Automatic diagnosis
Transparency through the model
This chapter describes, how diagnosis arises in Selmo and why it does not require additional programming is required.
Diagnosis in Selmo is:
not an afterthought feature
not a collection of error messages
not UI logic
But:
The visible deviation between modeled target behavior and the real actual state.
Where messages originate
Diagnosis arises exclusively in the model.
Specifically at the point where:
a state is active
a zone in this state has a meaning
a expectation or condition is violated
Typical sources of diagnosis are:
S-zones whose expected behavior is not meti-zones whose condition is violatedCMZ whose permanent condition is not met
plausibility checks (e.g. pair check)
Every diagnosis has a clear model origin.
There are no messages:
without zone
without state
without a defined expectation
Why no extra programming is necessary
In classic systems, diagnosis is often:
programmed manually
duplicated in many places
separated from the actual process
In Selmo this is not necessary because:
behavior is modeled explicitly
expectations are described unambiguously
monitoring is an integral part of the model
The model already contains:
what is expected
when it is expected
how to react in case of deviation
When the behavior is described, the diagnosis follows inevitably.
There are therefore:
no additional error queries
no "if error then message" logic
no manually maintained message texts
Relationship zone ↔ message
The zone is the smallest diagnosable unit in Selmo.
Each zone has:
a unique meaning
a clear context (sequence & state)
a defined HMI text
When a deviation occurs:
this exact zone is identified
this exact context is displayed
this exact expectation is named
The message thus answers:
Which zone is affected?
In which state?
Which condition is violated?
The diagnosis not only explains that something is wrong, but why the process cannot continue.
Determinism of the diagnosis
Since diagnosis arises from the model, it is:
deterministic
reproducible
not interpretation-dependent
Same situation means:
same message
same cause
same reaction
This applies:
in automatic operation
in manual operation
during commissioning
on restart
Distinction from classic error messages
For clarification:
Diagnosis ≠ PLC error bit
Diagnosis ≠ collective fault
Diagnosis ≠ UI text
Diagnosis ≠ operator interpretation
Diagnosis is:
Model-based transparency about deviations.
Benefits of automatic diagnosis
Through automatic diagnosis, the following is achieved:
troubleshooting becomes faster
commissioning becomes safer
operation becomes more comprehensible
documentation consistent
responsibility justifiable
Transparency does not arise from more messages, but from the right messages.
Summary
Automatic diagnosis in Selmo:
arises directly from the model
does not require additional programming
is clearly assigned to zone and state
remains the same in all modes of operation
If the model is correct, the diagnosis is also correct.
Last updated
Was this helpful?