Development of Intelligent Assistance System to Support
Eco-efficient Planning
Sarfraz Ul Haque Minhas and Ulrich Berger
Chair of Automation Technology, Brandenburg University of Technology Cottbus, Cottbus, Germany
Keywords: Production Scheme, Knowledge Management, Knowledge Repository, Environmental Impact.
Abstract: The automotive industry is facing challenges due to high mass customization and consequent
decentralization of manufacturing systems. Currently, the evaluation and optimization of eco-efficiency of
production processes is complicated due to time consuming LCA simulations and inexperience of
production planners to make respective decisions. This paper addresses this issue by developing ontology
based intelligent assistance system to support planner in environmental assessment of manufacturing of
customized production in decentralized manufacturing networks as well as decision making in production
planning.
1 INTRODUCTION
The manufacturing industry particularly the
automotive industry is facing challenges to remain
competitive. The increasing product customization
and continuously reducing time to market on the
automotive landscape has compelled manufacturers
to speed up their innovation process as well as
frequent restructuring of value chains. Additionally,
the shifting of manufacturing facilities from high
wage locations to low wage locations, high
outsourcing and frequent collaboration with external
partners has resulted in distribution of
manufacturing facilities. It can be realised by that
fact that currently 75% of the vehicle production and
50% of automotive research and development is
carried out by suppliers (Christensen, 2009).
Furthermore, the lack of skilled people to manage
change and enable customized production in
manufacturing facilities has put the complete
innovation process at risk. The decision concerning
selection of optimal production schemes for
customized products is quite complicated as it
requires optimization of manufacturing processes
not only based on conventional key performance
indicators such as cost, time, quality, flexibility but
also on potential environmental impact. Unlike the
conventional key performance indicators, the
knowledge concerning environmental impact as well
as assessment is not well versed and intrepretable for
machines and users to be taken as decision support
system. In addition to this fact, reducing number of
skilled workers, inexperienced planners to take
decisions based on eco-efficiency and the time
consuming environmental simulations demand
leveraging of exploiting concepts and innovative
methods related to knowledge management.
Focussing on the scope of this paper, it must be
added that the increasing number of activities in a
decentralized production environment results in
increasing environmental impact in terms of several
emissions and ecological wastes. The new
environmental directives and regulations needs new
metrics, advanced assessment methods and efficient
environmental assessment tools have limited the
potential environment assessment in conducting
holistic, reliable, quick and precise assessment.
Furthermore, the addition or removal of certain
environmental terms according to the new directives
and changing assessment methods and unfamiliarity
with the potential impact of combinations of
materials, process and resources on the
environmental impact of the production processes
has made manufacturers to ignore this factor in
production planning and optimization.
Consequently, the production planning in distributed
environment is not cost efficient any more. A strong
need has been felt for an efficient environmental
information system to support decisions on
environmental impact of the potential production
schemes. In the absence of such system, the planner
must be experienced enough to conduct precise and
331
Ul Haque Minhas S. and Berger U..
Development of Intelligent Assistance System to Support Eco-efficient Planning.
DOI: 10.5220/0004110903310334
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2012), pages 331-334
ISBN: 978-989-8565-30-3
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
efficient assessment. The probability of availability
of such experts is very low as the knowledge about
environmental terms, regulations, assessment
methods as well as the simulation tools is expanding
along with the expansion of manufacturing domain
with new materials, processes and resources such as
machines. Therefore, the planner cannot conduct
assessment without the knowledge assistance system
or tool and therefore be be facilitated with
information or knowledge assistance system to get
the desired information from the knowledge
assistance system upon request. This approach will
enhance productivity in planning.
2 KNOWLEDGE MANAGEMENT
IN DECENTRALIZED
PRODUCTION NETWORK
Production planning and scheduling refers to
activities that deal with selection and sequence of
production processes as well as the optimal
assignment of tasks to manufacturing resources over
a specific time. Several methodologies have been
introduced in the literature to enable computer aided
process planning namely feature based planning
(Cai, 2007; Mokhtar et al., 2007; Berger et al.,
2008), artificial intelligence i. e. neural networks and
genetic algorithms) based planning (Joo, 2005;
Monostori et al., 2000; Venkatesan et al., 2009;
Zhang et al. 1997) and knowledge based approaches
(Wu et al., 2010; Tsai et al., 2010). In a highly
individualized customer demand scenario i. e. one-
of-a-kind production, incremental process planning
has been proposed for extension or modification of
primitive plan incrementally according to the new
product features (Tu et al., 2000). Likewise, agent
based approach is used to enable manufacturing
organizations dynamically and cost effectively
integrate, optimize, configure, simulate and
restructure their manufacturing system as well as
supply networks (Zhang et al., 2006). It must be
added that the scope of optimization in planning is
mainly limited to the traditional key performance
indicators such as cost, time and quality. Mass
customization production scenario has however
made manufacturers focus on flexibility in
optimizing processes. The sebsequent
decentralization of manufacturing systems demand
planning decisions based on environmental impact
of the intended activity or process besides decisions
made on conventional key performance indicators.
From the literature, several approches and
methodolgies have been proposed convering several
distinct areas of planning but no significant
contribution has been made sofar that help accessing
and supporting decision making based on key
performance indicators. The knowledge
management framework for supporting
manufacturing system design is proposed by
Efthymiou et al., 2011a. The proposed framework
encompasses four constituent components of the
knowledge management system. The extended work
to this framework is described in work presented by
Efthymiou et al., 2011b, in which the ontology for
the manufacturing system is defined. The extended
work emphasizes structuring of knowledge to
support evaluation of manufacturing facilities and
activities based on cost, time, quality and flexibility.
There is no contribution made sofar in the scientific
literature that addresses the issues related to eco-
efficient evaluation as well as decisions support
system for optimal selection of production processes
for customized products. Moreover, commercially
available PLM tools as well as lifecycle assessment
tools are incapable of facilitating planners with the
environmental knowledge that support them in
decision making. Therefore, an intelligent assistance
system must be devised as an add-on to existing
planning tools that facilitate planners in make
decisions on eco-efficiency of manufacturing.
3 DEVELOPMENT OF
INTELLIGENT ASSISTANCE
SYSTEM
In this regard, the concept for intelligent assistance
system is described by Minhas et al., 2012. The
production scheme for manufacturing customized
products is retrieved from the commerical planning
tools in XML format. It contains information about
the whole production scheme for manufacturing any
customized product. The scheme consists of several
nodes each specifying the product information, the
process technological information and the resource
information corresponding to the process
technology. The ontology based knowledge
repository is enriched with the already existing
simulated environmental impact results of the
potential production schemes. The information
concerning the input (process node specifications)
and output information (simulated environmental
impact results) for the simulated case are stored
either manually in the knowledge repository by
experts or exchange through the xml file from the
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
332
conventional LCA tool such as SimaPro.
Figure 1: Architecture for web-based reasoner.
There are two possibilities for the user to query
the knowledge repository of the intelligent
assistance system. One way is to query in complete
natural language format. It triggers huge complexity
in processing the natural language sentence and
extracting the context of the query. The second way
is (see Figure 1) to restrict the degree of freedom of
user’s in specifying search. This is done with the
web-based configurator developed in PHP. The
SPARQL query is generated as a result of the
configurator input. It is communicated to the server
through web-service which then navigates the
required information concerning the potential
environmental impact of any desired process node.
The knowledge repository is regarded as multiple
instances of domain ontological concepts. This
ontology is represented in specific platform
independent open source language to maintain
consistency while navigating information as well as
the information exchange. This knowledge
repository is made capable of making conclusions
against the queries due to its inherent reasoning
capabilities. The reasoning part of the knowledge
repository is enriched with production and inference
rules associated with the material and process
environment of decentralized manufacturing.
The outcome of the navigation is the
environmental impact results by matching the same
case stored in the knowledge repository or inferred
from similar cases. The application is capable of
running on the web-based system and a web service
has been created to query the knowledge repository.
The planner before making environmental
assessment sends request to the server; the server
generates the required output in reponse to the
SPARQL query which navigates the information
into the ontology based knowledge repository.
The concepts for manufacturing domain relevant
to the environmental impact are modelled in an open
source ontology editor named as Protégé. At the
process level, material specifications, operation and
relevant resource information formulate the basic
concepts for manufacturing domain (see Figure 2) in
RDF format. Furthermore, semantic web rules are
defined to incorporate reasoning capabilities in the
intelligent assistance system. These rules are
programmed using JENA Engine.
Figure 2: Fundamental Ontology Concepts.
The user can send the request for the search of
environmental impact or emissions of certain
combination of materials. The web service based on
JAVA will generate a query which will navigate
inside ontology for retrieval of the information using
JENA inference engine.
4 SUMMARY
The knowledge management concepts can be
exploited in each of the planning area in the
automotive production networks to achieve quick
and reliable multi-objective decision making. The
artificial intelligent approaches and knowledge bases
implemented using ontologies with reasoning
capabilities, will be helpful in reusing, inferring
from as well as for making quick optimization. In
this paper, these concepts are being exploited
considering optimization of production processes
based on environmental impact assessments. The
addition of environmental impact factor as one of
the key performance indicator needs to be evaluated
first using commercially available software tools.
DevelopmentofIntelligentAssistanceSystemtoSupportEco-efficientPlanning
333
The representation of knowledge about material,
process and resource and its consequent impact of
environment using ontologies together with
reasoning capabilities will help the inexperienced
planners as well the experienced ones to get
knowhow concerning emissions that is anticipated to
be assessed as well as the environmental impact of
combination of potential product, process and
resource. The intelligent assistance tool runs on web
based system to facilitate geographically distributed
planners in building, updating, sharing knowledge.
Moreover, this knowledge helps in conducting
assessements regarding eco-efficiency in
manufacturing of customized products.
ACKNOWLEDGEMENTS
The work reported in this paper is supported by EC
FP7 Factories of the FUTURE Research Project e-
Custom "A Web-based Collaboration System for
Mass Customization" (NMP2-SL-2010-260067), to
configure production systems through web based
collaboration in a decentralized production
environment.
REFERENCES
Berger, U., Kretzschmann, R., Arnold, K. P., Minhas, S.,
2008, Approach for the Development of a Heuristic
Process Planning Tool for Sequencing NC Machining
Operations, Applied Computer Science, 4/2:17-41
Cai, J., 2007, Development of a Reference Feature-Based
Machining Process Planning Data Model for Web-
Enabled Exchange in Extended Enterprise, PhD
Dissertation, Shaker Verlag, Aachen.
Christensen, T. B., 2009, Integration of Environmental
Technology in Modularized Production Systems in the
Automotive Industry, in: Proceedings of Joint Action
on Climate Changes Conference, Denmark.
Efthymiou, K., Alexopoulos, K., Sipsas, P., Mourtzis, D.,
Chryssolouris, G., 2011,(a), Knowledge Management
Framework, Supporting Manufacturing System
Design, Proceedings of 7
th
International Conference
on Digital Enterprise Technology (DET2011), Athens,
28-30 Sep. 2011.
Efthymiou, K., Sipsas, K., Melekos, D., Georgoulias, K.,
Chryssolouris, G., 2011, (b), A Manufacturing
Ontology Following Performance Indicators
Approach, Proceedings of 7
th
International
Conference on Digital Enterprise Technology
(DET2011), Athens, 28-30 Sep. 2011.
Fjällström, S., Säfsten, K., Harlin, U., Stahre, J., 2009,
Information Enabling Production Ramp-Up, Journal
of Manufacturing Technology Management. 20/2:178-
196.
Joo, J., 2005, Neural Network-based Dynamic Planning
Model for Process Parameter Determination,
Proceedings of International Conference on
Computational Intelligence for Modelling, Control
and Automation and International Conference on
Intelligent Agents, Web Technologies and Internet
Commerce, Vienna, 28.-30. Nov., 117-122.
Minhas, S., Juzek, C., Berger, U., 2012, Ontology based
Intelligent Assistance System to Support
Manufacturing Activities in a Distributed
Manufacturing Environment, Proceedings of 45
th
CIRP Conference on Manufacturing Systems, Athens,
16.-18. May, 2012.
Mokhtar, A., Tavakoli-Bina, A, Houshmand, M., 2007,
Approaches and Challenges in Machining Feature
Based Process Planning, Proceedings of 4th
International Conference on Digital Enterprise
Technology.
Monostori, L., Viharos, Z. J., Markos, S., 2000, Satisfying
Various Requirements in Different Levels and Stages
of Machining using One General ANN Based Process
Model, Journal of Materials Processing Technology,
107/1-3:228-235.
Tsai, Y. L., You, C. F., Lin, J. Y., Liu, K. Y., 2010,
Knowledge-based Engineering for Process Planning
and Die Design for Automotive Panels, Computer-
Aided Design & Applications, 7/1:75-87.
Tu, Y., Chu, X., Yang, W., 2000, Computer-Aided
Process Planning in Virtual One-of-a-Kind
Production, Journal Computers in Industry, 41:99-
110.
Venkatesan, D., Kannan, K., Saravanan, R., 2009, A
Genetic Algorithm-Based Artificial Neural Network
Model for the Optimization of Machining Processes,
Neural Computing and Applications, 18/2:135-140.
Wu, M., Li, D., Ji, W., 2010, Knowledge-Based
Reasoning Assembly Process Planning Approach to
Laser Range-Finder, International Conference on
Computer Application and System Modeling
(ICCASM), Taiyuan, 22.-24. Oct., V2-686-V2-690.
Zhang, D. Z., Anosike, A. I., Lim, M. K., Akanle, O. M.,
2006, An Agent-Based Approach for e-Manufacturing
and Supply Chain Integration, Journal of Computers
and Industrial Engineering, 51/2:343-360.
Zhang, F., Zhang, Y. F., Nee, A. Y. C., 1997, Using
Genetic Algorithms in Process Planning for Job Shop
Machining, IEEE Transactions on Evolutionary
Computation, 1/4:278-289.
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
334
