Need help for great opportunity ( >5 houses )

[acekrystal]

Hello All!

I'm working on a funded projects where I try to push more open source solutions rather than expensive closed business solutions.

For one of my projects I got the awesome task to come up with a sensor plan to monitor all energy flows in a building.
We did a lot of conceptual and functional design and I think we have a very awesome Home Sensoring plan version 0.1
The awesome part is that I get to actually build it and it will be deployed in 5 to 20 houses for testing (depending on how many signups there are from the house owners)!

Awesome right!?!
I work on this project at my job, but also in my free time, together with a friend.. A lot of things are still new to us, but we’re making BIG experience steps.
So I'm really busy trying to get a set together for a Demo that I'm going to do at end of September, after that I need to build multiple of those sets.

Need some help working together.
We are currently doing this on our own right now, but would like to try and see what happens if we pull community's in this kind of projects. Or actually the other way around, what would happen if we use (business) project resources for improving opensource community solutions for everyone?
We and the company I'm working for (www.coreon.nl) are really open for idea's and like to support and work together with opensource idea's.


My request:
Even though I would love to do everything on my own, sometimes it is better to aks for help. Therefore I am looking for people who’d like to help me out on this project.


What am I looking for?
1. First of all, I really need you to be motivated to work with me on this project.
2. I’m currently trying to build the whole Arduino sensor nodes --> Raspberry/OdroidC1+ Gateway part, but I have some problems getting its setup (Latest news: *cough* burned my RFM modules..) I’m looking for some people who can help us getting started: build a working sensornode <---> Gateway for us as a starting point as soon as possible.
Extra Details bellow.

3. In case it turns out to be just as much fun as we’d expected it to be: we’re looking for someone to help us build multiple of the final sets later this year)
What do I offer you?
1. Free pizza!
2. We have limited resources, but we’ll pay for all the expenses you’d necessarily have to make to get here and help us out. Next to that you’ll receive a small fee for your hard work.
I’m seeing it as a day, weekend, or maybe a few day’s where we get together and start building.
3. We’ll provide and pay for all the required hardware . We also have soldering equipment, 3D printer, laser-cutter, Prototype PCB CNC machine and other hardware here in office for a nice DIY party, so no hard feelings if you might want to take advantage of that for your own project, just let me know

Details:
I'm lookig at github.com/abouillot/HomeAutomation/tree/master/piGateway connected over 433MHz to arduino (feather 32u4 RFM69HCW 433MHz) with the sensors: PIR, CO2, Temp, Hum.
Sensors are not locked in yet, but at the moment I'm using: Parallax-Pir and the SHD31-D for Temp and Hum. CO2 we are running an experiment on including: MQ135, MG811, TGS4161, K30, and maybe later adding a S8 LP to the test.
My current goal is not really getting the live data (though I see it is better to include this from the start), is currently busy busy writing Python code to have a simple way of adding new "measurements" into the DB. But we’re also looking at things like defining every sensor ID in the DB, and make every sensor node (arduino) with his unique NodeID pull all his sensorID's from the DB. This so we can later make a front end (web) interface, or include it in other projects like OpenHab. Where you can easely manage all your sensors, and edit things on arduino code level though the Web interface. But this is future music =^.^=. I also would like to add the DB to openhab, or maybe move to a better excisting DB platform. But for quick development changes we made our own to fit our ideas. I can see the ERD here BTW to give you an idea of what we build: www.lucidchart.com/invitations/accept/0346fa2a-ea08-4cdf-b5b4-cb58a01b98dc There are awesome ideas, there is some funding, but I can't learn, build and document fast enough to make them all into reality. So at Coreon they poked me for this idea to share it with the community's.


I hope someone shares my idea / interest.

[computourist]

Hello acekrystal ,

Sorry for the late reply, summer holiday ;-)

Interesting project you have. It seems like a real challenge to work out a project like this.

I think you will find that the approach taken in this forum is a bit different from yours:

- most of us use an Arduino-based gateway to connect the radio network to the MQTT broker.
the early attempts in 2015 to use a Raspi as a gateway were not explored any further I believe. The Arduino-based solution is stable (running more than a year without interruption) and fulfills our needs for now. Some have ported the gateway to an Arduino Mega to add more functionality.

- a fixed datamodel is used to identify nodes and sensors/devices. This enables the use of a universal gateway that handles different types of nodes (assuming the correct dataformat is used on the radio link) Configuration of nodes is not database driven like you propose.

- ISO writing proprietary code to read and store sensor data most of us use Openhab as a central controller. Openhab has a broad range of scripting possibilities and integrates easily with other brands and devices.

Please don't read this as a negative comment on your approach, I merely want to point out the differences so you know what knowledge / experiences to expect on this forum.

[acekrystal]

Update:
The past month I learned a lot and was able to get the RFM69 working directly on the Odroid C1+.
I now succesfully got a working test setup with a SHT31-D (temp. Hum.) and K30 (CO2) sensor, reporting in on the MQTT server on the Odroid through 433Mhz... YAY!

Hello computourist

Thanks for your reply!

I'm aware of the fixed Data model that is used in this code. I'm planning to stick with it to not distance my code from this project. But I will not yet change our RAW Sensor DB design to this Data module. I'm not really convinced that this is a good approache for scaling and flexibility. We are building our own listener/pre-processor in python, that will do a Data model conversion for the time being. In this case this python program will sit between the MQTT broker and our RAW sensor DB. But we use it for more things, Like HTTP GET listeners, or direct GPIO reading from the Odroid. This python program wil also be able to change settings of interval times by fields defined in the Raw sensor DB.
I'm planning to stick with this approach for the time beeing in order to not limit ourselfs by technical limitations of exciting platforms.

Though, I really like to add OpenHab later on, mainly as interface for all of this. I havent had the time to play with Openhab yet, but I'm hoping we can build something like a module/plugin for Openhab to add our project to it. Of course we should look if we can re-use the Openhab functions and Data models as much as possible, instead of keeping our own. But these are all next steps . A other reason for not using OpenHab is that I'm really new to this, so  feel like it should first learn it before I get myself into slapping something together that never would work together that way.


But, I'm open for advice or any other help.


And talking about help.
Is there somewhere an up-to-date list of the Data model used with the (pi)Gateway? esspecially about the sensor numbers used.
I would like to know were I should knock on the door if I would like to add more sensors. I'm having a CO2 sensor, but there is not number for it yet as far as I know.
And I need to know how this model handles multiple sensors of the same type. For science purposes I would like to monitor temperatures on 3 different hights in a single room.

I will also keep updates posted in here