Development of An Expert System For Demand
Development of An Expert System For Demand
Development of An Expert System For Demand
Manufacturing
To cite this article: Esra Kurt Tekez (2018) Development of an expert system for demand
management process, International Journal of Computer Integrated Manufacturing, 31:10, 970-977,
DOI: 10.1080/0951192X.2018.1481297
Article views: 71
ARTICLE
demand information to satisfy the customer’s needs. In this there is no web-based application on customer knowledge.
respect, the success of the manufacturing system, such as that Similarly, Khodakarami and Chan (2014) explored how customer
improved order fulfilment in terms of quantity, speed, lower relationship management systems support the internalisation
inventory levels and increased flexibility, is expected. It is process by providing learning opportunities. They used case
obvious that synchronising demand enables automated sys- study approach and interviews as the main technique for data
tem for evaluating real-time information and hence provides collection. Their results reveal that organisations often do not
mechanism for better production plans and successful manu- make good use of their customer relationship management
facturing system. systems’ capabilities to obtain knowledge from their customers.
The paper continues with literature review. Following that, In addition to these studies, Cheng et al. (2016) proposed archi-
ESs and the developed process for the integrated demand tecture for supply–demand matching based on complex net-
management will be presented. In the next section, an appli- works. However, their study merely consists of road map.
cation of synchronising between demand and supply is pro- Although they present suggestions on how demand manage-
posed and carried out at the office furniture factory. Finally, a ment might be such as task analysis, digital description, data-
conclusion is discussed in the last section. base, artificial intelligence and reasoning, there is no any
application on web-based integration in their study.
Most of the studies in literature as mentioned earlier focus
2. Literature review
on signification of information sharing for synchronisation
By integrating the demand and supply activities, companies can supply and demand and effects on business performance.
create competitive advantage. Therefore, most of the researchers However, the realisation of this integration with web-based
investigated the web-based supply with demand integration and applications still need more attention and research. Although
its influence on supply chain performance. For example, Frohlich the web-based applications have their own advantages, the
and Westbrook (2002) collected random sample from UK manu- literature review shows that there are still few studies on this
facturers and services, and there was strong evidence that subject. Some research along this line is already being carried
demand chain management led to the highest performance in out as summarised below.
manufacturing. Welker, Vaart, and Pieter (2008) examined the Chengalur-Smith, Duchessi, and Gil-Garcia (2012) examined
effect of business conditions on internal and external information information sharing via the web-based supply chain application
sharing using a multi-case study among SMEs (Small Medium known GEOPS (GE Operations), developing questionnaire. It
Enterprise). Their study showed that the role of information and stores data about supply chain participants and their activities
communication technology is also limited in supply links in in a centralised database and makes it readily available to
moderately complex business situations: catalogues are not avail- participants through a web browser. But, this application does
able on the Internet, and timing issues are still largely discussed not consists of individual customers. Moreover, although GEOPS
by phone. is capable of connecting companies in any arrangement, most
Indicating business benefits, Zhou et al. (2014) examined companies prefer to implement just a dyadic connection, limit-
two supply chain practices of sourcing and delivery as well as ing visibility to just their immediate suppliers/customers.
information quality and developed scales measuring them. Similarly, Thimm and Rasmussen (2013) presented a conceptual
This study showed that firms need to align supply chain framework and a rule-based approach for business networks.
practice with the level of their information quality in order to They aim to automate information provisioning to the network
achieve good overall business performance. Additionally, participants. Information provisioning in their research means
Chong and Zhou (2014) expressed that web-based demand that when certain expected events occur, information is auto-
management implementation has also a significant impact on matically provided to subscribers. Informed firms know about
service innovation performance in organisations. Moreover, the resources and competences of other firms too, and may
Gonzálvez-Gallego et al. (2015) studied to test the direct wish to engage them in co-production networks, through alli-
effects and the relationships of information technologies, ances and strategic partnerships. While this study focuses more
employing the hierarchical multiple regression analysis with on inter-firm collaboration, it misses individual customers.
proposed research hypotheses. They used questionnaires for Furthermore, Zhou et al. (2013) presented a method to
data collection and their results show that external and inter- analyse customer requirements from the perspective of infor-
nal information and communication technology capabilities mation. In their proposed method, it benefits from the enter-
and integrated information systems are important drivers of prise web page to offer customers product information. When
firm performance. customer finishes the selection, the enterprise will record their
Furthermore, Pillai and Min (2010) proposed a conceptual options in its database and send feedback to them. The online
model for a firm’s capability to calibrate supply chain knowledge. chat tool on enterprises’ websites provides customers with a
Although this study emphasises determinations on the necessity real-time communication channel. However, it mainly focuses
to calibrate information, how to calibrate the knowledge of on customer-facing functions and neglects manufacturer’
customer demand is not shown. For facilitating the detection capacity to fulfil customer demand.
of market opportunities, Fidel, Schlesinger, and Cervera (2015) Moreover, Saha et al. (2016) developed an ES for customer
regarded customer knowledge management as a strategic order management. This study identifies the attributes and
resource. They emphasised collaborating with customers within their importance to prioritise orders, predict order completion
the innovation process according to the obtained results. But, time and quantify strategic and operational levels risk indices
their research model is established hypotheses; meanwhile, for a capacity-constrained supply chain stage. It provides a
972 E. K. TEKEZ
3. Expert systems
4. The developing process to integrate demand with
ESs or knowledge-based systems are programmes that emu-
manufacturing system
late this decision-making process instead of a human expert.
They are computer programs embodying knowledge about a Effective demand integration provides efficient flow through-
narrow domain for solving problems related to that domain. out the supply chain. For companies, it is important to sustain
Figure 1 illustrates the basic concept of a knowledge-based ES. master production schedule with minimum deformation and
Knowledge acquisition involves the acquisition of knowledge obtain more customers without affecting capacity and utilisa-
from human experts, books, documents, sensors or computer tion negatively. Figure 2 shows the relationship between the
files (Turban, Aronson, and Liang 2005). receipt of the accurate demand knowledge and the manufac-
The knowledge base as seen Figure 1 is the part of an ES turing system’s success. If consistent customer demand infor-
that contains domain knowledge which may be expressed as mation does not flow, the company will not be able to
INTERNATIONAL JOURNAL OF COMPUTER INTEGRATED MANUFACTURING 973
information
Reduce inventory and
Manufacturing relationship man. inventory carrying costs
flow
Joint capabilities
planning
Order fulfillment Reduce manuf. cost
Matching demand due to improved
Customer and supply
relationship planning Increase customer
management Adjustments satisfaction and loyalty
Customized product
on planning development Increase flexibility
respond to differentiated needs. When demand exceeds sup- The domain knowledge is represented in a collection of pro-
ply, it can be either encouraged supply or discouraged duction rules in ‘IF–THEN’ form, and the data about the cur-
demand. It requires understanding the level of demand and rent situation are represented by a set of facts. The inference
matching with manufacturing speed. engine is responsible for comparing each rule in the knowl-
Manufacturing and sourcing plan is developed, identifying edge base with the fact in the database. Note that enterprise
causes of the mismatches and providing better supplier rela- resource planning called ERP system is information system in
tionship as possible solutions (see Figure 2). Because deter- the manner of enterprise wide which contains all data for
mining the requirements back through the supply chain software modules, such as manufacturing, distribution, pur-
enables to realise manufacturing operations in the right chasing, warehouse management and finance. This study pre-
place at the right time. Corrupted information about demand sents web-based interaction with ERP database in the
visibility and factory capabilities can lead to misguided capa- manufacturing company in order to evaluate material and
city plans, missed production schedules, redundant inventory capacity constraints to satisfy the demand. This is so signifi-
investment and loss of customers. Therefore, firms should be cant for indicating the ability of the company to generate
more responsive to their customers providing balance feasible production schedules. Another important aspect of
between demand and supply. Demand integration aids to the proposed structure is customer satisfaction. In this way,
enable consistent planning and fewer inventories, and hence firms can be responsive to fast-changing customer prefer-
reduces time and costs, increase customer satisfaction and ences. Also, by using developed ES, even sales people with
loyalty as well as flexibility. less technical training can field a customer inquiry using the
Due to the reasons explained earlier, companies need to questions built into the system.
synchronise dynamically customer demand and manufactur- Taking into account answers of customers according to
ing capabilities building the right customer relationships. Also, defined rules within the knowledge base in ES as well as
considering successful demand integration is related to infor- database of ERP, it helps users to supply product with the
mation technologies, ES technology is used in this study tak- desired attributes. These attributes are listed in options and
ing into account most manufacturing decisions require expert will act as the guidance for the user. When the selection
knowledge on how to understand and use the data to make process is completed, it is investigated from static product
decisions. Figure 3 demonstrates the proposed structure. The inventory in its database and sends response to the users. If
developed ES evaluates dynamic data which changes accord- there is a need, the manufacturing decision is evaluated for
ing to customer request using sales and manufacturing data. related product. Also, according to the condition of factory
Customers at retailer
Request Response
Manufacturing
Inference Engine Company
No
Customer Yes
Selection of product variant
approve
No
Figure 4. The flow diagram for integrating demand to manufacturing on web-based ES.
capacity, alternative due date to be presented to the customer ● If the given convenient date as alternative due date is
can be generated. Figure 4 illustrates the flow of this for accepted by the customer, the customer order is created
demand management integrated manufacturing. automatically on the ERP database via web-based ES at
The decision process of web-based ES is listed as follows: the distant retailer.
● If the customer does not approve alternative due date,
● The chosen product, features for related product and similar products in inventory according to customer
requested due date are asked to the customer, and a requirements are suggested to the customer.
new session is started for the customer. ● If customer does not approve these suggestions, appro-
● If there is any inventory of that product, the product priate orders and related customers of them are deter-
would be sold to the customer creating shipping order mined for the same product requested by customer in
automatically via web-based ES at the distant retailer. order to investigate alterability of due date.
● If there is no inventory of product, appropriateness of
the requested due date of order is analysed. In that
sense, firstly, it should be investigated whether having
This developed expert system also provides to increase the
sufficient capacity and material to satisfy the requested
utility and benefit gained from ERP. In addition to that, run-
customer requirement such as due date and quantity.
ning different algorithms on database and integration results
This is necessary for the rough cut capacity planning in
of algorithms with ES expands the consideration perspective
the master production scheduling. The condition for
in the proposed system.
checking capacity feasibility is as follows:
X
xsi Ct ; for every t
12It 5. An application of ES for integrating demand and
supply
Let It = {i|Ii = t}, which contains all the orders with due dates
in week t. Each order consumes the capacity in the last period This study is carried out at the office furniture factory. The
that could be used to produce. Let xi be the demand quantity enterprise has two main product groups. One of the groups is
of order i and si the standard time required by a single unit standard products whose specifications are defined by
item of order i. The available capacity limits the expected designers in factory, and the other group is customised pro-
output capability of a work centre or calculated capacity, ducts whose specifications are defined by customer. Of 1400
that is, the demand quantity of order multiplied by the corre- standard products, 328 have inconstant attributes like colour,
sponding standard times per unit. If the required capacity is width and length. So these products have 6276 variants that
greater than the available capacity, this leads to a bottleneck, can be sold. Variants of these products have 51,227 different
that’s for; constraints can take the form of limited capacity. specifications in ERP database. Customised products are
defined when a customer wants a new product and customis-
● In subsequent flow, the statement is evaluated which is ing products causes increment of variants and product speci-
too much demand relative to capacity at a given time. If fications in ERP database. Therefore, there is a need for
factory capabilities are not balanced with demand, alter- mechanisms in order to support decision balancing between
native due date is generated assigning the expected supply chain and sales operations in dynamic environment by
demand to the coming planning period and then integrating ES into the database. The proposed mechanism
asked to the customer for approval. provides to
INTERNATIONAL JOURNAL OF COMPUTER INTEGRATED MANUFACTURING 975
● place shipping order automatically from the factory on ● If there are not enough materials to produce the pro-
ERP through web-based ES when it is needed duct, is lead time feasible for procurement and manu-
● analyse requested due date of the new demand infor- facturing of related materials to begin manufacturing in
mation for feasibility when there is no requested product time?
in inventory
● generate new alternative due date, if the requested due
date is not feasible Figure 6 shows the logic block indicates the production
● place a new order automatically from the factory on ERP knowledge base of ES. The numbers in Figure 6 such as 1 and
through web-based ES when it is needed 0 represent results of database query; 1 and 0 mean yes and
● present similar products in inventory for the same due no, respectively. If the necessary items such as capacity, mate-
date requested by customer rial and lead time are enough, manufacturing of the product is
● determine appropriate orders and related customers feasible and approval block for sales is fired. After the approval
who can be accepted alteration for related order. of the customer, the customer order is created automatically
on ERP database by ES at the distant retailer.
If lead time for supply and manufacturing is not feasible
This paper enables that customer demand is integrated to the when there is no material of the requested product, alterna-
production planning processes in the factory using Exsys Corvid tive due date is generated and asked to the customer for
which is ES shell. Customer orders are evaluated depending on the approval. After the user answers, the results block is fired
schedule of factory by sales staff at the retailer using web-based and order status is seen in the screen. Following this, order
ES. That’s for, ES asks to the user some questions such as product approval block is run. Then, customer accepts or rejects the
type, variant, due date. For example, firstly, it is expected to choose order. If order is accepted by the user, the customer order is
a product as shown in Figure 5. created automatically on ERP database by ES. On the contrary,
After asking due date, a session is created for customer and if the customer does not approve alternative due date, ES is
then logic block for static inventory control is fired. Note that the looked for similar products in inventory in order to present to
knowledge base of the ES is created by using the logic blocks of the customer. If the customer chooses one of the similar
the Exsys Corvid. If there is any inventory of product which products, the results and approval blocks for sales are fired.
customer wants, ES also can notify to user the earlier due date When there is no similar product or the customer does not
as an alternative. Then, after the approval of customer, a ship- approve suggestions of similar products in inventory, it is
ping order is created automatically for sales of product. If there is investigated whether there is a customer who may be appro-
a new customer, it is necessary to add customer information into priate to change for the requested due date of customer
the ERP database before choosing the customer in ES. Hence, the orders. When appropriate customers are found, ES suggests
knowledge of customer is chosen at the last step in ES. the orders of customers which can be swapped for same
If there is no inventory of product, ES considers production products as a list in order to assist sales staff. Because, in
phase and runs production logic block. Production block looks that sense, it also can become important to prioritise among
for the following items: customer orders in order to satisfy the most important custo-
mers and to maximise profits. For this, the sales personnel
● Is there enough materials? would like to negotiate with the related customer about post-
● Is lead time for manufacturing feasible for due date? pone of order having same product for a time period together
● Is there enough capacity? with marketing activities such as a discount or free gift.
In this study, eight logic blocks are created in Exsys Corvid ● suggest some possible customers that can be appropri-
to develop ES. Note that the knowledge base of the Exsys ate for postponing of their order when it is needed.
Corvid is created by using the logical blocks. Every logic block
has its own ‘if–then’ rules and logic blocks are called from
another logic blocks related to case. Generating combinatorial In other words, customer orders are evaluated depending on
‘if–then’ blocks provides more condition for consideration the schedule of factory by sales staff at the retailer using web-
with less definition. based ES. Companies can effectively produce product based
on customer’s need with better demand visibility. Namely,
they can handle their manufacturing operations to achieve
economies of scale, and coordinate inventory replenishment
6. Conclusions to be sustainable. Providing proper use of the resources, the
This paper includes the development of a web-based ES as a significant benefits from the system, such as improved order
prototype integrating ERP database for demand integration. A fulfilment in terms of quantity, speed, lower inventory levels,
smooth two-way flow of information from the customer interac- and increased flexibility and growing revenue, are expected.
tion back into manufacturer is necessary in order to provide This research can be extended in several directions. Firstly,
demand information and visibility of manufacturing requirements this paper limits supply chain as a manufacturer and retailer
such as on-hand inventory, present and future material to meet taking into individual customers; hence, it can be expanded to
the related demand. This study focuses on demand management include inter-firm integration involving suppliers. Secondly,
practices and is carried out at office furniture factory. The pro- this study can be extended by adding prioritisation of custo-
posed ES provides to mer orders into the proposed ES. Finally, future research was
able to develop new innovative products with better under-
standing of customer’s current and future needs and expecta-
● capture the customers’ requirements in a consistent
tions enabling customer insight.
manner,
● compare them against the specifications of product sta-
tic inventory, Disclosure statement
● creating shipping order on ERP from the factory automa-
No potential conflict of interest was reported by the author.
tically at the distant retailer to fulfil the customer’s needs,
● evaluate for deciding to manufacture or not comparing
factory capabilities, References
● determine new alternative due date when it is needed,
Cheng, Y., F. Tao, L. Xu, and D. Zhao. 2016. “Advanced Manufacturing
● creating customer order for new product on ERP of the Systems: Supply–Demand Matching of Manufacturing Resource Based
factory automatically at the distant retailer to fulfil the on Complex Networks and Internet of Things.” Enterprise Information
customer’s needs and Systems 1–18. doi:10.1080/17517575.2016.1183263.
INTERNATIONAL JOURNAL OF COMPUTER INTEGRATED MANUFACTURING 977
Chengalur-Smith, I., P. Duchessi, and J. R. Gil-Garcia. 2012. “Information Negnevitsky, M. 2002. Artificial Intelligence: A Guide to Intelligent Systems.
Sharing and Business Systems Leveraging in Supply Chains: An New York: Addison-Wesley.
Empirical Investigation of One Web-Based Application.” Information Pham, D. T., and P. T. N. Pham. 1999. “Artificial Intelligence in Engineering.”
& Management 49: 58–67. doi:10.1016/j.im.2011.12.001. International Journal of Machine Tools & Manufacture 39: 937–949.
Chong, A. Y.-L., and L. Zhou. 2014. “Demand Chain Management: doi:10.1016/S0890-6955(98)00076-5.
Relationships between External Antecedents, Web-Based Integration Pillai, K. G., and S. Min. 2010. “A Firm’s Capability to Calibrate Supply Chain
and Service Innovation Performance. International Journal of knowledge—Antecedents and Consequences.” Industrial Marketing
Production Economics 154: 48–58. doi:10.1016/j.ijpe.2014.04.005. Management 39: 1365–1375. doi:10.1016/j.indmarman.2010.04.002.
Croxton, K. L., D. M. Lambert, S. J. Garcia-Dastague, and D. S. Rogers. 2002. Ryu, S.-J., T. Tsukishima, and H. Onari. 2009. “A Study on Evaluation of
“Demand Management Process.” The International Journal of Logistics Demand Information-Sharing Methods in Supply Chain.”
Management 3 (2): 51–66. International Journal of Production Economics 120: 162–175.
Exsys Inc. 2010. Exsys Corvid Manual, V5.2.1 Knowledge Automation Expert doi:10.1016/j.ijpe.2008.07.030.
Systems Development Software. www.exsys.com Saha, C., F. Aqlan, S. S. Lam, and W. Boldrin. 2016. “A Decision Support
Fidel, P., W. Schlesinger, and A. Cervera. 2015. “Collaborating to Innovate: System for Real-Time Order Management in a Heterogeneous
Effects on Customer Knowledge Management and Performance.” Journal Production Environment.” Expert Systems with Applications 60: 16–26.
of Business Research 68: 1426–1428. doi:10.1016/j.jbusres.2015.01.026. doi:10.1016/j.eswa.2016.04.035.
Frohlich, M. T., and R. Westbrook. 2002. “Demand Chain Management in Thimm, H., and K. B. Rasmussen. 2013. “Obtaining Informedness in
Manufacturing and Services: Web-Based Integration, Drivers and Collaborative Networks through Automated Information Provisioning
Performance.” Journal of Operations Management 20: 729–745. – A Modelling Framework and Active Database System Approach.”
doi:10.1016/S0272-6963(02)00037-2. International Journal of Computer Integrated Manufacturing 26 (11):
Gonzálvez-Gallego, N., F.-J. Molina-Castillo, P. Soto-Acosta, J. Varajao, and 1054 1065. doi:10.1080/0951192X.2012.
A. Trigo. 2015. “Using Integrated Information Systems in Supply Chain Turban, E. 1995. Decision Support Systems and Expert Systems. USA:
Management.” Enterprise Information Systems 9 (2): 210–232. Prentice Hall International Editions.
doi:10.1080/17517575.2013.879209. Turban, E., J. E. Aronson, and T.-P. Liang. 2005. Decision Support Systems
Jüttner, U., M. Christopher, and S. Baker. 2007. “Demand Chain Management- and Intelligent Systems. 7th ed. USA: Pearson/Prentice Hall.
Integrating Marketing and Supply Chain Management.” Industrial Marketing Vollmann, T. E., W. L. Berry, and D. C. Whybark. 1992. Manufacturing
Management 36: 377–392. doi:10.1016/j.indmarman.2005.10.003. Planning and Control Systems. 3rd ed. Homewood, Illinois: Irwin.
Kadiri, S., B. Grabot, K.-D. Thoben, K. Hribernik, C. Emmanouilidis, G. Cieminski, Welker, G. A., T. Vaart, and D. P. D. Pieter. 2008. “The Influence of Business
and D. Kiritsis. 2016. “Current Trends on ICT Technologies for Enterprise Conditions on Supply Chain Information-Sharing Mechanisms: A Study
Information Systems.” Computers in Industry 79 (C): 14–33. doi:10.1016/j. among Supply Chain Links of SMEs.” International Journal of Production
compind.2015.06.008. Economics 113: 706–720. doi:10.1016/j.ijpe.2007.04.016.
Khodakarami, F., and Y. E. Chan. 2014. “Exploring the Role of Customer Zhou, H., Y. Shou, X. Zhai, L. Li, C. Wood, and X. Wu. 2014. “Supply Chain
Relationship Management (CRM) Systems in Customer Knowledge Practice and Information Quality: A Supply Chain Strategy Study.”
Creation.” Information & Management 51: 27–42. doi:10.1016/j. International Journal of Production Economics 147: 624–633.
im.2013.09.001. doi:10.1016/j.ijpe.2013.08.025.
Lee, H. L., C. S. Kut, and C. S. Tang. 2000. “The Value of Information Sharing Zhou, Z., Z. Xiao, Q. Liu, and Q. Ai. 2013. “An Analytical Approach to
in a Two-Level Supply Chain.” Management Science 46 (5): 626–643. Customer Requirement Information Processing.” Enterprise Information
doi:10.1287/mnsc.46.5.626.12047. Systems 7 (4): 543–557. doi:10.1080/17517575.2012.763189.