https://reference.opcfoundatio.....docs/4.5.4

You can add the IDeviceHealthType  to your subclass.

Any Server vendor is free to add additional locales for LocalizedText values.

We have once more some problems with unclear specifications and behaviours: various companion specifications define LocalizedText for state machine state names - but with no informations regarding LocaleID.

E.g. MachineryItemState, MachineryOperationMode: defines different state and uses CurrentState[LocalizedText]

E.g. ISA95JobStateDataType uses StateText[LocalizedText]

But both of them in addition to the Part-16 doens't provide any information how to deal with localeID. The only one I've found is PackML - but they don't define a invairant locale.

What should we define/implement? Are the provided StateNames from the specifications implicit defined as invariant locale?

Best regards

Patrick

If the Method does not have arguments the associated arguments properties should not exist.

in general these state machines express different things. The State Machines in OPC UA for Machinery Part 1 express the state and operation modes of the machine (or component). These state machines are kept very generic, so that every domain can use them. It is up to the domains to use it like this or to add deeper information in Substate Machines. Examples for that can be found in OPC 40084-1 Chapter 7.4.1 or in OPC 40501 Chapter 8.3.10.

The State Machine in PackML rather expresses state of the production. Therefore it does not make sense from my point of view to map this with OPC UA for Machinery Part 1 without taking other information into account.

But furthermore there is the State Machine for the Job State in OPC 10031-4 OPC UA for ISA 95 Job Control (see figure 5). This state machine can be used alone or combined with OPC UA for Machinery. With the consideration of both you can combine the Machine Status and the status of the current job, which leads to a similar expression of the production as is done in Pack ML.

I hope this is helpful.

Best wishes

Heiko

I tried to find if it is allowed to have UA method argument without Value in Part 3: Address Space Model and Part 6 - Mappings but I could not find any clear hint. Wanted to be sure if it is allowed to have UA Variable as method argument without any Value tag which describes the method arguments as in following UAMethod?

<UAMethod NodeId="i=8888" BrowseName="GetMonitoredItems" Executable="false" UserExecutable="false">
  <DisplayName Locale="en">GetMonitoredItems</DisplayName>
  <References>
     <Reference ReferenceType="HasComponent" IsForward="false">i=2253</Reference>
  </References>
</UAMethod>

<UAVariable NodeId="i=8889" BrowseName="InputArguments" DataType="i=296" ValueRank="1">
  <DisplayName Locale="en">InputArguments</DisplayName>
  <References>
    <Reference ReferenceType="HasTypeDefinition">i=68</Reference>
    <Reference ReferenceType="HasProperty" IsForward="false">i=8888</Reference>
  </References>
</UAVariable>

In case UA method has no arguments, is there a need of corresponding UAVariable to describe the arguments?

Best regards

Ahmad

Currently I'm struggling with creating a mapping to enable customers to calculate KPIs with different companion specifications (because they have different machines form different vendors at the shop floor)...

• Is it possible to mapp PackML-States to MachineryItemStates or should this solved with a combination of MachineryItemStates & MachineryOperationMode?
• MachineryOperationMode mode is extremley wide and unclear (should this be the intention of a standard?) - possible interpretatons are
  • production mode like automatic / manual (e.g. single step) / maintenance (e.g. limitted speed, re-calibrate axis positions...)
  • defined from a operation mode switch with impact to safety (e.g. SLS)

My fist idea was something like this

• PackML-Wait-States
  (MachineryItemState = OutOfService) & (MachineryOperationMode = None) -> ABORTED
  (MachineryItemState = OutOfService) & (MachineryOperationMode = Maintenance) -> HELD
  (MachineryItemState = NotExecuting) & (MachineryOperationMode = None) -> STOPPED, IDLE, COMPLETE
  (MachineryItemState = NotExecuting) & (MachineryOperationMode = Processing) -> SUSPENDED
• PackML-Acting-States
  (MachineryItemState = Executing) & (MachineryOperationMode = Setup) -> RESETING, STARTING, COMPLETING, UNSUSPENDING, UNHOLDING, CLEARING, STOPPING
  (MachineryItemState = Executing) & (MachineryOperationMode = None) -> ABORTING
  (MachineryItemState = Executing) & (MachineryOperationMode = Processing) -> EXECUTE, HOLDING, SUSPENDING

But the guidline for OEE with PackML defines something like this:

Time machine is producing product (EXECUTE + HOLDING + SUSPENDING)

HOLDING and SUPENDING because bringing the machine to a controlled stop with not creating scrap (so there could be partial produced goods inside the machine).

Time machine is in faliure (HELD + STOPPED + ABORDET)

This is strange:

• STOPPED isn't defined as error-state
• HOLDING and ABORTING are acting states from a Hold or Abort command - in general there will be an Error to trigger this transition. so why are they not used for "time machine is in failure"? (this would have an impact to my mapping).

this would change the mapping to something like this

(MachineryItemState = Executing) & (MachineryOperationMode = Maintenance) -> HOLDING
(MachineryItemState = Executing) & (MachineryOperationMode = Processing) -> EXECUTE, SUSPENDING

what do you think? what would be the more appropriate interpretation? or am I completely wrong?

Best regards

Patrick

I tried to search regarding ListOfExtensionObject in OPC UA Part 6 - Mappings. In OPC UA Part 6 - Mappings an example is illustrated for ListOfExtensionObject. Recently I received a NodeSet file which contained an empty ListOfExtensionObject in blue color in following XML snippet.

Is it recommended to have ListOfExtensionObject in following way or is it not a valid way to describe ListOfExtensionObject  ? 

<UAVariable NodeId="s=128.OutputArguments" BrowseName="OutputArguments" DataType="i=296" ValueRank="1" ArrayDimensions="0" MinimumSamplingInterval="100">
<DisplayName Locale="en">OutputArguments</DisplayName>
<References>
<Reference ReferenceType="HasTypeDefinition">i=68</Reference>
<Reference ReferenceType="HasProperty" IsForward="false">i=12873</Reference>
</References>
<Value>
<ListOfExtensionObject xmlns="https://opcfoundation.org/UA/2008/02/Types.xsd" />
</Value>
</UAVariable>

Best regards

Ahmad

They are in best position to help with open62541 coding questions.

we are trying to add the OpcUARoboticsNodSet2 to our opc ua server of our robotic system.
The DI nodeset is integrated successfully and shown in the structure as well as our internal opc ua functionality in UA Expert.

The robotics nodeset is not yet visiable, but we don't have any build errors. Only when we are editing the server.cpp file with addObjectNode (from open62541, according to the sample server example https://github.com/umati/Sampl.....dRobot.cpp) the object with it's children is shown.

std::stringstream ss;
ss << "http://example.com/" << Name << "/";
m_nsIndex = UA_Server_addNamespace(m_pServer, ss.str().c_str());
UA_ObjectAttributes objAttr = UA_ObjectAttributes_default;

objAttr.displayName = UA_LOCALIZEDTEXT_ALLOC("", Name.c_str());
objAttr.eventNotifier = UA_EVENTNOTIFIERTYPE_SUBSCRIBETOEVENTS;

auto status = UA_Server_addObjectNode(
m_pServer,
UA_NODEID_NUMERIC(m_nsIndex, 0),
UA_NODEID_NUMERIC(nsFromUri(m_pServer, constants::NsDIUri), UA_DIID_DEVICESET),
UA_NODEID_NUMERIC(0, UA_NS0ID_ORGANIZES),
*open62541Cpp::UA_QualifiedName(m_nsIndex, Name).QualifiedName,
UA_NODEID_NUMERIC(nsFromUri(m_pServer, constants::NsRoboticsUri), UA_ROBOTICSID_MOTIONDEVICETYPE),
objAttr,
nullptr,
m_nodeId.NodeId);

The xml implementation for axis_1 to axis_7 for example is also not shown even if the motion device is added with the addObjectNode function.
Is it necessary to add all object manually or should the content of the robotics xml structure be shown by integration to the cmake file?

rudolf.pribis@stuba.sk said
Ok, I found something in OPC 10000-100: Devices Annex C (informative). I was thinking about something similar, like to use different views.

"In this case, the device is represented by an Object “MyDevice” where the vendor of the OPC UA

Application can provide its specific knowledge of the device. In addition, the Object has two

components called FDTView and AutoIdView in the figure, containing the information as defined in

the corresponding companion specifications."

I am looking forward also for other suggestions...

Hi,

you already found a good place describing a solution for your issue, that is composition.

If composition is not an adequate solution for your case, you can also use a specific ReferenceType defined in Part 23 of the base specification, called "RepresentsSameEntityAs" or one of its subtypes. This is for example used in OPC 30142: OPC UA for PROFINET Remote IO or OPC 10000-81 OPC UA FX.

"In this case, the device is represented by an Object “MyDevice” where the vendor of the OPC UA
Application can provide its specific knowledge of the device. In addition, the Object has two
components called FDTView and AutoIdView in the figure, containing the information as defined in
the corresponding companion specifications."

I am looking forward also for other suggestions...

Is there an explanation or guide how to in a proper way model an "asset" in several information models?

As I understood, different use cases requires different viewpoint on data for instance

- PackML for some KPI calculations

- ISA95 for connection with Production Planning

- OpenPLC for addressing operational values

- AAS for creation I4.0 Component, or Digital Twin.

How the interconnection between all these views should be made? Is there somewhere a best practices which describes references between modeling in several information models?