Hueberry (little project I'm developing)

jetboy

Not really Mac or Emulation related (although you'll see that a Mac is my main machine). I'll cut to the chase, it's a shameless plug for a little project I'm developing in my spare time. I'd really like some advice on this one given that most members of this board have damn good technical minds, and some probably have a good idea of how product development works. Below is a link to a quick video I put together:

https://vimeo.com/116406995

Using a Raspberry Pi I've made a little device that controls the lights (all Philips Hue bulbs) in my home. I've done this using the standard Raspbian distribution... however, I've set it all up the way I like it and put together a bunch of Bash scripts. The Bash scripts are by no means awesome, but I guess as a 'proof of concept' kind of thing, they demonstrate that I can not only make on/off scripts, but can also code some logic into it.

It has WiFi (via a $3 USB chip... daaaamn they are cheap now!!) and the OS is on a ~$10 SD card. All up (including the touch screen) the thing probably cost me ~$75 AUD (mostly before out dollar crashed... DOH!) Anyhow, given what's out there and given what Philips charges for their gear (and the fact this could do more than just control Hue lights)... I think it's pretty cheap.

Any advice on this one? I'm keen to turn it into a product and possibly do a bit of a kickstarter. Questions:
- What do you think I'll need to do/improve/develop in order to get it up to the kind of standard that people could be willing to back?
- What automation features could it be cool to have? I'm sorta hoping to have it work with my balcony garden (watering the plants), my aircon, my oven and my sound system. However, the lights are the priority and I want that control/setup to be REALLY good before I start going anywhere else.
- The design... is the lamp cool, or is that just me? (Is my idea a fundamental flaw, or could the idea of having an automation product as 'furniture' work?) My idea is to develop upon that idea and make it a polished style (rather than a raspberry pi stuck onto a cheapo lamp stand), but I like the wooden lamp look because it's normal/simplistic/friendly rather than nerdy IMO.
- Any other advice based on your opinions/skills/experience? In particular, any general product development advice? My goal isn't to make millions out of this thing, but for a while I've wanted to develop a product (even if it's just for say 100 people as a neat gimmick). I'd love to develop my knowledge of how product development/production works (in particular, in the IT world).

Thanks in advance and sorry about the shameless plug! :D

Edit: I made a little instructables site for it...
http://www.instructables.com/id/Hueberry/

mossy_11

I think it's kinda cool.

The technical side is a bit outside of my expertise (my Bash knowledge isn't great, and my electrical engineering knowledge sparse at best), but if you ever do decide to turn it into a product for people to buy I could probably cover it on Gizmag.

dickmedd

A few things I'd suggest:

You could lasercut/mill/3D-print a custom fit enclosure that both holds the Pi, and attaches to the lamp stand. I like a lo-fi aesthetic myself, but I think the Kickstarter market sometimes demands a bit more polish.
We use Pis for dashboards in our office, and have them boot into a fullscreen browser (Chromium lets you run a kiosk mode, as does Midori). As you're using Hue's RESTful interface, I'd be tempted to make a more customised web interface that runs fullscreen. I control an Arduino Yún at work with Flask and Requests, but I like Python :cheer:

(P.S. I really do like this stuff, so if you want any more input, give me a DM/email)

menace690

I've got an idea for you. Well, lights are generally controlled by a light switch... why not make this a wall switch replacement. This would remove any aesthetic concerns for people as well, no need to find furniture that "everyone" likes.

I think the PI would fit perfectly in place. A regular light switch cover cut to shape would be easy enough to test with.

As for the programming, I would definitely switch from Bash Scripts to making a real application of some sort. You could use python, C, anything really, since its Debian. It would add a minimum remove the terminal launch which looks unpolished.

_Em

If you really want to go with the lamp, how about this? http://internetsiao.com/wp-content/uploads/2009/09/imaclamp.jpg

Otherwise, I like the light switch idea -- or even better, a thermostat casing. Make an API for it so that you can remotely control it from an iOS/Android device and you can even toss the visual interface and just create a unit that hides INSIDE a light switch casing or similar.

vitaflo

Maybe I'm missing something but how is this different from say, SmartThings which is $99 and lets you control a ton of things (including Hue lights) out of the box on your phone and also has an API you can write your own scripts to for anything else you want it to do?

jetboy

vitaflo wrote:
Maybe I'm missing something but how is this different from say, SmartThings which is $99 and lets you control a ton of things (including Hue lights) out of the box on your phone and also has an API you can write your own scripts to for anything else you want it to do?

Sure fair question... although, I think I should make it clear that this is something I made myself, it's a DIY project and SmartThings is $199 for a starter kit according to Amazon (http://www.amazon.com/SmartThings-Smart-Starter-NEWEST-VERSION/dp/B00N97U7U4). So-far I've gotten the current iteration of my project down to roughly $65 USD. Not knocking SmartThings (looks cool and I'm sure a lot of people like you will prefer it)... however, I'm the DIY type and believe this is a solid project (which is worth me taking further). I'll try to explain why below.

Firstly just as a note... the project has progressed significantly and I'm now at a point where I'm going to hold back on releasing pictures/software updates until I have something that I'm 100% happy to show everybody (this was a demo in order to drum up a bit of interest and I think it achieved its purpose since it got ~800 views and I received a bunch of feedback both positive and negative). It's now a single python app without the lampshade stand.

My main answer is that this is a DIY project (not sure if I can calculate the value of the DIY experiences I have gained). Also, it's 100% open... it's Linux, a Raspberry Pi, Common components and an open-source python app. Yes, commercial solutions exist and I don't think I could argue that they aren't more feature complete/polished at this stage. However, IMO none quite fit my needs/wants.

I guess it's like those quadcopters? You can buy a decent one for $25 bucks from ebay now that has a camera, bluetooth, inbuilt tricks, a moulded design and they'll be guaranteed to work out of the box with no understanding of physics/designing quadcopters...etc. However, big communities exist where people spend months learning about the physics and putting together their own solutions using balsa wood/parts from the hobby store.

menace690

Good luck!