See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/277594812

Green Cockpit: Transparency on Energy Consumption in Manufacturing

Companies

Article in Procedia CIRP · December 2015

DOI: 10.1016/j.procir.2015.01.011

CITATIONS READS

16 2,026

6 authors, including:

Tobias Rackow Tallal Javied

Friedrich-Alexander-University of Erlangen-Nürnberg Friedrich-Alexander-University of Erlangen-Nürnberg
9 PUBLICATIONS 120 CITATIONS 16 PUBLICATIONS 234 CITATIONS

SEE PROFILE SEE PROFILE

Toni Donhauser Peter Schuderer

Friedrich-Alexander-University of Erlangen-Nürnberg Technische Hochschule Ingolstadt
22 PUBLICATIONS 183 CITATIONS 39 PUBLICATIONS 116 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Jörg Franke on 25 January 2016.

The user has requested enhancement of the downloaded file.

 Available online at www.sciencedirect.com

ScienceDirect
Procedia CIRP 26 (2015) 498 – 503

12th Global Conference on Sustainable Manufacturing

Green Cockpit: Transparency on Energy Consumption

in Manufacturing Companies
T. Rackowa*, T. Javieda, T. Donhausera, C. Martina, P. Schudererb, J. Frankea
a
Institute for Factory Automation and Production Systems, University of Erlangen-Nürnberg (FAU), Egerlandstraße 7-9, 91058 Erlangen, Germany
b
Technische Hochschule Ingolstadt (THI), Esplanade 10, 85049 Ingolstadt, Germany

* Corresponding author. Tel.: +49-9131-85-27176; fax: +49-9131-302528. E-mail address: tobias.rackow@faps.uni-erlangen.de

Abstract

Companies are forced by rising energy costs to seize control of their energy consumption to maintain contestability. There-
fore, transparency over the companies’ energy flux along the production process is required. Energy data management software
is helpful, but cost-intensive. Hence, especially small and medium sized enterprises (SME) spare this investment. In this paper,
the requirements for an energy controlling infrastructure in SME are elaborated, followed by a deduced software-architecture
which supports the respective controlling structure. Further, the prototypical realization of the corresponding tool “Green-
Cockpit” will be presented. The free of cost, open source and web-based tool is designed to help companies monitor, interpret,
analyze, plan and report their energy consumption. The Green Cockpit tool outperforms other energy management software at
management disciplines with its ability to not only analyze energy consumption, but plan and control it additionally.
© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
© 2014 The Authors. Published by Elsevier B.V.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of Assembly Technology and Factory Management/Technische Universität Berlin.
Peer-review under responsibility of Assembly Technology and Factory Management/Technische Universität Berlin.
Keywords: Energy controlling, energy management, operations management, open source data management

1. Introduction ers of energy metering equipment provide energy data man-

agement software. However, this software does not address
Many, especially producing companies, are aware of the the problem in an ideal way for two reasons:
urge to save resources. Hence Thamling et al. [1] observed in Usually, the software is cost intensive, which enlarges the
their study in 2010, that 50 % of all polled enterprises in the restraint of companies (especially SME) to invest into an
segment of small and medium-sized enterprises (SME) con- energy monitoring system. Further, the software modules
sider energy efficiency as an important topic. support mainly the analysis of energy consumption with an
Due to a lack of tools and instruments for monitoring and emphasis on the technical aspect. There is yet no satisfying
controlling of resource consumptions, there is often times a solution for the support of operative and strategic decision
missing of ability to identify and raise efficiency potentials. making as well as for the rolling planning process along with
On a methodological level, energy management systems pro- ongoing controlling.
vide support. With the framework, which is suggested for The following paper compiles the reasons for, and the re-
instance in the ISO 50001, energy streams become transparent quirements of an energy controlling in SME and describes the
in order to reveal saving potential. structure of a prototypical energy controlling tool “Green
Even though there is given a basic method of how to pro- Cockpit” which meets these requirements as described.
ceed for the evaluation of a company’s energy consumptions
hereby, a tool is missing which fulfills the ongoing need of
analyzing, planning and controlling energy consumptions
from an information technological perspective. Many suppli-

2212-8271 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of Assembly Technology and Factory Management/Technische Universität Berlin.
doi:10.1016/j.procir.2015.01.011
 T. Rackow et al. / Procedia CIRP 26 (2015) 498 – 503 499

2. Energy Transparency mentation of a controlling position and the split-up of the

controlling tasks. The extension of the regular controlling
If the metering of energy consumption is being executed in scope of duties with the field of energy involves, according to
a systematic, regular and repetitive way, one speaks of energy Gänßlein et al. [9], the following steps:
monitoring. If the gained data are being augmented with out-
put values (e.g. consumption per unit produced) and used in a 1. Provision or generation of energetic information as pro-
controlling circuit, one can speak of energy controlling, which cess input.
represents a tool for generating transparency. [2] 2. Processing or transformation of energy information via
As Hesselbach [3] notes, there are two fundamental ways appropriate controlling instruments within the processes.
of measuring energy consumptions in companies: 3. Provision of preprocessed energy information in a com-
prehensible form to the process customers.
x Top-down: Measuring the overall energy consumption (i.e.
for a production site by the energy supplier) along with Taking the management reporting as an example process,
subsequent derivation of information for the causative pro- energy information needs to be gathered, analyzed, reported
cesses. and discussed besides financial information, for instance.
x Bottom-up: Metering every single consumer (i.e. every
machine of a site by using energy meters) along with the 3.3. Configuration of energy controlling
derivation of information for the overall consumption.
The foundation of energy controlling is the energy moni-
For the reason of an adequate, relatable and accurate me- toring, hence the systematic and regular capture of all relevant
tering with low time expenditures, permanently installed me- energy data. According to Hesselbach [3], it requires the fol-
ters should be implemented according to the bottom-up ap- lowing components to realize an energy monitoring system:
proach, which measure the consumption according to defined
parameters. x Measuring equipment (i.e. intelligent meters)
x Transmission equipment (i.e. bus cabling)
3. Energy controlling in enterprises x Signal conversion
x Data logging (data base)
For organizational reasons, there is often times an institu-
tional segmentation of strategic and operative controlling, A data preparation and analysis is required for a sound
whereby both departments contain the fields planning, con- evaluation of the energy consumption after the elicitation. In
trolling and reporting in the corporate practice. this context, data preparation primarily means the generation
of comparability. Based on Schön [10], one can compare
3.1. Strategic and operative controlling energy data for instance as follows:

As Fischer et al. [4] clarify, the main difference between x actual-vs.-actual: Comparison of the energy consumption
operative and strategic controlling lies in the goal of both of a single consumer; for instance comparing a present
alternatives. While the strategic controlling aims to ensure the value with a historical value
sustainable and permanent existence of the company, the x benchmark-vs.-actual: Comparison of the energy consump-
operative controlling strives year after year for the improve- tions of two consumers (one of both being the benchmark)
ment of profitability and productivity. The operative control- x target-vs.-actual: Comparison of actual values with target
ling acts as supporting organ of the operative management values, which were planned in the past
level in order to meet these goals and sets its focus on the own x target-vs.-forecast: Comparison of target values with fore-
undertaking under consideration of tactical and operational cast values (values which are adjusted by actual values)
planning. The success of the operative controlling bases ac-
cording to Vollmuth [5] substantially on the timely detection The expressiveness of measured values can be raised
of variances from the defined goals. through the generation of appropriate key figures by associat-
With this classification, energy controlling can be assigned ing the values with output dates. Exemplary performance
to the range of duty of the operative controlling. Particulari- indicators are listed in table 1.
ties arise in the field of SME. Ihlau et al. [6] noticed in this
context, that SME often maintain a simple, fiscal character- Table 1. Exemplary list of energy performance indicators (EnPIs)
ized accounting system, have restricted information sources, EnPI Calculation formula Unit
have only limited internal controls and the planning of busi- Specific energy ‡”‰› ‘•—’–‹‘ Š
ness is not existent or not documented. consumption —–’—– ’‹‡ ‡•
‡”‰› ‘•—’–‹‘‘ˆƒ’”‘ ‡••
3.2. Incorporation of energy controlling in SMU Energy intensity Ψ
‘–ƒŽ‡‡”‰› ‘•—’–‹‘
Specific energy ‡”‰› ‘•—’–‹‘ Š
From the four solutions to organize controlling in SME, costs ‘•–•‘ˆ’”‘†— –‹‘ ‡—”‘
described by Peemöller [7], mainly two are deployed in cor- Revenue based ‡˜‡—‡ሺ’‡”’”‘†— –ሻ
porate practice according to Reiß and Denison [8]: the imple- Ψ
energy efficiency ’‡ ‹ˆ‹ ‡‡”‰› ‘•–•
500 T. Rackow et al. / Procedia CIRP 26 (2015) 498 – 503

After the data preparation and formation of performance Table 2 Requirements for the Green Cockpit
indicators, one can proceed with the energy analysis. The Category Requirement
analysis contains the comprehensive contemplation of the
monitored energy consumption. Administration Access control and user management
Due to the enormous amount of data, instruments for visu-
alization must be deployed in order to master the complexity. Interfaces Optional export into Microsoft Excel
According to Grahl [11], energy saving potentials can be Data Generation and compression of characteristic values
revealed with the help of these instruments by a macro analy- processing and EnPIs
sis, followed by a micro analysis for further detailing.
After identifying energy saving potentials, it is important Controlling Variance analysis (i.e. actual figure vs. target value)
to work systematically and focused on the realization of the
appropriate measures. The determination of realistic and Reporting Web based standard reports and ad hoc reports
achievable target values is therefore a suitable procedure.
Subsequently, the ongoing comparison of actual key figures/ 4.1. Fundamentals of data warehousing architecture
energy performance indicators with the defined target values
indicates whether the measures are expedient. According to Müller and Lenz [13], business intelligence
The process of determining target values is similarly arbi- (BI) deals with the gathering, preparation and analysis of
trary as the process of forming key figures. Depending on the internal (i.e. operational) and external data of all activities in a
goals and the processes which are to be optimized, each com- company, aiming for the creation of knowledge.
pany has to define its own, realistic target values. According The data gains economic significance by transforming mi-
to Gleich et al. [12], several alternatives of setting target val- cro data (e.g. energy consumption in the seconds range) into
ues from the corporate division “planning and budgeting” analytical data (aggregated, sorted and grouped operational
serve as aid for orientation: data). Analytical data can be derived from micro data via
OLAP-operators (online analytical processing) and stored in
x figures from previous years; data warehouses (DW) in the form of data cubes. Data cubes
x benchmark values; (as multidimensional data structures) allow for historic and
x calculated values (i.e. physical minima); statistical analyses as well as for the generation of key figures/
x top-down derived target values; EnPIs.
x negotiated target values. A comparison between the requirements of an energy con-
trolling with the features of BI shows, that BI fits perfectly for
The review and control process using target values can on- the realization of an energy controlling process. The core of a
ly be effective, if all process customers are being regularly BI-tool is the data warehousing process which contains the
and systematically provided with the latest, appropriate fig- process of data acquisition from internal (operational) and
ures/ EnPIs and target values. The control circuit is closed by external sources, the data transformation and data preparation,
the reporting system, subsequently. the process of loading the data into the data warehouse or data
According to Sicher et al. [4], the outcomes of reporting marts, respectively, and the OLAP-based data analysis.
systems can be differentiated between standard reports, devia-
tion reports and demand reports. With these reports, any pro- 4.2. Basic structure of the Green Cockpit
cess customer is able to initiate measures with the aim to
achieve the defined goals and monitor their impact in the The setup of Green Cockpit follows the basic structure as
aftermath. illustrated in figure 1. All relevant internal and external data
are extracted via interfaces from the connected data sources
4. Prototypical realization of the Green Cockpit and loaded into a staging area. In this temporary storage, the
data are being revised, targeting the homogenization via trans-
The Green Cockpit prototype covers the fields as listed in formation processes depending on their quality and eventual
table 2. The underlying processes are described in the follow- conflicts, and prepared for the generation of performance
ing. indicators.

data warehousing
data sources ETL-systems data warehouse analysis and distribution

document Extraction OLAP process customers

server
Transformation
operations
data base staging area Load data marts
(replicate, aggre- • reports • data mining
external gate, archive) • OLAP • ad-hoc analysis
web core data warehouse • dashboards

Fig. 1. Basic structure of the Green Cockpit

 T. Rackow et al. / Procedia CIRP 26 (2015) 498 – 503 501

After a successful transformation, the data are being saved 4.5. Data warehouse through “Palo Community Edition”
in a core data warehouse and under certain circumstances
outsourced into data marts for performance reasons. With For being able to evaluate the data from the core data
OLAP-based analyses of data and OLAP-queries, these data warehouse in a fast and flexible way via multidimensional-
are accessible for process customers, eventually. analyses, the OLAP-methodology can be used. According to
Kemper et al. [14], OLAP is being described by five criteria:
4.3. Data source
x Fast: Regular queries need to be processed within 5 se-
Internal and external data are usually stored on the data conds, complex queries within a maximum of 5 seconds.
servers/ hosts in a company. The access to these data is exe- x Analysis: An intuitive analysis and the possibility of any
cuted via an interface between the BI-System and the server. calculation must be ensured.
Ideally, relevant data such as manufacturing data (e.g. output x Shared: An effective access control and the possibility of a
quantity) and energy data (e.g. energy consumption) are ready multiple user operation must be provided.
for instant recall on one or several data bases from which they x Multidimensional: The possibility of a conceptual, multi-
can be called up at any time. For the purpose of the Green dimensional view must be existent independently of the
Cockpit these databases are being provided in the *.xls file underlying data base structure.
format. There are five file categories: energy data, financial x Information: Due to the scalability, the response time of
data, manufacturing data, target value data and core data queries must remain stable even at large amount of data.
(which are necessary for merging the categories).
The specialty of multidimensional analyses is the fact, that
4.4. Data processing (ETL) with “Talend Open Studio” (TOS) they have a multidimensional data room underlying. Subse-
quently, OLAP-systems can provide data analysis operations
The process of gathering, transformation and preparation such as pivoting, drill up/ down or split and merge with a very
of data is described by the ETL-process (extraction, transfor- quick response time.
mation, loading). According to Schön [10], the respective The “Palo Community Edition” is a license free basic
process is responsible for the correct transmission of all data package, consisting of tools for the realization of typical BI-
into the data warehouse: processes. The basic package provides all relevant compo-
nents which are necessary for the Green Cockpit prototype.
x Extraction: Extraction of all relevant data from the data The Palo OLAP server is the core component. It is operated as
sources and deposition in a staging area. a multidimensional in-memory-OLAP-server, in which the
x Transformation: Homogenization of all data formats and loading operations are being executed during the starting
data structures which are stored in the staging area accord- procedure of the OLAP-server application, which loads the
ing to the scheme in the core data warehouse by transfor- complete data base from the ETL-process in cube structures
mation processes. into the internal memory. Modeling the data cubes with Palo-
x Loading: Under consideration of the rules for extraction internal tools happens through the provided web module,
and transformation, the data are being loaded into the core which needs to be installed on a central server. This leads to
data warehouse. the advantage, that the Green Cockpit can be operated from
any computer with a common web browser. Depending on the
The implemented tool (Talend Open Studio) defines itself access rights, the user has access to the contents of Green
as a code generator and converts virtually modeled processes Cockpit interface via the user interface. “Palo Pivot” serves as
into source code, which can be executed subsequently. In a a tool for the creation of ad-hoc reports.
first step, all internal and external data are being extracted
from the data sources and loaded into the staging area. As for 4.6. Standard reports in Green Cockpit
the prototype, the source system consists of several *.xls files,
the staging area of a *.csv file. In a second step, the data base, Reports can be created on the user interface of Green
the data cubes and all required dimensions are being compiled Cockpit. After successful login, the user can administrate
along with their dimension elements and loaded eventually. standard reports, which means create, delete, edit, and import
When data base, data cubes, dimensions and elements have or export reports. The four standard reports, provided in the
successfully been compiled, the transformation process is Green Cockpit prototype, are described in table 3.
ready to start. The transformation processes are preparing the Figure 2 shows an exemplary screenshot of the standard
buffered data of the staging area in a way, so that they can be report “energy planning”. Area (1) outlines a mask for plan-
loaded into the data cubes. This data preparation contains, ning the energy consumption of all production sites. In the
besides the homogenization and merging of different data columns of (2), the target factors (as described in table 3) can
table, the formation of pre-aggregates. be set for each plant. The fields in (3) show a tabular compari-
The modeling of the ETL-processes terminates with the son of actual and target (planned) values which are displayed
loading of all prepared data into the cubes and the generation graphically in (4). The fields in (5) serve the purpose to navi-
of key figures according to the specified rules. These rules gate between the different reports (see table 3).
lead to the feature, that any user query (e.g. EnPIs) can be
calculated in real time from the underlying data.
502 T. Rackow et al. / Procedia CIRP 26 (2015) 498 – 503

comes to energy controlling on a management level. Howev-

Fig. 2. Exemplary energy planning mask of the Green Cockpit er, first tests in companies showed that the implementation
into the process data stream is a complex matter, since an
Table 3 Standard management reports provided in Green Cockpit
energy metering infrastructure, which is installed in the after-
Report Description math, does not necessarily have an interface with the operat-
Creation of a key figure ranking: ing data system.
x Tabular and graphical display for top 5 plants and final Figure 3 shows the draft of a data stream which can be im-
Top five products in dependence of the set parameter plemented in a very easy way, since it uses features of exist-
ranking x Management support in the search for saving potential ing software end equipment with minor additional adjust-
at a glance ments only. The interpretation of the energy meter data (im-
x Adjustable parameter: time period, key figure/ EnPI pulses) is being taken care of by a software which is usually
Provision of information of the individual plants: provided by the meter fabricator. The energy data can then be
x Tabular and graphical display of all plants by quarters in exported to a *.xls file similar to the export of operation data
Timing dependence of the set parameter by operating data systems. The data files can finally be loaded
diagram x Depiction of key figures over a time course for an early to a “Green Cockpit data base” for raw data which is then the
identification of trends point of contact for the Green Cockpit as it is described in this
x Adjustable parameter: year, key figure/ EnPI paper.
Planning of energy consumption in the production process
x Definition of target values (key figures/ EnPIs) for a Consumer 1 Consumer 2 Consumer 3
selectable period of time and production sites. meter meter meter
x Display of the deviation between actual and target value Raw data
Energy for the monitoring of measures. Operating data (impulses)
Data logger
planning (i.e. output quantity)
x A first draw for target values is suggested by a multipli-
Interpretation and transformation
cation of actual values with target factors (i.e. increase of the impulses into evaluable
of production by 5 % along with increased energy effi- figures (e.g. kWh)
ciency by 1.5 % in the following quarter. The resulting
values can then be modified if necessary. Operating data system Software
Already existent in the of the energy meter equipment
Analysis of key figures for final products
company provider
x Tabular and graphical display of the figures of a plants
final products for two time periods in dependence of Automatic and regular export of data into .xls files
production site and key figure
Analysis Server
x Management support for comparing key figures of final Folder with .xls files
products for different time periods
(e.g. Q1/’13 vs. Q1/’14)
x Adjustable parameter: Time period 1 & 2, key figure, Green Cockpit data base
Production site x Energy data
x Production data
x Company data
5. Conclusions
Communication
With the development of the Green Cockpit prototype, a
tool is being designed which meets the identified require-
ments for an energy controlling in SME. Hence, it helps man- Green Cockpit
agers to identify critical energy related figures. Discussions
with energy data management software producers indicate, Fig. 3. Data stream for Green Cockpit 2.0
that the Green Cockpit prototype strikes a new path when it
 T. Rackow et al. / Procedia CIRP 26 (2015) 498 – 503 503

Furthermore, testing the Green Cockpit gave the insight, [3] Hesselbach J. Energie- und klimaeffiziente Produktion: Grundlagen,
Leitlinien und Praxisbeispiele. Wiesbaden: Springer Vieweg; 2012.
that it can be of interest to evaluate electricity expenditures in [4] Fischer TM, Möller K, Schultze W. Controlling: Grundlagen, Instrumente
the context of other expenditures, such as personnel or mate- und Entwicklungsperspektiven. Stuttgart: Schäffer-Poeschel; 2012.
rial. The next version of Green Cockpit will have an interface [5] Vollmuth HJ. Controllinginstrumente. 5. akt. Aufl. München: Haufe;
2011.
to accounting tools which will provide information that the [6] Ihlau S, Duscha H, Gödecke S. Besonderheiten bei der Bewertung von
operating data system cannot provide. KMU: Planungsplausibilisierung, Steuern, Kapitalisierung. Wiesbaden:
In the end, managers are designated as users of the Green Imprint: Springer Gabler; 2013.
Cockpit. However, some managers might not be coath to [7] Peemöller VH. Controlling: Grundlagen und Einsatzgebiete. 5., Aufl.
Herne, Berlin: Verl. Neue Wirtschafts-Briefe; 2005.
work in a web-based environment. Subsequently, it is to be [8] Reiß H, Denison E. Unternehmensführung und Controlling: Aus der
tested, if a Microsoft Excel based version of Green Cockpit Studienserie Erfolgsfaktoren im Mittelstand: Deloitte Mittelstandsinsti-
can provide the same flexibility and functionality without tut. Bamberg; 2008.
[9] Gänßlein S, Losbichler H, Horváth P, Michel U, editors: Internationaler
losing the performance advantages of a web based tool. Controlling Verein ICV. Green Controlling: Relevanz und Ansätze einer
"Begrünung" des Controlling Systems. Gauting, Stuttgart; 2011.
Acknowledgements [10] Schön D. Planung und Reporting im Mittelstand: Grundlagen, Business
Intelligence und Mobile Computing. Wiesbaden: Gabler Verlag; 2012.
[11] Grahl A. Handbuch für betriebliches Energiemanagement: Systematisch
The research project Green Controlling is funded by the Ba- Energiekosten senken. Berlin: dena; 2009.
varian State Ministry of Education, Science and the Arts. The [12] Gleich R, Hofmann S. Planungs- und Budgetierungsinstrumente: Innova-
authors are deeply grateful for this support and are responsible tive Ansätze und Best-Practice für den Managementprozess. 1. Aufl. s.l:
Haufe Verlag; 2006.
for the contents of this publication. [13] Müller RM, Lenz H. Business Intelligence. Berlin, Heidelberg: Imprint:
Springer Vieweg; 2013.
References [14] Kemper H, Mehanna W, Baars H. Business intelligence - Grundlagen
und praktische Anwendungen: Eine Einführung in die IT-basierte Ma-
nagementunterstützung. Business Intelligence - Grundlagen und prakti-
[1] Thamling N, Seefeldt F, Glöckner U. Rolle und Bedeutung von Energieef- sche Anwendungen 2010.
fizienz und Energiedienstleistung in KMU: Prognos AG. Berlin; 2010.
[2] Rackow T, Schuderer P, Franke J. Green Controlling: Ressourcenorien-
tierte Steuerung von Unternehmen. Zeitschrift für wirtschaftlichen Fab-
rikbetrieb 2013; 108(10): p. 773–777.

View publication stats