LABORATORY 2.0

Towards an Integrated Research Environment for Engineering Mechanics

Jonas Schulte, Reinhard Keil

Heinz-Nixdorf-Institute, University of Paderborn, Fuerstenallee 11, 33102 Paderborn, Germany

Johann Rybka, Ferdinand Ferber

Department of Engineering Mechanics, University of Paderborn, Warburger Str. 100, 33098 Paderborn, Germany

Keywords:

Engineering, Integration, Laboratory, Mechanics, Modularization, SOA.

Abstract:

Cooperation Support Systems, respectively CSCW-Systems, increasingly offer standardized interfaces to al-

low their integration into university-wide IT infrastructures. However, several disciplines (e.g. engineering

and medical science) require the use and seamless integration of additional applications to meet researchers’

requirements and support collaboration in a sustainable manner. This articles outlines the possibilities to inte-

grate high-tech laboratories into existing IT infrastructures to strengthen the exchange of information among

teaching, research, and industry. Since, laboratory components are usually characterized by proprietary in-

terfaces, we replaced these manufacturer-speciﬁc interfaces and protocols by a service-oriented architecture

for laboratories. Therefore, functionalities of laboratory components will be encapsulated as a service and

made accessible by Linux Field-Bus Couplers. The modularization of the laboratory allows the connection

to the world of e-learning documents. This article highlights the symbiosis between research in engineering

and teaching at universities. The authors explain in the article that not only the research can take a signiﬁcant

inﬂuence on teaching, but also vice versa, the teaching is a part of the later researches in laboratories.

1 HIGH-TECH LABORATORIES:

INACCESSIBLE KNOWLEDGE?

In many academic disciplines the transfer of knowl-

edge between top level research and teachings in uni-

versities is sluggish and unacceptable for modern re-

search institutes (see (Nahar et al., 2001) and (Potoc-

nik J., 2007)). Particularly, research facilities that use

complex and expensive testing equipment are often

reserved for a small number of researchers. Hence,

costly acquired research results are slowly dissemi-

nated and the transfer of current research results to

lectures is barely given. A second challenge are

proper backup and archiving solutions for test results

as well as the related experimental parameters. Yet,

the knowledge about the relation between test evalu-

ations and additional parameters of the test set-up is

fundamental for the later re-use of obtained results.

In the area of materials engineering, compo-

nents are subjected to cyclic thermo-mechanical

stress (Mahnken, 2008). Thermal shock is an extreme

form of thermo-mechanical stress of components,

which occurs particularly in components of mechan-

ical engineering. These highly specialized laborato-

ries usually consist of many components from differ-

ent manufacturers. Most of today’s laboratory com-

ponents provide only proprietary and vendor-speciﬁc

interfaces. Hence, the laboratory does not provide

enough ﬂexibility in terms of adjustments and reor-

ganizations (see (Nahar et al., 2001) and (Potocnik J.,

2007)). Furthermore, this entails media breaks, which

prevent the transfer of research results to teaching as

well as to cooperation partners from industry.

In this article we present a service-oriented ap-

proach for laboratory architectures. The objective is a

modularized thermal shock laboratory, allowing more

ﬂexibility in respect to the test set-up and easy data

exchange among the laboratory and existing applica-

tions of an university-wide infrastructure. Standard-

ized interfaces are necessary for easy processing and

re-use of test results. A laboratory component can be

addressed in a much more ﬂexible way by encapsulat-

ing its functionality as a service and making it acces-

sible by using Linux Field-Bus Coupler. This mod-

407

Schulte J., Keil R., Rybka J. and Ferber F..

LABORATORY 2.0 - Towards an Integrated Research Environment for Engineering Mechanics.

DOI: 10.5220/0003483504070412

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 407-412

ISBN: 978-989-8425-56-0

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

ularization of the laboratory and the standardization

of the interfaces allow a seamless integration of the

laboratory into the existing university infrastructure.

Knowledge creation and knowledge transfer should

no longer be considered separately, since they inﬂu-

ence each other. For us high-tech laboratories play a

central role for university teaching, they are a source

of knowledge. New learning and teaching opportu-

nities will arise by the close coupling of knowledge

creation and knowledge transfer. To guarantee efﬁ-

cient access to latest research results in the context of

courses, high-tech laboratories have to be consistently

integrated into university-wide infrastructures.

This article is organized as follows. In section 2

a service-oriented architecture for laboratories will be

explained. As an example ﬁeld of application the im-

plementation is realized for a thermal shock labora-

tory. Thereto, we ﬁrst refer to the service encapsula-

tion of laboratory components (section 2.1) and then

the use of Linux Field-Bus Coupler

for the modular-

ization of the laboratory network will be discussed.

The section 2.3 deals with safety aspects in the lab-

oratory environment. Section 3 tries to answer the

question how to facilitate and accelerate the trans-

fer between research and teaching. Afterwards sec-

tion 3.1 introduces a framework for building collabo-

rative learning and working environments. Further-

more, the framework supports in integrating exist-

ing heterogeneous systems into complete networks to

consolidate system convergence. Section 4 presents

LTM-SOLA, a service-oriented laboratory applica-

tion. This application is a browser-based graphical

user interface to plan, prepare, control, and coordi-

nate thermal shock experiments. Finally, we discuss

our results in section 5 and give a short outlook to

further work.

2 A SERVICE-ORIENTED LAB

ARCHITECTURE

The intention to develop a service-oriented laboratory

architecture has two main reasons. On the one hand

to arrange laboratory components in a ﬂexible way by

modularization of the control software. This allows

researchers to adjust the laboratory regarding the ac-

tual needs as well as replacing existing and adding

new devices to the laboratory’s integrated network.

On the other hand the service-oriented architecture in-

tends to enable a simple and sustainable integration of

laboratories into existing information infrastructures.

Linux Field-Bus Coupler (LFBC) – a 32-bit ARM pro-

cessor system with an embedded Linux operating system.

Thereby the knowledge ﬂow between research in lab-

oratories and teaching in courses can be sustainable

improved.

2.1 Service Encapsulation of

Laboratory Equipment

The objective is to avoid media breaks between dif-

ferent systems to allow a continuous information

ﬂow. Although it is impossible to reduce every media

break, e.g. those that occur when shifting from digi-

tal to analog media and vice versa. However, media

breaks have to be reduced in order to ensure system

convergences (Keil-Slawik, 2005).

In order to consistently reduce media breaks and

to ensure an integrated system network, which in-

cludes the laboratory, uniﬁed interfaces have to be

developed for all laboratory devices. The issue is that

laboratory devices are typically equipped with propri-

etary protocols. Uniform interfaces, or even a consis-

tent API to control the devices do not exists. Thus, the

idea to develop a service-oriented architecture for lab-

oratories may be a suitable approach. Thereby, func-

tions of the laboratory components will be encapsu-

lated as a service. In (Ferber et al., 2008) the authors

describe a possibility to make laboratory components

accessible using web services. For this purpose Java

classes have been developed that make the laboratory

device’s functionalities accessible via modern web

services and encapsulate proprietary protocols. This

concept with uniform interfaces and protocols makes

laboratory components accessible from “any” appli-

cation (like e-learning, CSCW applications or digital

libraries).

Nevertheless, the disadvantage of the presented

approach is that communication between Java-based

services and the laboratory components still require

a central master computer. This master computer is

responsible for the entire interaction and is a single

point of failure. In addition the performance and re-

sponse time of the control software is an important

factor when using laboratory services in real exper-

imental execution. Hence, it is valuable to mod-

ify the concept of web services by replace the cen-

tral master computer with a service-oriented applica-

tion, that is modularized and thus very ﬂexible. To

make laboratory components more independent how

the web services have made it, the LFBC infrastruc-

ture will be introduced. The company WAGO

pro-

vides industry standard Field-Bus Couplers that are

suitable for our purpose due to its network interface

and an embedded Linux as operating system. This

WAGO



is a German company for electrical intercon-

nections and automation (http://www.wago.com).

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

408

approach is unique, because the most laboratories use

the PLC technology. WAGO provide a wide range

of ﬁeld bus modules for any devices. So only one

manufacturer is involved, the system is extensible and

easily programmable in C. The LFBC coupling the

manufacturer-speciﬁc interfaces and protocols of the

laboratory devices in blocks that are run cyclically to

update the signal states. So any device is equipped

with a ﬁeld bus modules and the speciﬁc functional-

ity that is embedded in a building block and running

in the execution cycle of the LFBC. These function-

ality is accessible from the local network by bidirec-

tional inter-process communication ﬂow, called net-

work socket.

2.2 Intelligent Laboratory Components

In section 2 the advantages of a service-oriented ar-

chitecture for laboratories are emphasized. Beside it

scientists wants to focus on their core tasks such as

planning, setting up, the actual execution, and later on

the evaluation of thermal shock experiments. The for-

mer laboratory architecture had the fundamental dis-

advantage that an experiment is non-exclusively in-

ﬂuenced by the control software. Instead, the execu-

tion is affected by external conditions (a central mas-

ter computer and a Siemens PLC

The solution presented in this article breaks the

rigid link between the central master computer and

laboratory devices by using LFBCs (see section 2.1).

This allows completely new experiment scenarios, for

example, to determine the PID

control parameter

prior to the actual experiment execution. The PID

controller calculation (algorithm) involves three sepa-

rate parameters, and is accordingly sometimes called

three-term control. The PID parameters are usually

determined once for a speciﬁc control system. How-

ever, the conditions in the thermal shock laboratory

are continuously changing, because scientists experi-

ment with various samples that differ in their mate-

rial composition and therefore require speciﬁc PID

parameters. Thus, the PID deﬁnition has to be ad-

justed individually for each test set-up.

Figure 1 shows a comparison between a standard

2-point control and the advanced PID control. The

blue and the black lines stand for the desired values.

The purple line represents the actual values using PID

controller and the yellow line constitutes the actual

The PLC is a memory-programmable control unit.

A proportional-integral-derivative controller (PID con-

troller) is a generic control loop feedback mechanism (con-

troller) widely used in industrial control systems – a PID

is the most commonly used feedback controller. (Further

information: http://mhf-e.desy.de/e638/e1770/).

Figure 1: Comparison Between 2-Point Control and Auto-

matic PID Control.

values using 2-point control. Obviously, the PID-

based parameters signiﬁcantly improve the automatic

control.

Conditioned by the encapsulation of laboratory

devices as services, associated with the use of LFBC

infrastructure, the calculation of appropriate PID pa-

rameter was made possible. Thus, the induction heat-

ing can be adjusted to the current test set-up for an

efﬁcient automatic control. The presented functional-

ity is only a fraction of what opportunities the system

offers for integration of building blocks.

2.3 Safety Aspects of the New

Laboratory Architecture

In laboratories often heavy and dangerous devices are

used, hence a faulty test set-up and inappropriate use

may lead to fatal accidents. In addition, failures may

imply huge ﬁnancial losses. To avoid this kind of

inappropriate use a major task of the Siemens PLC

was to ensure the safe execution of thermal shock ex-

periments. For example, the safety door of the test

chamber must be closed before the induction heating

can be switched on. This precaution is implemented

on the PLC. In the new laboratory architecture the

Linux Field-Bus Coupler has to take over this task

and thereby ensure the safe experiment execution. In

a LFBC the input and output signals are presented as

a process image. To facilitate the access to this rela-

tively complicated process image, a program was de-

veloped which provide a high-level access to the pro-

cess image. The access on a high-level is done by a

program called “iocontrol”, which works like an op-

erating system and implements the PLC functionality

for a LFBC. This allows to call the modules cycli-

LABORATORY 2.0 - Towards an Integrated Research Environment for Engineering Mechanics

409

cally that are responsible for controlling one or more

devices.

3 KNOWLEDGE CREATION AND

KNOWLEDGE TRANSFER

There exist numerous approaches to sophisticated im-

plementations of virtual laboratories that can support

students to learn practical work without the risks in

real laboratories (Ramat and Preux, 2003). Never-

theless, it is insufﬁcient for the universities to edu-

cate their students only in virtual laboratories. Rather

there is the requirement to prepare students as good as

possible by practical scenarios and assistance in real

laboratories for the industry’s needs. Furthermore,

high-tech facilities cooperate in many ways with the

industry and especially with medium-sized compa-

nies, for those who cannot afford an own test bench.

Experiments that are performed in the high-tech labo-

ratory such as the thermal shock laboratory represent

the information demand of many engineering compa-

nies.

In addition the authors follow this approach to

keep the delay between the knowledge creation and

knowledge transfer in teaching as low as possible.

Hence, the integration of the new laboratory architec-

ture into existing infrastructures is essential. The pre-

sented laboratory infrastructure in section 2 not only

enhances the ﬂexibility regarding the arrangement of

laboratory devices, but also allows simpliﬁes the in-

tegration of the laboratory into existing information

infrastructures. This includes in particular digital li-

braries, in which experiment results and experiment

parameters can be stored permanently and centrally.

Since, digital libraries are used as knowledge bases

for teaching and e-learning platforms in many ways,

the connection between the laboratory and digital li-

braries is very important. Through the direct link me-

dia breaks are eliminated and experiment results can

be directly or after a appropriate preparation used for

teaching. Before give an idea of how to use the results

in teaching we will introduce you in the WasabiBeans

Framework, which provide a wide range of rights and

user management, important for give accessibility to

the corresponding results.

3.1 System Integration with the

WasabiBeans Framework

To reach a direct link between different systems or

even across system classes they must be conform

connected to an integration layer or a message-bus

(Schmidt et al., 2005). Only in this way a complex

exchange of information between these systems can

be effected. One possibility to ﬁnd adequate support

here is the WasabiBeans framework (Schulte et al.,

2008). WasabiBeans is a framework for building col-

laborative learning and working environments and the

integration of heterogeneous systems into a system

group. This framework relies on a JBoss Application

Server

platform and therefore allows the use of many

existing standards such as JAAS or JCR with that the

ﬂexibility in the directory- and persistence layer can

be ensured.

Figure 2 shows the new laboratory architecture

that can be connected with arbitrary applications us-

ing WasabiBeans as a service-oriented platform for

system integration. The decision for the use of

WasabiBeans framework has to be justiﬁed, in par-

ticular, that the fast transfer of information, without

a break in media should be ensured between research

and teaching. To use the collaboration between sci-

entists in the laboratory on the one hand and the user

group, which will beneﬁt form the experimental re-

sults, should be effectively enhanced by the develop-

ment of an integrated infrastructure. The data model

of WasabiWeans implements the concept of virtual

knowledge spaces (Hampel et al., 2004). Therefore,

the framework is ideally suited to structure and orga-

nize information as well as collaborative work with

documents. An essential function is the manage-

ment of experiment results and teaching materials for

courses. Thus, there are novel teaching and research

possibilities by having ﬂexible capabilities of storage

as well as the ﬁne tuned rights and user management.

3.2 Bundling of Laboratory Services as

WasabiBeans-Module

In section 2.1 we discusses the advantages of encap-

sulating laboratory components and make them ac-

cessible as services. To access the individual ser-

vices better, they are grouped together as a module

and added to the WasabiBeans framework. Due to

the generation of a WasabiBeans module it is possible

to let the services run on the same JBoss AS, which

had deployed the WasabiBeans framework. This has

the great advantage that all calls to services of other

modules, such as a service for storing documents in

A message-bus denoted in the information technology

a class of software solutions that support the integration of

distributed services.

The JBoss Application Server (JBoss AS) is the world’s

most widely used Java application server. Available online

at: http://www.jboss.org/jbossas.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

410

WasabiBeans

Virtual

Knowledge Spaces

LTM-SOLA

Frontend

Encapsulation of

Laboratory Services

Documents and

Publications Server

Seam

Framework

Apache

Jackrabbit

EJB 3.0

JBoss Application Server

110 01010

001111101

110 01010

001111101

Instruction Set for

Induction Heating

Service-

Encapsulation

Group of

Researchers

Figure 2: Serivce-Oriented Laboratory Architecture Based

on LFBC.

a digital library, to be carried out only local. This

means in particular that no RMI calls or web service

calls are necessary. Performance measurements have

shown that this is a speed increase by a factor of up

to 1000 is possible. Another reason for the creation

of a module is the easier use of the new lab services

from existing applications. This concerns not only

those who already use the WasabiBeans framework,

but also the applications of a university-wide infras-

tructure that have not yet been fully integrated into the

system group. The WasabiBeans framework offers a

variety of cooperative support services and therefore

facilitates the interface to existing applications. The

developer has a high degree of ﬂexibility through the

use of the framework. For example, events can be

triggered by the completion of an experiment, that im-

ply actions required in other applications.

4 LTM-SOLA -

SERVICE-ORIENTED

LABORATORY APPLICATION

The developed application LTM SOLA

is a key

ﬁgure to forward the modularization of the labora-

tory and creates a large degree of ﬂexibility with the

service-oriented approach concerning to prepare of

thermal shock experiments and archiving the results.

Through the provision of web services is set to techni-

cal standards and standard interfaces are offered. By

internal structure of the business logic as JavaBean

classes, the bundling of the provided services is pos-

sible. Access to each service is coordinated with the

speciﬁed interfaces and allows the use of business ob-

jects.

Lehrstuhl fuer Technische Mechanik Serviceorientierte

Laborapplikation

LTM-SOLA provides services, which take over

the planning, preparation, control and coordination of

thermal shock experiments. In the Editor view (see

Figure 3), temperature proﬁles in form of a tempera-

ture curves can be created that is used to control the

induction heating or the cooling device. Also various

conﬁguration options are available in the Scheduler

view that are essential for the execution of the thermal

shock experiments. Here, for example, the number of

heating and cooling cycles can be deﬁned, or the se-

lection of the temperature curves are created in the

editor can be done. In selecting the method of heat-

ing the user can choose between the Standard method

with a temperature curve, the TwoLines method for

controlling the temperature gradient or the SelfLearn

method for determining the PID parameters. During

an experimental, the processes in the monitor view

will be followed in real time. In a chart, the actual

and set temperatures are reviewed alongside the ex-

periment.

Figure 3: Editor-View of LTM-SOLA.

LTM-SOLA was designed based on the new re-

quirements and can be used to control laboratory

components, but it was particularly concerned to in-

teroperability with other systems and a great value

where placed on the cooperative experimental exe-

cution. In LTM SOLA it is possible to deﬁne roles

for each individual page views, so that not every user

can do the actual control of the laboratory equipment.

With a login cooperation partner and another univer-

sity member get access to the editor’s view in order to

create a heating proﬁle for the experiment. The heat-

ing proﬁle can be then saved in the system and later

loaded from a scientist on the ground to execute this

experiment. Through the use of WasabiBeans frame-

works, the storage of experiment results in different

repositories is possible. LTM SOLA for example, can

store the experiment results form the Result view in

the digital document and publication server Miless for

permanent archiving. More information about Miles

are (Gollan et al., 1999). Miles is in turn integrated

LABORATORY 2.0 - Towards an Integrated Research Environment for Engineering Mechanics

411

into various systems such as an e-learning system, as a

source of knowledge. In this way, experiment results

can be accessed from different systems without media

breaks to bring teaching and research closer together.

5 CONCLUSIONS

High-tech institutions cooperate in many ways with

the industry and especially with medium-sized enter-

prises that are not able to operate their own labora-

tory due to the high costs. Furthermore, universities

try to educate their student in a professional way. On

the one hand by involving students in current research

projects and on the other hand by taking the neces-

sities of the industry into account. The cooperation

process between academics, industry partners and stu-

dents was interrupted by the lack of standardized in-

terfaces. Figure 4 shows the thermal shock laboratory

as an integral part of a university-wide system infras-

tructure. The systems are not isolated anymore, but

linked by a service-oriented approach to an integrated

systems network for scientiﬁc research and work pro-

cesses.

Planning System

Knowledge Management /

E-Learning-System

Digital Library /

Document Management

transport

teaching units

research,

publish,