Wireless Mesh 4-20mA sensor

Hi all.
I'd like to ask some direction about a product I want to develop.

My customer has some sensors distributed along his manufacturing plant and has plans to implement all equipments inside an automation project.
As could use a lot of sensors (4-20mA), instead buying a lot of analog input PLC cards maybe use BLE Mesh devices can reduce his costs for not buying PLC cards and infrastructure installations.
What I ask is, as nRF52 has BLE Mesh, Thread and Zigbee which protocol should I consider and why?
Consider I must use Siemens PLC's (S7-1200) to control his plant and equipments, which way could be implemented a gateway BLE Mesh to communicate to those PLC's?

  • I am re-phrasing the question to understand this better, you are stating that you can use a set of analog sensors with mesh to get the analog data to the Siemens infrastructure in the factory. This as you suggested getting the analog data over mesh to a mesh gateway with a OPC-UA or similar end point so it can be queried by the rest of the OPC-UA infrastructure.

    Under this assumption the main things you need to check are the bandwidth that you need from the analog sensors to the gateway, you want to know the time criticality of the data etc. The best way would be to allow the mesh units to generate the events based on the set points and not stream the analog data continuously to the gateway. Continuous streaming requirement will require a bit more rigorous analysis based on the bandwidth needed, timeliness etc.

    The Thread packets are larger as compared to Bluetooth mesh so they can be more easily mapped to an IP frame, however all meshes have bandwidth constraints so that will be significant point to look at. Additionally I am assuming that all the nodes sampling the analog inputs are not powered by batteries so power is not a constraint. 

    Hope this helps, please post more info on bandwidth and timeliness needs so a better analysis can be provided.

  • instead buying a lot of analog input PLC cards maybe use BLE Mesh devices can reduce his costs for not buying PLC cards and infrastructure installations

    Are you intending to develop these "BLE Mesh devices" ?

    It's highly unlikely that the cost of development for one installation will be less than the cost of off-the-shelf units ...

  • Hi,
    at first will be used only for this project but we also will keep developing for others purposes.

  • Is my understanding about the problem correct ? can you comment on that ? thanks

  • I mean, really you should pick the right tool for the job - and you can only really find that by minimizing the cost function given by your constraints.  So really you need to analyze your constraints.  These should include actual financial cost of materials, development time, development resources and training, etc etc.  However, all else being equal, if you have a network with at least several nodes, relatively small amounts of data per node, relatively low thoroughput constraints, and want the possibility of rapidly connecting to some existing commercial networks, then I've been under the impression that zigbee is the way to go.  See here: https://www.eetimes.com/bluetooth-thread-zigbee-mesh-compared/
    But of course, if delivery time is a major constraint, and all you have it a development team that knows ble mesh, then unless you can justify training them, (re)hiring new people, or hiring out the work (which is certainly possible if you think you'll make it back in business volume) then you might want to stick with what you know/have.