PyI40AAS – Python Industry 4.0 Asset Administration Shell

The PyI40AAS project aims to provide an implementation of the Asset Administration Shell (AAS) for Industry 4.0 Systems, compliant with the meta model and interface specification provided in the document “Details of the Asset Administration Shell” (v2.0.1).

Features

• Modelling of AASs as Python objects (according to DotAAS sec. 4)
• (De-)serialization of AAS objects into/from JSON and XML (according to DotAAS sec. 5)
• Reading and writing of AASX package files (according to DotAAS sec. 7)
• Storing of AAS objects in CouchDB
• Compliance checking of AAS XML and JSON files

Project Structure

The PyI40AAS project provides the aas Python package with 5 submodules:

• aas.model: The AAS metamodel implemented in python
• aas.adapter: Adapters for various file formats and storage backends
• aas.compliance_tools: Compliance checker for AAS files
• aas.util: Provides utilities
• aas.examples: Example data and tutorials

License

The PyI40AAS project is dual-licensed under the terms of the Eclipse Public License - v 2.0 and the Apache License Version 2.0. Choose either of the two licenses that better suits your needs.

SPDX-License-Identifier: EPL-2.0 OR Apache-2.0

For more information, especially considering the licenses of included third-party works, please consult the NOTICE file.

Dependencies

PyI40AAS requires the following Python packages to be installed for production usage. These dependencies are listed in setup.py to be fetched automatically when installing with pip:

• python-dateutil (BSD 3-clause License)
• lxml (BSD 3-clause License, using libxml2 under MIT License)
• urllib3 (MIT License)
• pyecma376-2 (Apache License v2.0)
• Werkzeug (BSD 3-clause License)

Optional production usage dependencies:

• For using the Compliance Tool to validate JSON files against the JSON Schema: jsonschema and its dependencies (MIT License, Apache License, PSF License)

Development/testing/example dependencies (see requirements.txt):

• jsonschema and its dependencies (MIT License, Apache License, PSF License)
• psutil (BSD 3-clause License)

Getting Started

Installation

For production usage and building applications with PyI40AAS, we recommended installation from PyPI:

pip install pyi40aas

Example

The following code example shows how to create a Submodel with a Property serialize it into an XML file using PyI40AAS:

Create a Submodel:

from aas import model  # Import all PYI40AAS classes from the model package

identifier = model.Identifier('https://acplt.org/Simple_Submodel', model.IdentifierType.IRI)
submodel = model.Submodel(identification=identifier)

Create a Property and add it to the Submodel:

# create a global reference to a semantic description of the property
semantic_reference = model.Reference(
    (model.Key(
        type_=model.KeyElements.GLOBAL_REFERENCE,
        local=False,
        value='http://acplt.org/Properties/SimpleProperty',
        id_type=model.KeyType.IRI
    ),)
)
property = model.Property(
    id_short='ExampleProperty',  # Identifying string of the element within the submodel namespace
    value_type=model.datatypes.String,  # Data type of the value
    value='exampleValue',  # Value of the property
    semantic_id=semantic_reference  # set the semantic reference
)
submodel.submodel_element.add(property)

Serialize the Submodel to XML:

from aas.adapter.xml import write_aas_xml_file

data: model.DictObjectStore[model.Identifiable] = model.DictObjectStore()
data.add(submodel)
with open('Simple_Submodel.xml', 'w', encoding='utf-8') as f:
    write_aas_xml_file(file=f, data=data)

Examples and Tutorials

For further examples and tutorials, check out the aas.examples-package. Here is a quick overview:

• aas.examples.tutorial_create_simple_aas: Creating an Asset Administration Shell with all required objects via Python code
• aas.examples.tutorial_storage: Manage a larger number of Asset Administration Shells in an ObjectStore and resolve references
• aas.examples.tutorial_serialization_deserialization: Export or import AAS objects to/from JSON and XML documents
• aas.examples.tutorial_aasx: Export or import AAS objects with auxiliary files to/from AASX package files
• aas.examples.tutorial_backend_couchdb: Store, manage and update AAS objects in a CouchDB database

Compliance Tool

The PyI40AAS project contains a compliance tool for testing xml and json files is provided in the aas.compliance_tool-package. Following functionalities are supported:

• create an xml or json file compliant to the official schema containing example Asset Administration Shell elements
• create an aasx file with xml or json files compliant to the official schema containing example Asset Administration Shell elements
• check if a given xml or json file is compliant to the official schema
• check if a given xml, json or aasx file is readable even if it is not compliant to the offical schema
• check if the data in a given xml, json or aasx file is the same as the example data
• check if two given xml, json or aasx files contain the same Asset Administration Shell elements in any order

Invoking should work with either python -m aas.compliance_tool.cli or (when installed correctly and PATH is set correctly) with aas_compliance_check on the command line.

For further usage information consider the aas.compliance_tool-package or invoke with python -m aas.compliance_tool.cli --help respectively aas_compliance_check --help.

Contributing

If you plan contributing code to the PyI40AAS project, please get in touch with us via E-Mail first: m.thies@plt.rwth-aachen.de

Codestyle and Testing

Our code follows the PEP 8 -- Style Guide for Python Code. Additionally, we use PEP 484 -- Type Hints throughout the code to enable type checking the code.

Before submitting any changes, make sure to let mypy and pycodestyle check your code and run the unit tests with Python's builtin unittest. To install the required tools, use:

pip install mypy pycodestyle

Running all checks:

mypy aas test
pycodestyle --max-line-length 120 aas test
python -m unittest

We aim to cover our code with test by at least 80%. To check test coverage, you can use coverage:

pip install coverage
coverage run --source aas --branch -m unittest
coverage report -m

Contribute Code/Patches

TBD