Functional characterization of supply and demand for automatic responses of ERP systems to customer inquiries.

Teich, Tobias ; Mildenberger, Udo ; Richter, Matthias 等

Abstract: The following article presents an overview of a newly-created scientific-methodical approach for automatic comparison of customer's most individual demand with the present supply of applicable capabilities and resources in a virtual organization. This is part of an interdisciplinary project at the University of Applied Sciences in Zwickau where different institutions are involved to cover relevant fields of application. In cooperation with the Institute of Management and Information (IMI), the Institute for Production Technique (IfP), the section of machine construction as well as the department of electronic techniques, the authors are devising methods for functional information retrievals on resources to cope with higher efforts on working in supply chains.

Key words: Supply Web, ERP, Evolutionary Algorithms, XML, functional information retrieval

1. INTRODUCTION

As a result of globalization companies need to face harder competition. If they are located in labor intensive countries, it is even more difficult to get ahead from other competitors. Assignments depend more and more, next to the feasibility, on the reaction time to customer requests. Reducing some barriers in harshly competitive businesses, one solution can be to work in a virtual organization as a part of a temporary supply chain. Some critical aspects of such virtual structures can be settled in legal framework or in unfair behavior of single actors, e.g. maximizing individual benefit. In conjunction with a reticulately expanded chain of partners, it is essential to prove terms of technological and economical issues to establish influential criteria. That means, for gaining partnerships you need to know contemporarily which one can produce a specific piece of customer's demand to evaluated conditions. In conclusion to lower response time, an automatic comparison of company's supply and customer's demand needs to be developed. The authors intend to give a brief overview of an approach that is capable to attend this issue. The key elements are functional descriptions of both sides: supply and demand, based on business's capacity and initiated by the customer, making machine-aided processing feasible.

2. PRESENT SITUATION

2.1 Relevant production types

The ways to generate added value by producing an individual solution for a customer or partner in a virtual organization are described in three order related production types:

* Single item production

* Serial production

* Mass customization

Single item production seems to be the sole possibility for conforming to all requirements of a supply web, but has a clear disadvantage in costs due to a longer lasting response on specific queries. Manufacturing large restricted amounts of the same type of product with minimal variances is represented by serial production. Isolated implementing this production type can lower costs dramatically by economies of scale, nevertheless it doesn't match exactly the requirements of an order related type of fabrication. As an interim solution combining both ways, mass customization was developed. Ideally, it ought to merge cost efficiency and individual customizing a product. Mass Customization however is not able to accomplish all needs of a supply web with meshed relationships in-between.

2.2 Matching of supply and demand

Basically, there are two ways to deal with a partner's product related inquiry. One is to compare the existing portfolio of product data in an ERP system with criteria of a potential customer. If it bears a close resemblance to present products, the alternative with highest similarity will be offered, similar to the concept of mass customization.

Another possibility of replying to a partner's request is to consolidate specific knowledge from responsible professionals of production, controlling and logistics department, often spread across over several locations. After extensive communication and coordination processes, they decide about feasibility, calculated prices, terms of delivery and payment. Altogether these procedures of treating with a product request are not able to cope with the noble planning goal in virtual organizations: Shrinking the time between an incoming demand and the beginning of production drastically (Davidow & Malone, 1993).

3. DESCRIPTION OF THE NEW APPROACH

The following presentation of the approach, described by the authors, consists of seven sequential steps which are concerted to each other (figure 1).

[FIGURE 1 OMITTED]

3.1 XML schemata to describe customer request

A customer request consists of customer data such as name and address of the company, a CAD description of the wanted product and supply data, e.g. price, quantity and physical description required by the customer.

The first step is to transform these dates into a communication language, in this case XML (Extensible Markup Language). For validating this description the authors have developed a new metadata model characterized by a XML schema.

XML, as a meta language to describe documents in the WWW (World Wide Web) or as a communication standard of heterogeneous IT structures, is used to describe the vectors for supply and demand. A detailed description and specification of XML can be found on the webpage of the W3C (World Wide Web Consortium), but shouldn't be subject of this paper (W3C 2007).

XML schema, as a way to describe a pattern for XML document structures, is developed by the authors to describe geometry as well as customer's additional needs. As a newer standard, schemata support more data types than anterior DTDs' (Document Type Definitions) and are completely written in XML syntax. It can be generated from an object oriented UML data model and in a second step it validates database originated XML files. For example, it can make sure that generated CAD drawings are grammatical correctly described in XML (Bormann & Bormann, 2002).

This schema constitutes a generative grammar to check a document containing a specific request with validation rules according to XML. Substantial background for developing these rules is a functional description of resources and capabilities in manufacturing companies that allows comparing them with each other.

3.2 Functional description of company resources

After specifying customer's request, available resources and capabilities of a company are represented in an innovative manner--the functional description of operating facilities and equipments in a virtual organization.

Relevant elements are the human resources for building process, materials and substances, utilized space of machines and technical constructions as well as instruments and manufacturing techniques. The main focus underlies modeling the characterization of machines and instruments that influences specific product related criteria by manufacturing methods' parameters.

For abstract efficient modeling, all kinds of processing instruments are separated into four groups. The first one is characterized by their capability of changing geometrical and technological attributes of materials. This makes it affordable to determine process steps, associated with their order in production sequence.

Second group consists of assembling machines and instruments which put parts together. They are clearly specified with geometric shapes or features that can be connected additionally by process values.

Third cluster of resources illustrates the transportation and handling capability. Material positions changes with defining elapsed time and speed of movement.

Last group contains stock equipment to provide and allocate material in boxes or temporary stock racks.

3.3 Generating of production plans and process variants

Hence, the logical conclusion of describing customer's request and modeling resources is now to generate routings and may create many variants to build a product. For example, a weld between two parts can be materialized by a laser or common technology. For this reason process variants are created, so that technical potential can be tapped in with the given resources.

3.4 Comparison of supply and demand

Comparing functional supply and demand is the most important step after representing the technical environment. When facing one inquiry to different supply vectors, modern algorithms for information retrieval will help to match them. The perfection would be a supply that completely applies to the individual demand.

The output of this step is a feasibility study that shows the gradual fulfillment of the request in technological, geometrical and physical aspects. This listing will give a first impression about ecological and economical factors and is the successor for the next step.

3.5 Permanent evolutionary scheduling

After analyzing technological feasibility, production costs and delivery date must be calculated for decreasing the amount of variants. Each will be integrated simulatively into an environment emulated in a MRP system. This step tends to use genetic algorithms and evolutionary strategies for realizing resource planning and machine scheduling.

3.6 Evaluation of process variants

Remaining process variants will be measured by a matrix of economical requirements, restricted by the request as well as company's economical restrictions. The reticular structure of process variant plans requires ant colony algorithms optimizing multi-criteria conditions. Another effect can occur by a cross-company optimization of the supply chain that leads to characteristic transparency of virtual organizations.

3.7 Generating of final supply

Finally, the customer as a virtual partner in the supply web gets a presentation of solutions which vary by price, date and delivery conditions as well as specific product implementations. Additionally, a final offer includes different kinds of realizations from various partners. Completing with an overview of negotiation strategies, the choice is made easier. As already mentioned human behavior could have negative effects to the transparency of assignments.

4. CONCLUSION

The main focus of the project constitutes generating automatic responses to customer inquiries, based on functional descriptions of supply and demand. Quite contrary to mass customization, the newly developed approach is able to fulfill the trend of pure individualization considering technological potential and business environment of organizations.

The usage of modern algorithms supports contemporary realization of customer requests and allows multi-criteria evaluation of alternative solutions. After evaluating single steps, the project team is going to integrate the approach into an existing IT environment. The consequence of this project will lead to a significant reduction of response time between customer's request and organization's response.

5. REFERENCES

Bormann, U. & Bormann, C., (2002). Concepts Content-Representation & Markup-Languages. SPC TEIA Lehrbuch-Verlag GmbH, Berlin, p. 372, ISBN 978-3935539586.

Davidow, W. & Malone (1993). The virtual organization: The customer as a Co-Producer. Campus-Verlag, Frankfurt/ New York, p. 106, ISBN 978-3593349473

W3C (2007). Official webpage of the w3c for XML. Available from: http://www.w3.org/XML/, last visited Accesed:2007-05-28.