Sample won for personal project = Home Assistant

Banana Pi SBCs design BPI-M2+/BPI-M2 Zero/BPI-P2 Zero(H3)
1

Hi,

I won this Banana Pi M2+ and is really happy to received it.

Evolution steps :

  1. I bought its acrylic case
  2. I bought its OV5460 CSI camera
  3. I installed ARMBian image on a microSD and boot it perfectly

I though to be able to use this BPi M2+ for even a tiny personal cloud called Cozy but as I already have a Banana Pi R1 router with a 2.5" onboard SATA harddrive I finally replaced its operating system named OpenWRT by Bananian (Debian like). Now I got Bananian working on this BPi-R1 I need to try changing it to ARMBian to stay constant in my used operating systems. By using even Bananian or ARMBian on this BPi-R1 I will use it for the tiny cloud Cozy (due to its integrated harddrive).

Then I though use the BPi-M2+ for Home Assistant called Gladys project. But I still does not have home automation protocol driver (ZWave, 433MHz, or other) then to reduce cost I decided to use MQTT through the existing personal network. MQTT is the main secure (using SSL/TLS) way to communicate with object of things (self created with Arduino Nano).

Even if Gladys project is able to use MQTT I saw it run with Node.js I did not studied. To reduce the use of BPi-M2+ memory I decided to use Home Assistant in python project because python and specifically Home Assistant has its own web server. It means I do not need to run Web Server (Apache or NGINX) before running on it the home assistant that use Node.js. Just running Home Assistant is enough then wil just communicate with small network packet by using MQTT protocol.

The main reason I decided to use Home Assistant in place of Gladys project is then the smallest size of Home Assistant in place of Gladys project. But also a better control of its use due to python langage. GPIO could stay easily used with python. Camera (OpenCV) could also be used with python (Home Assistant kernel).

Then the next steps will be : 4) Install operating system on the onboard integrated emmc 5) test using the onboard bluetooth to talk to a portable bluetooth speaker allowing to take the speaker where I change the room place in my house. 6) test using integrated micro (on the portable speaker) to see if I may ask some small sentences to Home Assistant 7) using GPIOs for events doorbell and entrance door switch detection 8) use a small USB hub to connect USB keys when back home to see who is back or gone and then let Home Assistant saying what is needed to those people

As you can see I still have a lot of work but is so busy in other things I need more time to progress on it.

Still sincerely thanks for your sample I will really appreciate to progress in the Home Assistant project.

1 Like
2

thank you for you support ,hope you can share your project on forum.:slight_smile:

3

Great :wink:

The test to install it on the eMMC was good but it seems to not really use the installed eMMC. I will then try this later when the project will become more functionnal.

Home Assistant has been installed properly and works great.

The new step to progress will be pairing the BPi-M2+ with an external Bluetooth speaker to let the BPi-M2+ telling what I want (CPU temperature, …). I suppose this step must be faster than the previously.

After that I will use the work of somebody else to simultaneously use Home Assistant and the face recognition (OpenCV) program. As I saw Home Assistant let more than 800 Mb of free memory I suppose this face recognition may then be used simultaneously.

When those two points will be finished I will try to progress with GPIO’s.

Then the BPi-M2+ could be put as door peephole with a doorbell button. It will becomes the perfect Home Assistant keeping an eye on my house :wink:

Have a nice day.

Miguipda :wink:

4

Following this guide:

I managed to run it on a banana pi m2u, as a build i used a raspjessie.

Be careful not to run the command:

Sudo apt-get upgrade