Hacker News Reader

Comment by wernerb | Hacker News Reader

This looks really cool. I wonder though if my bed would not then be dependent on your servers at some point as I consider such a product as a long-term investment. For example, can it live without the server-side aggregation? Would you consider opensourcing parts of the server to enable people to use their own servers in the long run? Would you consider adding a respectable API that follows the ideas of for example, the philips hue? http://www.developers.meethue.com/philips-hue-api

View on HN

It's nice that you mention because we inspired both our API and architecture to Philips Hue: the REST api is served by both our cloud servers and by the device locally.

The app connects to the unit via wi-fi when you are home, which means you can control the product even if the internet (or our service) is down.

Opensourcing a lot of our stack is part of the plan, we take a lot from OSS and we want to give back. As far as the hardware, we'll make sure it's a pleasure to hack on, remember that we have to do it every day :)

How much of this do you have working already? I've been using Zeo and miss them a lot, tracking sleep phases or heartrate through this seems very ambitious.

Also, some physical switch for microphone could be nice..

We never record the audio from the mic, we are only capable of detecting loud sounds or a sustained vibration (snoring).

It's a really cool idea and great job on getting this far; I think I might order one.

Does that mean both the device and your cloud have apis, or does it mean that an app which communicates only with the device can do the same things as your app which talks to both?

Any 3rd party app can do everything our app can do, the APIs we use internally are the ones that will be publicly available.

This is a commitment that inspires cooperation. The quantified self community needs gadgets like these, but also open platforms that encourages tinkering.

I just purchased a Luna Developer Pack[1].

[1]http://lunasleep.com/extras/

Nothing but respect for that. You've pushed me over the edge and I've preordered one now :)

That sounds incredible, I'm excited!

This is a very valid concern because I used to own a Zeo headband that did a lot of the same stuff, and when the company went belly-up the cloud service disappeared, leaving all their customers with a useless bedtime fashion accessory.

Really encouraging to hear the developer's reply that the their APIs will be accessible over wifi and that they'll even be open-sourcing some of their stack. Wonder if they specifically learned from what happened with Zeo. I'll definitely consider this as a replacement.

Actually Zeo had a very nice serial connection and I never uploaded any of their data, pushing it in realtime to my HA system. But yeah, I cry a lot that they are gone, it's almost impossible now to get a headband replacement.

> product as a long-term investment

Having an internet outage (or DDOS attack) render your "smart" home appliance unusable should be considered unacceptable product design. That said, heated mattress pads on the market now have a finite lifespan. The heating elements gradually lose their ability to heat, necessitating replacement after a perhaps 3-5 years, give or take. Unless they've engineered a much longer-lasting heating system, this significantly limits the window of risk.

If they have engineered a pad with significantly longer-lasting heating elements they should trump that fact LOUDLY, as it'll be a big plus to existing fans of heated mattress pads.

+1 on the API.

Reply on news.ycombinator.com