Single Source of Truth: Integrated Process Control and Data Acquisition

System for the Development of Resistance Welding of CFRP Parts

Michael Vistein

, Monika Mayer

, Manuel Endraß

and Frederic Fischer

German Aerospace Center, Center for Lightweight Production Technology, Augsburg, Germany

{michael.vistein, monika.mayer, manuel.endrass, frederic.ﬁscher}@dlr.de

Keywords:

Thermoplastic Composites, Resistance Welding, Process Control, Data Acquisition, SCADA.

Abstract:

For the development of a novel (industrial) process, in particular within a research environment, a very ﬂexible

and adjustable control and data acquisition system is required. Traditional SCADA systems often are not

designed for frequent changes. To facilitate the development process, the storage of all relevant data from

process parameters to measurement data at a single location, as a “single-source-of-truth” is desirable. This

paper introduces an integrated process control and data acquisition system that is built around the open-source

central data storage system “shepard” which facilitates the evaluation of the process and offers potential for

inline-optimization. The system is evaluated by the example of the production of a thermoplastic component.

As part of the European Clean Sky II large passenger aircraft project, the German Aerospace Center produces

the 8-meter long upper shell of the multifunctional fuselage demonstrator (MFFD) made from thermoplastic

composites. The frames are attached to the skin by resistance welding, which is done using an actuated gantry

system. This novel process has the potential to disrupt standard aircraft assembly by dust-less welding in a

fully automated yet interactively customizable manner which is hence relevant for every process development

context.

1 INTRODUCTION

The development of a novel production process usu-

ally requires a large number of experiments which

need to be documented and later evaluated carefully.

During the evolution, different sets of process param-

eters may be required and a large amount of measure-

ment data can be aggregated. To facilitate this pro-

cess, the German Aerospace Center (DLR) has cre-

ated the shepard software (storage for heterogeneous

product and research data)(Haase et al., 2021; Krebs

et al., 2021). Using a REST (Representational State

Transfer) interface, data can be stored and retrieved

from shepard in a structured way. Connections be-

tween different data items can be created, both in a

semantic way (e.g. as predecessor and successor rela-

tion) and in a temporal way. A key concept of shepard

is the ability to store a vast amount of different data

at a single location, therefore employing a “single-

source-of-truth” approach. Furthermore, shepard is

developed as an open-source project for maximum

https://orcid.org/0000-0001-6998-0017

https://orcid.org/0000-0002-4448-9501

https://orcid.org/0000-0003-4133-0834

https://orcid.org/0000-0001-5421-4336

reach and features an on-premises approach to store

sensitive data locally, which can be vital in research

projects.

In industrial applications, SCADA (Supervisory

Control and Data Acquisition) systems are very com-

mon. Using these systems, the operator can control

and monitor the production process and modify rel-

evant process parameters (often called recipes). Ag-

gregated data can also be provided to a MES (man-

ufacturing execution system) for further processing.

Common industrial solutions are for example WinCC

(Siemens) or Aveva. For use in research applica-

tions, these solutions have some drawbacks. They

usually are tailored to a speciﬁc process, which does

not change much after being commissioned, whereas

in research the development of a process often is a pri-

mary concern. Furthermore, licenses can be very cost

intensive. There have been some approaches creating

low-cost and/or open-source SCADA systems, both

for industrial (e.g. (Phuyal et al., 2020), (Merch

et al., 2017)) and academic use cases (e.g. (Vargas-

Salgado et al., 2019), (S¸ahin and

Is¸ler, 2013)). Sev-

eral of these systems are build around an open ar-

chitecture using the OPC/UA (Open Platform Com-

munications/Uniﬁed Architecture, (OPC Foundation,

592

Vistein, M., Mayer, M., Endraß, M. and Fischer, F.

Single Source of Truth: Integrated Process Control and Data Acquisition System for the Development of Resistance Welding of CFRP Parts.

DOI: 10.5220/0012161500003543

In Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2023) - Volume 1, pages 592-599

ISBN: 978-989-758-670-5; ISSN: 2184-2809

2022)) protocol for a vendor-independent communi-

cation between different components and the control

system.

This paper introduces a ﬂexible, open-source

SCADA system called ProcessControl together

with a real-time monitoring system called Process-

Monitoring that are based on top of the shepard data

storage system. This allows to beneﬁt from the advan-

tages of shepard, in particular from the single-source-

of-truth approach that avoids scattering process pa-

rameters, measurement logs and documentation over

several places. A large number of experiments can be

tracked exactly with the parameter sets that have been

used in combination with measurement data, which

allows the process engineer to improve the process

continuously. This paper describes the SCADA sys-

tem using the example of a resistance welding ap-

plication within the Clean Sky II project “Multifunc-

tional Fuselage Demonstrator” (MFFD).

The remainder of this paper is organized as fol-

lows: Section 2 provides an overview of the resistance

welding use-case. In section 3 the SCADA system is

explained in detail. Section 4 evaluates the beneﬁts of

the process control system from the process engineers

point of view. Finally, section 5 draws a conclusion

and provides an outlook of further improvements.

2 MULTI FUNCTIONAL

FUSELAGE DEMONSTRATOR

The DLR, together with Premium Aerotec, Airbus

and Aernnova are part of the Clean Sky II project

Multifunctional Fuselage Demonstrator and respon-

sible for single part manufacturing and the assem-

bly of the 8-meter long upper shell structure, made

from carbon ﬁber-reinforced low-melt Polyarylether-

ketone (CF/LM-PAEK). The single part manufactur-

ing and assembly combines several innovative pro-

duction technologies such as in-situ consolidation

with automated ﬁber placement (AFP

ISC

), ultrasonic

welding (spot and continuous) and electrical resis-

tance welding(Fischer et al., 2022; Endraß et al.,

2022a). To validate all processes, a full-scale, but in

length to 1 meter reduced test shell has been manu-

factured on track towards the upper shell manufac-

turing as part of the de-risking strategy for technol-

ogy validation. Because the whole structure is made

from CF/LM-PAEK, thermoplastic welding technolo-

gies are key enablers for dustless assembly. Dur-

ing assembly, the shell is ﬁrst stiffened by stringers

using continuous ultrasonic welding. Next step is

the integration of frames to the skin by resistance

welding. The connecting surfaces of the frames are

Figure 1: Custom-made weld modules for resistance weld-

ing of frames to skin (left). Resistance welded composite

frame within the test shell (right).

Figure 2: Weld bridge anchored to the placement tooling.

called “attached ﬂanges” and are directly welded to

the skin laminate using resistance welding. The in-

tegral frames in the upper shell are split in the keel

region, in order to enable the integration. There-

fore, frame-couplings are welded to the frame ends

to connect both sides. Finally, right-angled elements

(cleats) are welded to certain stringer and frame inter-

sections to further stiffen the structure. In this section,

the manufacturing hardware and the resistance weld-

ing process for frame integration, which is focused in

this paper, are explained.

2.1 Overview

In order to validate the custom-made welding infras-

tructure, processes, data acquisition and the manufac-

turing execution system six frames have been welded

to the test shell. Each frame 1 (cf. ﬁgure 1) is con-

nected to the skin laminate 2 by up to 18 welded

attached ﬂanges 3 using a unique weld-module 4

which provides both the necessary welding current,

as well as mechanical pressure to join both composite

parts.

Due to a reduced stringer pitch towards the fuse-

Single Source of Truth: Integrated Process Control and Data Acquisition System for the Development of Resistance Welding of CFRP Parts

593

lages horizontal split, different length conﬁgurations

of the attached ﬂanges have to be encountered and

thus require individual process parameters for the

welding process (i.e. current, voltage and time) with

the necessity to be even optimized and adopted in

value and sequence during test shell manufacturing.

The welding modules (cf. ﬁgure 1 left) are attached

to a one degree-of-freedom gantry system. This so-

called weld bridge, cf. ﬁgure 2 can be positioned

accurately using a rack and pinion system. Every

frame is aligned into a ﬁxture on the weld bridge and

subsequently positioned on the accurate location on

the skin in ﬂight-direction. Afterwards, the attached

ﬂanges are welded to the skin one by one.

2.2 Resistance Welding for Shell

Assembly

Resistance welding is, besides induction and ultra-

sonic welding, known as one of the most matured

welding technologies applicable for joining of high-

performance thermoplastic composites. In resistance

welding a welding strip, the so-called welding ele-

ment, is placed in the bondline. The welding element

consists out of a conductive implant (e.g. stainless-

steel mesh or carbon ﬁbers) and in case of electri-

cally conductive adherents (the skin and the attached

ﬂanges) additional insulation (e.g. glass-ﬁber or neat

resin layer) above and below the implant. While

maintaining the adherents under weld pressure, an

electrical current ﬂow through the conductive implant

leads to joule heating, melting of the bondline while

increasing the polymer chain mobility under reduced

viscosity and allows for polymer diffusion between

welding element and adherents(Ageorges et al., 2000;

Yousefpour et al., 2004). Since heat is generated in

between the adherents the introduced power can be

balanced to the polymer needs, and a structural join is

generated under minimal invasive component defor-

mation. Processing parameters for resistance weld-

ing of LM-PAEK were matured on coupon level at

DLR’s static resistance welding test bench, based on

a sophisticated Design of Experiments (DoE) study.

Numerous welds were performed validating attenu-

ation losses during water coupled ultrasonic testing,

fracture load in single lap shear testing and failure

mode by fracture surfaces analyses. For detailed in-

formation the reader is referred to our publications

(Endraß et al., 2020; Endraß et al., 2022b; Thom

2021). However, process parameters need to be

adopted at transfer from coupon to full scale level

due to changes in boundary conditions, e.g. differ-

ent thermal conductivity of components and tools and

transition resistance. Thus, ﬂexible process parame-

Motors, Valves,

Sensors

TwinCAT PLC

ProcessControl

ProcessMonitoring

Shepard

Measurement

equipment

EtherCAT

OPC/UA

REST (HTTPS)

Figure 3: Overview of system architecture.

ter modiﬁcation is essential during this early develop-

ment phase. Within the MFFD’s upper shell manu-

facturing frames, frame-couplings and cleats are inte-

grated by electrical resistance welding. In total 652

resistance welding operations are performed varying

in geometrical dimensions of part and bondline, weld

tool conﬁgurations and positions within the structure.

Hence, the allocation of the deﬁned weld parameter

sets to its respective operation positions and traceable

data management is indispensable for the automated

assembly and data evaluation.

3 PROCESS CONTROL AND

DATA ACQUISITION

The overall system for frame welding of the MFFD

test shell consisted of several components which are

coupled using standard networking protocols. The

project employs three pieces of software that have

been developed at DLR: shepard, ProcessControl and

ProcessMonitoring. In the following sections, at ﬁrst

the overall architecture is introduced, followed by a

more in-detail description of the main components.

3.1 Architecture

Figure 3 shows the overall architecture of the welding

system. The hardware of the weld bridge is controlled

using a Beckhoff TwinCAT programmable logic con-

troller (PLC) running on a standard industrial PC.

The motor controllers, values, etc. are connected to

the PLC using the industrial ﬁeldbus protocol Ether-

CAT(Jansen and B

uttner, 2004). The gantry is actu-

ated by two Beckhoff AM-series motors, driven by an

AX8000 servo system. The pneumatic system is con-

trolled using Festo CPX-series valves, and the weld-

ing power supply by analog 0 V to 10 V signals.

For quality assurance purposes, measuring equip-

ment is installed. This consists of several high-

precision analog input terminals (Beckhoff ELM se-

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

594

ries) that can measure the voltage and current ap-

plied to the welding elements, as well as temperatures

from thermocouples. Data for the pressure measur-

and is collected via OPC/UA. The measurements are

taken by the specialized ProcessMonitoring software,

which collects data from different sources, performs

a live visualization of the process and stores the mea-

surement data in the central shepard database.

The overall process is controlled by the Process-

Control software. This software communicates with

the PLC using the OPC/UA protocol and uses the

shepard database as central data store.

3.2 Shepard Data Storage

During the execution of a manufacturing process a

large variety of process data accrues, often from a

multitude of different sensors. The acquisition and

structured storage is essential for quality assurance

steps, as well as for inline process optimization. A

precise annotation of time and location for all data is

essential for later analysis, as well as a concise knowl-

edge of all process parameters that have been used. In

order to avoid redundancies, a single central storage

(“single source of truth”) is desirable.

To facilitate the data storage, the DLR has devel-

oped the shepard software(Haase et al., 2021; Krebs

et al., 2021). Shepard can be accessed using a REST-

based API and provides ready-to-use client imple-

mentations for a wide variety of programming lan-

guages (e.g. Java, C++, Python, . . . ). The recently de-

veloped shepard ecosystem has been selected as base

technology to evaluate the usability and suitability of

the system within a larger research use-case.

The shepard system data model consists of two en-

tities: Collections and DataObjects. Collections ag-

gregate DataObjects and serve as a container for a sin-

gle project, workpiece, etc. A DataObject represents

a single step, action, result, etc. within a project. A

hierarchical structure can be formed among DataOb-

jects by using a parent/child relation (every Data-

Object can have exactly one parent and an arbitrary

number of children). Furthermore, a temporal rela-

tion between two DataObjects can be modeled using

the successor relationship. DataObjects can contain

references to further data such as StructuredData (ar-

bitrary data in JSON (JavaScript Object Notation) for-

mat), Timeseries (temporal sequence of data points)

or Files (binary ﬁles).

3.3 Process Control

The overall process is controlled by an additional

in-house developed open-source software called

ProcessControl, which acts as SCADA system. A ma-

jor goal during the development of the ProcessControl

software was the concept of a “single source of truth”.

To achieve this goal, ProcessControl stores all rele-

vant data within shepard.

A main concept of ProcessControl is the recursive

decomposition of the production process into smaller

parts. Ideally, several small parts share common pa-

rameter sets and programs and thus allow for reuse.

The frame welding process for the MFFD demonstra-

tor can be decomposed into two layers: frames and

attached ﬂanges. The process parameter set for the

frames contain the position of the frame, such that the

associated program can move the weld bridge into the

proper position. The process parameter set for the at-

tached ﬂanges contains all necessary information for

the welding process (such as voltage, current, dura-

tion and pressure).

Figure 4 shows the overall structure of data that

is stored for the demonstrator part. ProcessControl

makes use of the shepard DataObjects and Structure-

Data references. On the top level, a single DataObject

for the project (i.e. in this case frame welding) is used.

Different projects can be stored in the same or in mul-

tiple shepard Collections.

The data structure for a project consists of two

parts. The ﬁrst part (left, dashed part of ﬁgure 4) is

the speciﬁcation of relevant metadata and is attached

to the main ObjectNode as StructuredData. This con-

tains the deﬁnition of parameter sets (i.e. names and

types of data variables, but no concrete values) and

the manufacturing programs necessary for produc-

tion. For one production step, a single set of process

parameters must be deﬁned, but several programs can

be speciﬁed, which can be useful for multi-stage pro-

cesses. In the case of the MFFD demonstrator, for

each ﬂange of a frame at ﬁrst a contact measurement

is performed before the ﬁrst actual welding is per-

formed.

The second part (right, solid part of ﬁgure 4) de-

ﬁnes the structure of the part (i.e. the number of

frames and the number of ﬂanges for each frame).

Furthermore, at least one instance of (concrete) pro-

cess parameters must be present for every step. More

than one instance may be present if a step is tried sev-

eral times with varying parameters in an experimen-

tal setup. To achieve traceability, no process parame-

ters are ever overwritten, but rather a new instance is

added.

The initial setup of the project structure and pro-

cess parameters is done automatically by the Process-

Control software using a CSV ﬁle (comma sepa-

rated values) and by supplying an appropriate JSON-

formatted project deﬁnition ﬁle. The project deﬁni-

Single Source of Truth: Integrated Process Control and Data Acquisition System for the Development of Resistance Welding of CFRP Parts

595

Upper shell

demonstrator

Frame

integration

Project

definitions

Frame

Process

Parameter

definition

Program stage 1

Program stage

…

Flange

Process

Parameter

definition

Program

Frame 1 (using

frame

parameters)

Parameters 1

(e.g. bridge position)

Parameters …

Flange 1 (using

flange

parameters)

Parameters 1

(e.g. voltage, current)

Parameters …

Execution 1

Execution …

Flange …

Frame 2 Frame …

DataObject

StructuredData

Collection

Reference to

parameter set

used in trial

Figure 4: Structure of the MFFD demonstrator project.

tion ﬁle contains all programs and requires detailed

knowledge about the underlying hardware systems

and should be written by an automation expert. The

CSV ﬁle contains only the process parameters and

should be written by a process expert.

The ProcessControl software has been designed as

a manufacturing execution system with a strong fo-

cus on supporting experimental production scenarios.

Therefore, a graphical user interface has been created

which allows the user to modify any process param-

eter during the experiment, and also to execute a sin-

gle step and stage multiple times with different pa-

rameters (provided that the underlying process allows

multiple executions). This interface is displayed in

ﬁgure 5.

Every time a step is executed, a new DataObject

node is created in shepard and metadata such as ex-

ecution time and the concrete parameters that have

been used are attached to the new node. In addition,

a free text comment can be stored, enabling the ex-

pert to comment live on the experiment, allowing for

later precise evaluation. Furthermore, this node is also

used by the ProcessMonitoring software described in

section 3.4 to store measurement data. By combining

all relevant data at a single location, reﬁnement of the

process or experiment can be facilitated.

The ProcessControl software core is agnostic to

the process and the underlying hardware. Exten-

sion modules provide hardware connectivity. For the

Figure 5: Process parameter editor, showing parameters for

two phases of the welding process: CM (contact measure-

ment) and IP (initial voltage pulse). For each phase the time

(t), voltage (U), predicted current (IP), maximum allowed

current (IM) and applied pressure (p) can be speciﬁed.

MFFD demonstrator, communication using OPC/UA

has been implemented. The TwinCAT PLC provides

an OPC/UA server, while the communication module

in ProcessControl acts as OPC/UA client. The com-

munication module provides basic operations such as

writing a value to a node or waiting for a node to

change value. Using these operations, process param-

eters can be transmitted to the PLC and a start ﬂag

can be set. Further execution can be halted until a

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

596

Figure 6: Example for live GUI of ProcessMonitoring. Top

left: voltage, top right: pressure, bottom left: current, bot-

tom right: temperature.

success or error ﬂag is returned by the PLC. In the

case of an error, the whole process can be interrupted.

ProcessControl orchestrates the process and its data

ﬂow, hence the main program logic is expected to be

implemented in the underlying PLC.

3.4 Data Acquisition and Monitoring

Data acquisition and monitoring of relevant measur-

ands is an important part of a SCADA system. In

an industrial production process, this can be highly

customized with no need for changes after commis-

sioning. In contrast, an experimental setup or process

development demands a ﬂexible solution to adapt to

new requirements by means of new data sources and

measurands providing other insights into the process.

ProcessMonitoring offers interfaces to handle in-

coming time series from different data sources and

to add consistent timestamps. Important metrics for

the quality of welds are voltage, current, temperature

and pressure in the weld area that have been applied

over time (cf. ﬁgure 6). Two different data sources

are currently in use: OPC/UA for pressure values and

EtherCAT for voltage, current and temperature. Mea-

surements from a National Instruments card can also

be acquired simultaneously, but this was not needed

in this use case.

The pressure values are retrieved by the Twin-

CAT PLC from the pneumatic components. Process-

Monitoring receives the values by using the open-

source open62541 library for implementing the

OPC/UA connection. The OPC/UA communication

channel between the TwinCAT PLC and Process-

Monitoring enables (near) real-time communication,

reliably achieving latencies of 10 ms or less.

Current and voltage are measured by high-

precision analog input terminals (type Beckhoff

ELM3704-0000). The temperature between skin and

attached ﬂange is recorded by thermocouples con-

nected to Beckhoff EL3314-0002 terminals.

Figure 7: Resulting graph of welding operation generated

by ProcessMonitoring.

ProcessMonitoring communicates directly with

the Beckhoff measurement equipment by using an

EtherCAT master stack based on the Simple Open

EtherCAT Master (SOEM)(OpenEtherCAT Society,

2019) project. A temporal resolution of 1 ms for the

EtherCAT communication channel has been achieved

and is sufﬁcient. Therefore, a cycle loop below 1 ms is

crucial and can be accomplished using the Win32 API

function “QueryPerformanceCounter” submitting a

resolution of 1 µs on a Microsoft Windows 10 PC. So

EtherCAT frames can be handled and measured val-

ues can be extracted within 1 ms.

During the welding process the values for each

measurand are buffered and displayed live in the

graphical user interface (GUI, see ﬁgure 6). The GUI

has been developed using the Qt toolkit.

After the last welding phase has been ﬁnished,

ProcessMonitoring saves all relevant data in CSV ﬁles

and creates a graphics ﬁle of the measured curves (cf.

ﬁgure 7). All ﬁles are attached to the corresponding

ObjectNode in shepard that is also storing the process

parameters and comments of the expert.

4 EVALUATION

ProcessControl and ProcessMonitoring have been

used heavily during the manufacturing of the test

shell. Since the test shell manufacturing was imple-

mented as part of the de-risking strategy towards the

manufacturing of the MFFD’s 8m full-scale demon-

strator, processing parameters for resistance welding

and the weld setup were modiﬁed and optimized sev-

eral times with respect to reliable processing. An ex-

tensive Design of Experiments (DoE) -based paramet-

ric study targeting the deﬁnition of processing param-

eters (voltage, current, time and pressure) for resis-

tance welding of LM-PAEK within the DLR’s test

bench was conducted in order to deﬁne a baseline

set of parameters for later frame integration(Endraß

et al., 2022b). The optimized set of parameters were

set as baseline processing parameters within Process-

Single Source of Truth: Integrated Process Control and Data Acquisition System for the Development of Resistance Welding of CFRP Parts

597

Control and tuned, varying single parameters during

test shell manufacturing within the ﬁrst 90 weld op-

erations (ﬁve of six frames). The very last 18 weld

operations on frame six, were conducted using the

compiled and optimized parameter set. Since param-

eter variations from the baseline automatically gen-

erate a new tree for the parameter set within Process-

Control linked to the later execution ﬁle, transparency

and traceability of implemented changes is achieved.

This approach facilitates streamlined data retrieval

and subsequent analysis, accelerating the research

process and fostering reproducibility.

On the other hand, ProcessMonitoring allowed for

(near) real-time visualization of the actual process cy-

cle and played a pivotal role in the inline interpreta-

tion of the experimental data especially for the pro-

cess expert.

Using the MFFD frame-welding use-case, it could

be demonstrated that the shepard ecosystem is suit-

able for use in a large research project. With both

shepard

and ProcessControl

being open-source ap-

plications, a ﬂexible and cost-effective way of creat-

ing a SCADA system in research is provided.

5 CONCLUSION AND OUTLOOK

Using the tools presented in this paper, it was pos-

sible to create a control architecture that allowed the

process engineer to concentrate on conducting the ex-

periments. The tools were made with the research

use-case in mind and allowed the process to be ad-

justed ﬂexibly. Standard hardware components could

be used and integrated with minimal additional cost.

The shepard ecosystem has been selected as main

data storage for all process steps in the ﬁnal assem-

bly of the MFFD upper shell demonstrator, not just

for frame welding. Main reasons for the development

and use of shepard were the on-premises storage ca-

pabilities, its ability to adjust ﬂexibly to many differ-

ent types of data, the possibility to connect different

data to each other, and also the availability as open

source software. ProcessControl is the ﬁrst SCADA

software based on top of shepard.

The traceability of all experimental data, begin-

ning with the initial process parameters up to the ﬁnal

measurement results has been greatly increased by us-

ing a central data storage. The possibility to modify

process parameters and to annotate the experiments

during the initial setup of the experiment has proven

to be very helpful during the manufacturing of the test

https://doi.org/10.5281/zenodo.5091603

https://doi.org/10.5281/zenodo.8262579

shell. By storing all relevant information at a single

location, the time necessary to evaluate the experi-

ments can be reduced.

Based on the experiments and the data that has

been recorded during the manufacturing of the test

shell, initial parameters for the production of the ﬁnal

demonstrator could be created by the process expert.

To increase the reliability of the welding process, tar-

get envelopes for current and voltages have been de-

ﬁned. A recent extension of ProcessMonitoring pro-

cesses these target envelopes (stored as parameters

within shepard) and checks the measured live values

against the envelope. If a deviation occurs, a signal is

transmitted to the PLC which can react and, if neces-

sary, immediately abort the welding process.

The beneﬁts of centralized data storage in com-

bination with ﬂexible and vendor-independent tools

can be useful in many scenarios for process develop-

ment and research. Currently extensions are being

developed which allow to use the tools in projects

without a central PLC, for example with industrial

robots. Some manufacturers, e.g. KUKA, provide an

OPC/UA interface to their controllers which allow to

upload robot programs on-the-ﬂy. By using this inter-

face, ProcessControl could not only trigger the execu-

tion of programs (as it is usually done by the SCADA

system), but also to upload robot programs at the time

they are required. This allows to store the reference

version of programs together with all other relevant

information in shepard and avoids spreading data on

multiple systems such as robot controllers, data ac-

quisition systems and developer workstations.

ACKNOWLEDGMENTS

This project has received funding from the Clean Avi-

ation Joint Undertaking (CAJU) under grant agree-

ment CS2-LPAGAM-2020-2023-01. The JU receives

support from the European Union’s Horizon 2020 re-

search and innovation program.

DISCLAIMER

The results, opinions, conclusions, etc. presented in

this work are those of the author(s) only and do not

necessarily represent the position of the JU; the JU is

not responsible for any use made of the information

contained herein.

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

598

REFERENCES

Ageorges, C., Ye, L., and Hou, M. (2000). Experi-

mental investigation of the resistance welding for

thermoplastic-matrix composites. Part I: heating ele-

ment and heat transfer. Composites Science and Tech-

nology, 60(7):1027–1039.

Endraß, M., Engelschall, M., Jarka, S., Bauer, S., and Fis-

cher, F. (2020). Deliverable D2.1.8-55 (DEK), Opti-

mization of welding parameters of LM PAEK. Tech-

nical report, Clean Sky 2, D2.1.8-55.

Endraß, M., Engelschall, M., Mayer, M., G

answ

urger, P.,

Vistein, M., Sch

onheits, M., Jarka, S., Bauer, S., Fis-

cher, F., Schuster, A., Larsen, L.-C., Ferstl, S., and

Kupke, M. (2022a). Robust assembly - quality as-

sured welding technologies for full-scale applications.

In SAMPE Europe Conference 2022, Hamburg, Ger-

many.

Endraß, M., Thom

e, A., Gadletz, V., Bauer, S., Jarka, S.,

answ

urger, P., Fischer, F., Ferstl, S., Larsen, L.-C.,

and Kupke, M. (2022b). Resistance welding of low-

melt polyaryletherketone: Process deﬁnition and opti-

mization. In 20th European Conference on Composite

Materials, Lausanne, Switzerland.

Fischer, F., Endraß, M., Deden, D., Brandt, L., En-

gelschall, M., G

answ

urger, P., Mayer, M., Vistein,

M., Sch

onheits, M., Schuster, A., Jarka, S., Bauer, S.,

Hellbach, O., Larsen, L.-C., Kupke, M., and Ferstl, S.

(2022). How to Produce a Thermoplastic Fuselage. In

Herrmann, A., editor, ITHEC 2022, 6th International

Conference & Exhibition on Thermoplastic Compos-

ites, volume 6 of International Conference & Exhibi-

tion on Thermoplastic Composites. M3B GmbH, Bre-

men, Germany.

Haase, T., Gl

uck, R., Kaufmann, P., and Willmeroth, M.

(2021). shepard - storage for heterogeneous prod-

uct and research data. Online, https://doi.org/10.5281/

zenodo.5091604.

Jansen, D. and B

uttner, H. (2004). Real-time Ethernet:

the EtherCAT solution. Computing and Control En-

gineering, 15:16–21(5).

Krebs, F., Willmeroth, M., Haase, T., Kaufmann, P., Gl

uck,

R., Deden, D., Brandt, L., and Mayer, M. (2021). Sys-

tematische Erfassung, Verwaltung und Nutzung von

Daten aus Experimenten. In Deutscher Luft- und

Raumfahrtkongress 2021. Deutsche Gesellschaft f

Luft- und Raumfahrt.

Merch

an, D. F., Peralta, J. A., Vazquez-Rodas, A., Min-

chala, L. I., and Astudillo-Salinas, D. (2017). Open

Source SCADA System for Advanced Monitoring of

Industrial Processes. In 2017 International Confer-

ence on Information Systems and Computer Science

(INCISCOS), pages 160–165.

OPC Foundation (2022). OPC 10000-1: UA Part 1:

Overview and Concepts. Online, https://reference.

opcfoundation.org/Core/Part1/v105/docs/.

OpenEtherCAT Society (2019). Simple Open Ether-

CAT Master Library. Online, https://github.com/

OpenEtherCATsociety/SOEM.

Phuyal, S., Bista, D., Izykowski, J., and Bista, R. (2020).

Design and implementation of cost efﬁcient SCADA

system for industrial automation. International Jour-

nal of Engineering and Manufacturing, 10(2):15.

Thom

e, A. (2021). Ermittlung und Validierung

der Prozessgr

oßen beim elektrischen Wider-

standsschweißen von faserverst

arkten Hochleis-

tungsthermoplasten und deren Einﬂuss auf die

Verbindungsqualit

at. Bachelors thesis, HTWK

Leipzig.

Vargas-Salgado, C., Aguila-Leon, J., Chi

nas-Palacios, C.,

and Hurtado-Perez, E. (2019). Low-cost web-based

supervisory control and data acquisition system for a

microgrid testbed: A case study in design and imple-

mentation for academic and research applications. He-

liyon, 5(9):e02474.

Yousefpour, A., Hojjati, M., and Immarigeon, J.-P. (2004).

Fusion bonding/welding of thermoplastic compos-

ites. Journal of Thermoplastic Composite Materials,

17(4):303–341.

S¸ahin, S. and

Is¸ler, Y. (2013). Microcontroller-based

robotics and scada experiments. IEEE Transactions

on Education, 56(4):424–429.

Single Source of Truth: Integrated Process Control and Data Acquisition System for the Development of Resistance Welding of CFRP Parts

599