IIoT Unified Namespace Enterprise Example

Поділитися
Вставка
  • Опубліковано 8 вер 2024
  • In this video, Walker cites an example of a Unified Namespace in building a data hub.
    Thanks for watching!
    Subscribe!
    👉 bit.ly/SubTo40...
    LinkedIn 👉 bit.ly/40Solut...
    Twitter 👉 bit.ly/40Solut...
    Want to learn more?
    Join our community Discord Server
    👉 bit.ly/Industr...
    Take the FREE IIoT Mini-Course
    👉 bit.ly/iiotmin...
    #IIoT #Industry40 #DigitalTransformation

КОМЕНТАРІ • 17

  • @abhishekvigh9
    @abhishekvigh9 Рік тому +1

    I love your channel and learning concepts here. I love how you simplify what would otherwise seem complex topics :)

  • @boontecklee592
    @boontecklee592 2 роки тому +6

    I have 3 points to make.
    1) I'm horrified that your example of Unified Name-Space (UNS) is hardware centric instead of process centric. By this I mean you are retaining data from PLC, MES, ERP nodes, etc at their respective node tag in the UNS. In this way, you are just creating a central data hub, not a UNS. Imagine if the PLC hardware is replaced by several IIot devices in the future, then you will have to restructure your namespace at the hub, and make corresponding changes at each and every client! This will defeat the purpose of using UNS in the first place. Instead, if you structure the same data into a hierarchy of process related UNS, starting from nodes in the Enterprise level, followed respectively by levels for Site, Plant, Line, Unit, Machine, Loop, etc (ref. ANSI-SP88) , then irrespective of your hardware replacement, the tags in the UNS will remain unchanged. This should be the way to structure a UNS.
    2) You mentioned that the data hub is a close loop for the entire enterprise's data. This is not so. Instead, it is a "hub and spoke" model where data is radiated to connected clients through a single layer (rather than passing from one hierarchical layer to another).
    3) You mentioned that at the data hub, all PLC tags should be replicated into the UNS. This is not advisable, since it will only cause data overloading. As all PLC programmer knows, the PLC memory space is populated with at least two types of tags, the Internal Logic Tags (for internal logic operations) and the Process Tags (representing process status and outcomes). What need to be replicated at the data hub are the Process Tags, not the Internal logic Tags. Big data analytics and AI should only act on the Process Tags.

  • @oguzhanucar608
    @oguzhanucar608 Рік тому +1

    When a machine stops, how can we capture the leading signal from PLC to understand the root cause of stoppage to use it in the OEE downtime analysis? Because, in our case when there is a machine downtime, within milliseconds, there are several signals are being sent as MQTT message from PLC to server. How can we distinguish the leading signal from lagging ones?

    • @4.0Solutions
      @4.0Solutions  Рік тому

      It may be time to do an update on this one.

  • @patchoreep2
    @patchoreep2 4 роки тому +3

    Hi Intellic Integration!
    I have a question concerning topic granularity/organisation.
    The sparkplug specification demands this structure:
    namespace/group_id/message_type/edge_node_id/[device_id]
    You seem to prefer the ISA-95 equipment hierarchy (me as well, because of the finer granularity of "things").
    If you use the ISA-95 structure, is your solution still sparkplug compliant?
    Also, why does the sparkplug definition only use the group_id to distinguish between Site/Area/cell/unit?
    Thanks for the very informative videos!

    • @samueljaneiro4701
      @samueljaneiro4701 4 роки тому +1

      I'd love to have an answer to this comment. How does the Sparkplug specification namespace fit into the Unified/Enterprise namespace? Is it just another directory within the Enterprise example? Do you have to choose between the ISA-95 structure or the Sparkplug one? Can you have a custom namespace and still have it be Sparkplug compliant? What would that imply?
      It would be great to have an example that addresses all of this.

    • @4.0Solutions
      @4.0Solutions  3 роки тому +1

      We are going to do a video reply to this Jef!

  • @tazhhh
    @tazhhh 4 роки тому

    Thanks author, I enjoy your channel!
    Is there any formal education or any online course that you would recommend to acquire the basics to be able to work with digitalization of industries?
    Regards,
    Tazhi

    • @4.0Solutions
      @4.0Solutions  4 роки тому

      Yes! We actually just launched our Online mentorship program! bit.ly/mentorshipProgram

  • @saipra
    @saipra 4 роки тому +1

    what softwares are used to create namespaces?

    • @4.0Solutions
      @4.0Solutions  4 роки тому +1

      We use Ignition, FactoryStudio, Highbyte, or even an open MQTT Broker.

    • @FreeTrol
      @FreeTrol 3 роки тому

      @@4.0Solutions thx the answer I was looking for :-)

    • @avafab
      @avafab 3 роки тому +1

      @@4.0Solutions Is there a way to create the unified namespace using AWS?

    • @4.0Solutions
      @4.0Solutions  3 роки тому

      @@avafab Adding this to our video topic list Fabrizio. Thank you!

    • @nzxtifyjanne1902
      @nzxtifyjanne1902 3 роки тому

      @@4.0Solutions But all this information needs to be stored somewhere in order to analyze it I guess? Could you use a simple SQL Database for that or do you need something more like a data warehouse?

  • @oguzhanucar608
    @oguzhanucar608 Рік тому

    Come on man :D Machine learning wants every tag? That's not right man. Could you please tell me what the machine learning algorithm does with static data? Answer is nothing. The static data you take from PLC which is not changing through the process of machine is nothing but a straight line in the multidimensional space which does not have any variance. ML algorithms require variance in the data to have predictions. I love your videos, I love what you are teaching about IIoT integration but don't go ML topics. I'm not sure this is your area.