Question, not problem - Open Source linux version of a Provisioner (like the Android and IOS Mesh apps)?

[HeneryH]

Very much a beginner at this so if this is inappropriate for this forum do not hesitate to let me know.

Starting out trying to develop a BT Mesh system.  I purchased a nrf53 Dev Kit, a couple of Thingy:52 devices, some standard BT-Mesh Light Bulbs and some Adafruit Feather Dev Boards.

Also have a Raspberry Pi-4 with a rebuilt kernel to enable the crypto APIs, a BlueZ 5.63 update and it is detecting devices using the meshctl command line.

I have an IOS device with the Nordic Mesh app and the light bulb manufacturer app which both allow me to detect and manage a local mesh network.

** Much of the root base items seem like they are OK.

Now I want to think about a Raspberry Pi web app that will allow me to manage the mesh network.  The goal is that if here are more than one users needing to manage the mesh, then how can multiple users share admin when using their respective IOS devices?

Question 1 - Is this a stupid path to even think about?

Question 2 - Are the multiuser limits of the IOS app real?

Question 3 - Are there any linux open source apps that anyone has seen where I can collaborate?

Thanks for listening to the noob questions.

[Elfving]

Hello Joseph!

No worries, we are always available here to answer your questions.

Question 1 - Is this a stupid path to even think about?

Not at all. Both using a Rasberry Pi as a Provisioner and using several apps can work. How you would share the admin privileges with the Rasberry Pi would be up to you, but with the nRF Mesh applications this is already possible to a degree:

In the settings menu, press "Provisioners", and press the add Provisioners button to add more. And then you can use the json format to exchange info between the different Provisioners, which I think should be as easy as importing and exporting it through the app. You should however assign different address spaces to each Provisioner so that you can avoid any issues. I don't know if that would make this less an option for you.

Question 2 - Are the multiuser limits of the IOS app real?

Would you mind expanding on that?

Question 3 - Are there any linux open source apps that anyone has seen where I can collaborate?

You can have a look at Zephyr if you want to. What type of projects did you have in mind? I don't know if I would be able to suggest something to you either way, but maybe other developers in this forum would.

Best regards,

Elfving

[HeneryH]

Thank you for your kindness Elfving.

Your answer to (1) that there is some degree of sharing provisioner rights/abilities is new news to me.  That solves the problem I described in (2).

So, The answer to (2) is that my interpretation of limits to provisioners being only the single device that started the network in incorrect.  See the answer to (1) for more details.

In question (3) I am thinking, maybe incorrectly, of having a more server based network management solution rather than a model where people use their phones.  That server could be designed with some visual mappings of some type to indicate where the nodes are. 

For instance, here is a use case that I am tinkering with...

My mother is elderly at 90 and is mostly pretty independent.  She loves her independence but at her age she is susceptible to getting dizzy if she bends over and stands back up too quickly.  I created a simple BT LE puck that advertises its environmental metrics and a count of the number of button presses it has detected.

I have a simple Node-Red scenario running on a Pi that listens for the advertisements with the metrics and when it detects the button press count increments, then it sends an alert.  The alert in her case is a SMS to her children.  She really did NOT want a commercial system that sends emergency workers to her home upon alert.  She is old school and would rather die that be embarrassed

My problem with the current BLE system is mostly that with the RPi-zero that the server is running on has poor BT range.  Not sure if it is the embedded antenna or some type of RF interference.

BUT - Being a geek, I wanted to try Mesh 1) to fix the problem at hand and 2) wow, this can easily be expanded to neighbors by simply using mesh light bulbs in a string of homes front lamp posts.  They are close enough that the mesh seems like it could work.

Just a geek trying to learn and maybe make a workable solution.

A server based dashboard would likely be better than the nRF Mesh app.

[Elfving]

Wow Joseph,

Let me first off all give you props for thinking this much about your mother, her concerns and well being 

Like you mentioned with the Rasberry Pi, one thing to consider is the distance the signal can reach. Indoor (considering potential obstacles etc) we usually say that Bluetooth has a range of about 10m, and outdoors it can have about 100m. Though of course you could increase the range by increasing the output power, for info on what distances you could achieve with this take a look here. If you were to use Mesh you should optimally cover the same area with some redundancy of relay nodes as well. So the light post relay node system could work for the communication between the homes, but you would need more nodes inside the homes as well to get a good coverage.

Another idea is to use Bluetooth Long Range to get a better connection between your Bluetooth puck and the Rasberry Pi, and see how far that takes you (I can't really give any estimates on how much an indoor connection distance would increase using it).

To me it sounds like an LTE solution would be the best fit. You could for instance base it on the Thingy91, our cloud platform and the SMS example. The GPS would also cover the need for visual mapping of the units.

Best regards,

Elfving