Overview

This repository contains the application of the OPC UA export and lifting scripts at the Foundry Institute at RWTH Aachen University. Our goal is to capture all semantic information (e.g. unit, range, accuracy, etc.) from all sensor nodes using a suitable XPath expression.

To achieve we use the results of the project Transforming-XPath-to-OPC-UA. The basic idea behind their project is to simplify data access in OPC UA by transforming XPath expressions to OPC UA. To achieve this they:

• created a local export of the OPC UA server
• created an XML image based on the export
• created XPath expressions
• transformed these XPath expressions to OPC UA to extract data

Since we cannot publish the OPC UA server data of the machines we use, we utilize the Prosys sample server for demonstration.

Requirements

• python 3.7
• opcua-asyncio Version 0.9.0 (e.g., run pip install asyncua==0.9.0)
• OPC UA Server/Prosys Example Server

Usage

The program consists of the following four scripts:

• extractor_FI.py
• fileAnalysis.py
• OPCtranslator.py
• initializer.py
• RESTexporter.py

The three folders additionalFiles, exports and semantics are created automatically if they do not exist, but the files such as the certificates for the servers must be added manually in the additionalFiles folder. To execute the scripts the file initializer.py must be executed. The machines to be extracted can be defined as arguments in the command line call to the initializer.py file. (open the command line in the directory where the files are located, at least one machine name must be specified) For example:

• If we want to run the scripts on all machines

  python .\initializer.py InjectionMolding DosingFurnace RealTimeMS HPDC SprayHead RetrofitCS

• If we want to run the scripts only on the Dosing Furnace and Spray Head

  python .\initializer.py DosingFurnace SprayHead

At the FI Institute the execution of the initializer.py is done automatically via a cronjob with the command:

30 1 * * 0 /usr/bin/python3.7 initializer.py InjectionMolding DosingFurnace RealTimeMS HPDC SprayHead RetrofitCS

The cronjob is executed every Sunday at 01:30.

The extracted and processed json files are made available via a REST API server. For this we execute the file RESTexporter.py from the command line. For example:

• In general

  python .\RESTexporter.py

• At the FI Institute

  /usr/bin/python3.7 RESTexporter.py

The data of the individual machines can be accessed via the url in the form http://ip:port/machinename:

Machine	url
Injection Molding Unit	http://127.0.0.1:8080/InjectionMolding
Dosing Furnace	http://127.0.0.1:8080/DosingFurnace
Real Time Measurement System	http://127.0.0.1:8080/RealTimeMS
Spray Head	http://127.0.0.1:8080/SprayHead
HPDC Machine	http://127.0.0.1:8080/HPDC
Retrofit Cell Sensors	http://127.0.0.1:8080/RetrofitCS
Prosys Server	http://127.0.0.1:8080/ProsysExServer

Specs

The scripts run at FI Institute on a Raspberry Pi 4 with the following specifications:

• OS: Raspbian Buster x64 (Debian 10)
• CPU: Broadcom BCM2711, Quad core Cortex-A72 (ARM v8) 64-bit SoC @ 1.5GHz
• RAM: 8 GB
• Directly connected to the machines via Ethernet

The computer on which we tested the scripts has the following specifications:

• OS: Windows 10 Education 64 bit
• CPU: Intel Core i5-4210U @ 1.70GHz - 2.4GHz
• RAM: 16 GB

Requirements to connect to the machines

	nothing	username	password	certifiate	private key
Injection Molding Unit		X	X
Dosing Furnace		X	X
Real Time Measurement System	X
Spray Head				X	(X)
HPDC Machine		X	X	X	X
Retrofit Cell Sensors		X	X
Prosys Server	X

We use the same certificate for both the Spray Head and the HPDC Machine. The certificate is created as described here. The certificate and the private key are located in the additionalFiles folder. In general, the Spray Head does not require a private key, but since the set_security_string method of FreeOpcUa requires one, we need to specify it.

additionalFiles

This folder contains the necessary certificate and private key for exporting from the respective machines as well as the UNECE_to_OPCUA.csv file to translate the extracted UnitIds to a human readable format.

exports

This folder contains the exports and calculated XML images from the different machines. To get the desired information we search the server for all nodes with the data type "EUInformation". Then we get the corresponding sensor node via the "parent" expression and get all children resp. sensor metadata from it via the "child" expression. The results of this are also located as .json files in this folder.

Since the HPDC Machine does not use datatypes like EUInformation and EURange, no results are found. If we search the server instead for nodeids ending with ...min or ...max we get 14 sensor nodes.

	runtime of export creation on Raspberri Pi [sec]	number of extracted nodes	runtime of xml image calculation [sec]	number of nodes in xml image	number of cycles detected while calculation the xml image	runtime of semantic expression on Raspberry Pi[sec]	number of sensor nodes in semantic expression
Injection Molding Unit	121	4,161	11	4,016	124	12	25
Dosing Furnace	105	2,035	3	1,363	2	1	0
Real Time Measurement Systems	52	1,940	2	1,357	0	2	191
Spray Head	144	1,875	2	1,414	0	1	0
HPDC Machine	88	2,190	3	1,902	1	1	167
Retrofit Cell Sensors	1,142	35,257	563	34,643	0	5,645	135
Prosys Server	118	4,331	7	3,064	0	1	21

semantics

This folder contains the results of the semantic expression. The results are saved as a json file in the format 'Machinename'Analyzed.json. For example:

• InjectionMoldingAnalyzed.json
• RetrofitCSAnalyzed.json