An Analysis of Factors influencing Decentralization in the Context of RFID Systems in Manufacturing

Of all the discussions in information technology, few have been as contradictory as that of information technology and (de)centralization. The ultimate goal has been to determine an appropriate arrangement for information technology resources such as hardware, software, communication equipment, data and people within organizations. With the advent of ubiquitous computing technologies, this long-standing debate has become revitalized. Thus, it was the aim of this dissertation to address this topic by investigating the impact of ubiquitous computing technologies and radio frequency identification technology in particular on the decentralization of data and functions in the manufacturing domain.

To answer the identified research question and its sub-questions, a qualitative research approach was used and eleven cases from seven different industries were investigated. In summary, the following steps went into this research: (1) describing ubiquitous computing applications in manufacturing, specifically those that use decentralized data and functions, (2) identifying the factors that have an influence on the distribution of data and functions in the information system architecture, (3) explaining how the different factors influence the information system architecture by comparing both the centralized and decentralized solution, (4) extracting cost components and benefits from the explanation and building an integrated cost-benefit model.

By analyzing the cases of ubiquitous computing applications in the manufacturing domain, six main factors could be identified as having an influence on the distribution of data and functions: (1) synchronization, (2) external data exchange, (3) security, (4) standardization, (5) flexibility and (6) response time. For each identified factor both the centralized and the decentralized solution was investigated, which resulted in the formulation of theses on the (de)centralization decision and in a cost-benefit model. Based on these findings, thirteen management implications were extracted to help practitioners with the decision of how to design a ubiquitous computing application.