Hacker News new | past | comments | ask | show | jobs | submit login

From an efficiency/sustainability point of view, I have a basic discomfort with the idea of throwing away all the wifi/controller stuff each time you need a new bulb. I know LEDs don't run out very fast, but they do run out. This has cost implications too.

What about a setup where there are two levels of screw: the LED bulb screws into a fixture with the wifi stuff inside, which then screws into the light fixture? Then when the bulb goes out, you can just unscrew the LEDs without having to throw away all the other components.

Maybe there's something I'm missing though; I am by no means knowledgable about electronics.




It's funny, I had the exact opposite worry. With a rated life of 25 years, the WiFi components are likely to become obsolete long before the bulb itself quits. If I buy one of these, am I going to end up maintaining an obsolete WiFi network in the 2030s just to run my household lights?


The bulbs communicate with their master bulb using 802.15.4, so you'd just need one replacement master that supports whatever new wifi standard emerges. Of course the company itself also needs to still exist in 25 years...


Given how long most wifi stuff I've seen lasts, and the fact that the bulbs are rated for 25 years, I'm more worried about the wifi components failing and bulbs becoming useless, of the software becoming obsolete and not being able to run on newer devices.


From an efficiency point of view, what about all of these bulbs which will now be drawing power 24/7 for the wifi components, along with everything else in the house that has a standby function?

Just get LED bulbs and use the off switch.


Only the master bulb runs WiFi, the rest run something (zigbee?) that's lower power.


The WiFi radios 2 decades from now will probably be much different, anyway. Just look how quickly we've come since 802.11 A.


Even more reason for the double screw idea. You can upgrade the WiFi radio when your network becomes incompatible in the future.


This is definitely feature overload for a typical person.

I find it funny how on HN there arguments that non-replaceable batteries in MBPs are okay, and at the same time, that a small gadget must have user-replaceable radio.


It'll be a bit costlier, but as long as it comes in one piece (with separate parts screwed into each other), then it'll look exactly the same to the typical person.

Except that those who bother to read the manual will find out they can save some money by not chucking away the whole thing.


Non-replaceable batteries are not okay. And while I've seen plenty of Mac fans here, I can't say I've seen anybody defend this particular anti-feature.


Personally, I prefer to have better industrial design + a more solid feeling device to a removable battery (though all things equal, a replaceable battery would be nice...it just is far down my list). And I've seen similar defenses quite a bit.


This is a hardware project, not a software one. In virtual world we can talk about keeping features to the minimum and iterating the product as fast as we can. But we don't live in post-scarcity world where energy is free. Designing products for more-less planned obsolescence, while maybe good from marketing point of view, strikes me as wrong and harmful to the global society.


I think that this is because non-replaceable batteries in laptops and non-replaceable radios in lights are not directly analogous.

For one thing (I assume) you could get the battery replaced/fixed at an Apple store if you needed; the real issue is doing it yourself. Retrofitting new radios into old bulbs seems less likely to be a service that will be offered.

Additionally, the lifespan of the computer is 20 years less than the lifespan of the bulb. With the computer, if you had to get a new one when the battery ran out, in most cases it wouldn't be too much of a tragedy since you would be looking forward to a bigger better machine anyway. For the bulb, though, the wifi component could be obsoleted within a few years, while the actual light source has the capacity to run for another 20. So for each object, the ratio of when one part fails to when the user would want to replace it anyway is quite different.

Plus the non-replaceable batteries have the theoretical up-sides of leading to better build quality and industrial design, which is less apparently true of the bulb.


HN is not homogenous.


It's important to note that the batteries are replaceable with a visit to the Genius bar. In the case of this light, there is no such option. You won't be able to take 20-30 lightbulbs from your house and upgrade the radio in five years, you'll just have to replace the lightbulbs.


I'd be generally more worried about the power supply components. There's a tiny switchmode supply in there, which means capacitors. Which would be my main concern for early failures, I'd think far earlier than the LED under normal conditions.

Hopefully they'll be able to arrange good quality caps for manufacturing, even better if the whole unit can be opened up and serviced. Although it's questionable whether 99.5% of the market would ever try and fix it.


No, I agree. This idea is braindead because it's much more effective to focus on a device you plug in between the outlet in your wall and the devices you want to control power to. A standard already exists for half the project which is called x10.


Have you ever worked with X10? It's a standard finalised in 1975, and this shows.

Take a look under "Commands getting lost", "Relatively slow", "Limited functionality", "Interference and lack of encryption" for a quick overview: https://en.wikipedia.org/wiki/X10_(industry_standard)#Limita...

Certainly there are good things about having a standard for home automation, but there are many good reasons for not adopting X10.

(I take heart in the fact this project has some committed hacker types involved with it, Andy Gelme was a founder of the Melbourne hackerspace and I absolutely believe the Kickstarter project when it says the device will have a specified API and be hackable. Not an open standard, but at least interopable.)


Right, because there's no difference between the average person changing a lightbulb and messing with their wiring.

This is probably how things are going to work until its commonplace for houses to be built with x10.


I think this is targeted at ceiling lights ... and it isn't easy to futz with in-the-wall wiring for the typical user.


Sadly, most of the x10 adapters don't work with LED bulbs. (the "lamp" ones, the "appliance" ones should be okay)


Only one master bulb has a full WiFi setup, which consumes 2-3W. The rest of the bulbs communicate with the master using an 802.15.4 mesh network (which is very very energy efficient).


" the LED bulb screws into a fixture with the wifi stuff inside"

I would also think that the electronics could be small enough to simply be an insert between one of the conductors and the bulb. There are, for example, inserts (size of a coin iirc) that simply can flash a bulb when seated in between the screwbase and the bulb.


This is possible, but the features would be limited to only on/off and basic dimming.


The hardware in question costs only pennies, probably less than the shell of the bulb itself.


This is a brilliant idea




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: