Skip to content
OPC UA made easy - main repository of the Unified Architecture Framework
C++ Python CMake C Batchfile Makefile
Branch: master
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.
build restoring build folder by adding a .gitignore file in the build folde… Aug 27, 2014
doc added gitignore files to html output dirs May 15, 2013
examples Fixed missing link library. Nov 1, 2019
lib Add all code May 14, 2013
src Enabled c++11. Nov 12, 2019
unittests Merge pull request #124 from WimPessemier/master Jul 20, 2017
.gitignore Merge branch 'remove_uafc' May 11, 2015 support for security was added since v2.0.0 Nov 19, 2015
changelog.rst.txt Update to unified automation sdk v1.7.1. Oct 24, 2019
dependencies.rst.txt added some info about which SDK is needed Jul 30, 2014
install_linux.rst.txt updated the installation procedures to inform about the SDK library c… Jun 3, 2015
install_windows.rst.txt updated the installation procedures to inform about the SDK library c… Jun 3, 2015
lgpl-license-3.0.txt Add all code May 14, 2013
version.txt Update to unified automation sdk v1.7.1. Oct 24, 2019


UAF - the Unified Architecture Framework

Project website?

  • Main code repository:

  • Git Branches:

    • master: tested and production-ready, has the latest API and the latest tested features.
    • long-term: tested and production-ready, but API and feature set are frozen. Only bugfixes are applied.


  • The Unified Architecture Framework (UAF) is an open source software framework to simplify the development of OPC UA applications on Linux and Microsoft Windows, in C++ and Python.

  • OPC UA (OPC Unified Architecture) is the next generation standard for secure, reliable and scalable industrial communication.

  • In short, the UAF / pyUAF is OPC UA made easy!


  • The intent of (Py)UAF is to facilitate a web-browser-like experience for the plant floor. The idea is that you can browse the machinery of a whole facility (or even multiple facilities) by following the links ("references") between the data points ("nodes") of a plant-wide live model ("address space"). Similar to a web-browsing experience, it shouldn't matter to the client whether these data points are hosted by a single server, or spread over multiple servers. Also similar to a web-browser, none of the servers have to be known a priori by the client. New servers (e.g. new machines, redundant servers that come online, ...) can be discovered automatically by the UAF, and their data points can be browsed/read/written/monitored/... just like any other server to that was already known to the client.

  • In practice, the UAF is a framework (or "toolkit") that helps you to create advanced OPC UA applications with minimal effort. Essentially, it takes care of some technical aspects (such as Session and Subscription management, server discovery, address resolution, ...) so that you can concentrate on the functional aspects of your software application. Of course, some technical aspects (such as Session and Subscription management) are also exposed by the API, so that you can keep full control over them if you don't want the UAF to take care of them.

  • Some very concise and easy to understand examples can be found here:

  • A lengthier tutorial can be found here:

  • The following example (a single example, split up into several pieces) shows you some advantages of the UAF, in Python. In C++, the equivalent code would be a bit more verbose, but the C++ API is essentially the same as the Python API.

    • Intuitive API:

      All code from the Software Developers Kit is wrapped into some convenient namespaces and classes (many more than can be seen below of course).

      import pyuaf
      from pyuaf.util            import Node, NodeId, QualifiedName
      from pyuaf.util            import RelativePathElement, LocalizedText
      from pyuaf.util.primitives import UInt16
      from pyuaf.client          import Client
      from pyuaf.client.settings import ClientSettings
    • Automatic discovery:

      Just provide one or more Discovery URL(s), and the UAF will then automatically and periodically perform the discovery for you.

      settings = ClientSettings()
      settings.applicationName = "myClient"
      # Create the client with the given settings, and it will start to discover
      # the servers in your network:
      myClient = Client(settings)
    • Easy node addressing

      You can address nodes

      • in an absolute way (via "NodeIds" and "ExpandedNodeIds")
      • or in a relative way (via "RelativePaths").

      The UAF will resolve the addresses of the nodes, regardless on what server the nodes are hosted.

      # first define some URIs (Uniform Resource Identifiers):
      ns      = "MyOrganization"      # the "namespace URI" of the organization or company
      plc12   = "PLC/12/Server/URI"   # the "server URI" of the OPC UA server running on some PLC
      # now define some absolute addresses:
      tank5      = Address( NodeId("Tank5"    , ns), plc12) )
      tank5_name = Address( NodeId("Tank5Name", ns), plc12) )
      # now define some relative addresses (Tank5/Sensor3 and Tank5/Sensor3/Status):
      tank5_sensor3        = Address( tank5, [ RelativePathElement(QualifiedName("Sensor3", ns)) ] )
      tank5_sensor3_status = Address( tank5, [ RelativePathElement(QualifiedName("Sensor3", ns)),
                                               RelativePathElement(QualifiedName("Status" , ns)) ] )
      # the UAF accepts even relative addresses to relative addresses! So you could also do:
      tank5_sensor3_status = Address( tank5_sensor3, [ RelativePathElement(QualifiedName("Status", ns)) ] )
    • Automatic address resolution

      You can now read/write/monitor/... the Value (or any other attribute) of the nodes that we addressed above. Even if these nodes are hosted by multiple servers, the UAF will automatically:

      • create the necessary Sessions and Subscriptions to the correct servers (or re-use existing ones if possible)
      • resolve the addresses (i.e. translate the relative paths to absolute ExpandedNodeIds)
      • group the targets of the request into optimal service calls (so the UAF ideally only has to perform one service call per server)
      • perform the necessary service calls (i.e. invoke the actual OPC UA Read/Write/... service)
      • join and re-order the results of the service calls (so that the user doesn't notice that the call was actually split up into multiple calls).

      This can be seen in the following lines: reading or writing just takes a single line of code!

      # let's read 
      #   - the tank name         (a LocalizedText, which may be exposed by an OPC UA-enabled PLC) 
      #   - and the sensor status (an UInt16, which may be exposed by some OPC UA-enabled smart sensor)
      result =[tank5_name, tank5_sensor3_status])
      if isinstance(result.targets[0].data, LocalizedText):
          name   = result.targets[0].data.text()
      if isinstance(result.targets[1].data, UInt16):
          status = result.targets[1].data.value
      # we can also write a new name
      result = myClient.write( [tank5_name], [LocalizedText("Oil tank", "EN")] )
      if result.targets[0].status.isGood():
          print("OK, the new name was written successfully!")
    • Persistent monitored items

      You can create monitored items once, and then forget about them...

      • even if the server that hosts your monitored items is not online yet!
      • even if the server that hosts your monitored items dies, and the relative addresses of your monitored items now suddenly point to nodes hosted by another (redundant) server!!
      def myCallback(notification):
          print("New sensor status received: %d"
      myClient.createMonitoredData([tank5_sensor3_status], notificationCallbacks = [myCallback])
    • More stuff

      • The UAF client-side encourages you to define your node addresses in terms of Namespace URIs and Server URIs instead of the corresponding IDs. URIs are well-defined and static, while IDs may change between server restarts or server failovers. Since the UAF takes care of the mapping between URIs and IDs for you, you never have to deal with IDs, which means that your code remains stable, and your applications are robust to server restarts or failovers.

      • The UAF client-side keeps track of the mapping between temporary client-assigned handles and temporary server-assigned IDs. Since server-assigned IDs may change during the lifetime of your client application (in contrast to client-assigned handles, which are assigned just once by the UAF), client applications based on the UAF are robust to server restarts or failovers without effort.

      • The UAF client-side can automatically perform follow-up requests if a single request wasn't sufficient to get all data, and join the partial results into one complete result. This can happen in case of browsing large address spaces, or fetching a large amount of historical data. So you can browse a whole server or fetch a long list of historical data values with a single line of code, without having to worry about the need for possible follow-up requests.

      • The UAF Client class has a number of convenience methods (read/write/startMonitoringData/...) but also a generic processRequest method that can process fully configurable ReadRequests, WriteRequests, MethodCallRequests, ... It means that all configurable service parameters from the OPC UA SDK are exposed by (Py)UAF as well.

      • Dive into the documentation or the examples to find out more!


  • The UAF is based on the commercial C++ OPC UA Software Developers Kit from Unified Automation. A demo version of this SDK can be downloaded from their website for free: For the client-side of the UAF (the only side currently implemented), only the Unified Automation Client SDK is needed. The same will apply for the server-side of the UAF, once it is implemented.

  • More info about the dependencies: see dependencies.rst.txt



  • The Python documentation is online here.

  • The C++ documentation you can easily generate yourself by using Doxygen. Use the Doxyfile in the src directory.


Currently only the OPC UA client-side is implemented. Many services are fully implemented (see list below) but some less-frequently used services are still pending.

  • Supported services on the Client side:
    • read (synchronous + asynchronous)
    • write (synchronous + asynchronous)
    • method call (synchronous + asynchronous)
    • translate browse paths (synchronous)
    • create monitored data items (synchronous)
    • create monitored events items (synchronous)
    • browse and browse next (synchronous)
    • read historical data (raw data + modifications) (synchronous)
    • set publishing mode (synchronous)
    • set monitoring mode (synchronous)
  • Supported services on the Server side:
    • nothing yet so far!



  • Author: Wim Pessemier

  • Contact: W**.P******** (replace the asterisks)

  • Organization: Institute of Astronomy, KU Leuven (Belgium)

  • Project website:


  • This program is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

    This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

    You should have received a copy of the GNU Lesser General Public License along with this program. If not, see


Project Stats

You can’t perform that action at this time.