Documentation from the model
Evidence & Audit
This chapter describes, how documentation is created in Selmo and why it does not have to be created or maintained manually.
Documentation in Selmo is:
no additional effort
not a separate document
not a subsequent result
But:
A structured view of the formal model.
The fundamental problem of classic documentation
In classic automation projects, documentation is often:
temporally downstream
decoupled from the code
incomplete or out of date
hard to trace
Typical situations:
the code works, the docs are outdated
changes are not applied consistently
behavior cannot be conclusively proven
audit questions must be interpreted
The problem is not a lack of diligence, but the lack of a common source of truth.
The model as the single source
In Selmo the model is that source.
The model already contains:
process (sequences, states)
Structure (plant, hardware zones, zones)
behavior (Bit-Control)
Monitoring (interlocks, CMZ, checks)
reactions (stop, release, diagnosis)
That provides everything that is necessary for documentation.
What is described in the model, can be documented. What is not modeled, cannot be documented.
Which documentation is generated from the model
From the model the following can be automatically derived:
Process descriptions → states, transitions, sequences
Structure overviews → plant, hardware zones, sequences, zones
Safety arguments → interlocks, CMZ, plausibility checks
Diagnostic descriptions → zones, expectations, deviations
Operating concepts → manual and automatic operation, HMI semantics
This documentation is:
consistent
up-to-date
unambiguously
fully traceable
Traceability and audit
Through the formal model it is always traceable:
which state is active
which conditions apply
which monitoring applies
why a reaction occurs
For audits this means:
no interpretation of code
no assumptions about behavior
no detached descriptions
Instead:
An explicit, verifiable behavior model.
Documentation and changes
Changes are made in Selmo:
to the model
not to individual documents
This means:
every change is central
all derivations remain consistent
no diverging document versions
Change in the model = change in code, diagnostics and documentation.
Distinction from classic documents
To clarify:
Documentation ≠ PDF
Documentation ≠ textual description
Documentation ≠ collection of comments
Documentation is:
a formal representation of the model from different perspectives.
Benefits of model-based documentation
Through model-based documentation the following becomes:
engineering more transparent
operation safer
commissioning more efficient
auditing easier
responsibility provable
Documentation becomes from a mandatory document an integral part of the system.
Summary
Documentation in Selmo:
is generated automatically from the model
is always up to date
is clearly traceable
supports evidence and audit
Whoever understands the model, also understands the documentation.
Last updated
Was this helpful?

