dollar unit doesn’t sell.
And yeah, given that the reset sequence is so absurd, the video itself makes sense. This is one rare case where a video really is the sensible way to deliver information. It’s just the information that’s being delivered that’s absurd.
Yes, incredibly amazing design. I suspect that there might have been a clock/timer off camera but maybe there’s some other method that video editing wizards use. I wonder how many takes they did?
Honest question here: What is the point of a “smart” light bulb? I mean, I can sort of see a “smart” light fixture–so you can use your smartphone or intracranial neuro-implant or whatever to turn on the living room lights when you’re a block away from the house; or turn on the lights in the bedroom and the hall and the kitchen when you get up to get a glass of water in the middle of the night; or turn off the lights in the kitchen when you get back in bed and realize you left them on. But why put the “brains” (hardware and software) in the thing that’s literally proverbial as “something that gets changed out”? (I know modern post-incandescent bulbs last a lot longer than the old-fashioned “
” ones, but I would still think a table lamp or a floor lamp or a chandelier would last even longer than that.)
They last waaaaaay longer. A smart bulb is better than a smart fixture because you don’t want to have to design a billion different types of fixtures to replace the ones everyone already has/likes/creates.
I mean, I’m sure there is some market for smart fixtures but you are going to get rich selling bulbs that turn any fixture into a smart one.
I just bought one. It’s more useful than I expected.
The primary advantage of a smart bulb vs a smart plug/smart fixture - I already have those - is dimming. Also, color control if you have a bulb capable of that.
But “smart switches can dim also?” you say. Yes, but LEDs have a big problem with dimming precisely when the dimming is being done from the power source. What tends to happen is the bulb dims a little, then starts to flicker or just goes dark.
It’s complex to explain why (it has to do with the power supply in the bulb), but in any case, smart bulbs fix the problem. The bulb always gets full power, and then internally can precisely dim itself down all the way, hitting 1% brightness in principle.
Anyways you can now get multiple smart bulbs, group them together in the Alexa app, then laze in bed or the couch and set the light to 1% or 50% or full, etc.
It’s something you can’t do well at a higher level of granularity. And the primary use case I have is I can settle into my bed and then later turn off the lights without having to get up.
Damn you. Now my 3-level bedside lamp disgusts me. Sitting there, demanding I reach for it. I’m throwing it out this second.
There are smart light fixtures.
But I like my current, dumb light fixtures and I want to be able to talk to the lights to tell them to turn on and off - so smart light bulbs. (I admit - it’s “the clapper.” Just more expensive and slightly more specific).
Hmmm…
My favorite comment was:
Again, a hardware button means you need physical access to the bulb. That might not be easy if it’s mounted 15 feet overhead, or inside a fixture that is not easy to access. I’ll bet you’d get pretty annoyed if you were constantly having to get out a ladder or get a screwdriver to pair your phone.
I can tell you how bad UIs like this get made. Someone probably had a bright (heh) idea to use power cycles to program the bulb. They made the case to management, probably pointing out that it would save on the switch and that it would be super cool to be able to remotely control the bulb programming from a light switch. It was probably ‘free’ in that it didn’t take any extra hardware - just some software. That lowers complexity and reduces failure modes and reduces unit cost.
So the engineer got his way, and the bulb was designed to work this way. They may have brought in a UI guy later, but tied his hands by demanding he work with the system they designed. And there may have been more limitations - for instance, flicking the switch on and off quickly could cause the bulb to wear out sooner, and could screw up any other regular bulbs on the same circuit. So the UI guy might have been given some parameters for how quickly the switch can be thrown.
At some point, someone might have said “You know, this UI sucks.” But by that time the ship has sailed and no one is going to redesign the bulb hardware because a UI guy doesn’t like the interface. So they live with what they’ve got, and they blame the UI guy for not doing better even though he had no choice given what he had to work with. I saw this kind of thing many, many times. It happened to me many, many times.
This is why UI design should always be done FIRST, and prototypes should be tested before a line of code is written or any hardware built. But in my experience, that almost never happens in companies that aren’t primarily UI focused. They all CLAIM that they care about UI, but the way they insert UI design into the engineering lifecycle shows that they really don’t care.
The optimal setup would be a device model that can’t fail this way in the first place. For something like a smart bulb, you need the thing to handle all possible cases with as few user required paths as possible.
The firmware shouldn’t be able to fail in a way that needs resetting. Either it should be connected to (wifi or p2p radio, depending on the bulb kind) or not. The only data the firmware should write should be the info for this remote device.
If no connection can be established, the bulb should enter a mode where it can be accessed and all more complex interactions should be handled by an app on the user’s phone or tablet or GE’s remote servers.
No, you’re going to get rich by selling replacement switches when the ones that came with the light fixtures wear out from being turned on and off so many times.
Yeah, General Electric should get on that. 
o_o
>_<
o_0
There’s an accessory button on my key fob? Where?
Good luck writing a wifi/bluetooth/zigbee/whatever stack that can’t crash. It’s probably not impossible, but it’s very expensive. Way more expensive than putting a hardware switch on every bulb.
The process in the video isn’t for pairing. It’s a factory reset. Most users will never need to do the thing in the video (or get out a ladder and push a hypothetical button).
But you’re right. Even if they had a hardware button, you’d still want a sequence like this that didn’t require getting to the bulb. Because some of them are a pain to get to. So a hardware button doesn’t solve the problem, and you still have to make this video and be made fun of.
Well, yeah but that’s pretty hard to foresee. Imagine being at the design meeting where this was being considered and saying “But what if the instructional video we publish is so unintentionally hilarious that it goes viral and our brand suffers a PR hit?”
I think you’re just being snarky, but on the off chance you misunderstood my comment, I didn’t mean that it would be hard to make this video, just that it was a well-made video for its intended purpose.