A New Design for Open and Scalable Collaboration of Independent Databases in Digitally Connected Enterprises

David Levermore, Gilbert Babin, Cheng Hsu
2010 Journal of the AIS  
Digitally connected enterprises" refers to e-business, global supply chains, and other new business designs of Knowledge Economy; all of which require open and scalable information supply chains across independent enterprises. Connecting proprietarily designed and controlled enterprise databases in these information supply chains is a critical success factor for them. Previous connection designs tend to rely on "hard-coded" regimes, which do not respond well to disruptions (including changes
more » ... failures), and do not afford these enterprises sufficient flexibility to join simultaneously in multiple supply chain regimes and share information for the benefits of all. The paper develops a new design: It combines matchmaking with global database query, and thereby supports independent databases to interoperate and form on-demand information supply chains. The design provides flexible (re-)configuration to decrease the impact of disruption, and proactive control to increase collaboration and information sharing. More broadly, the new results contribute to a new Information System design method for massively extended enterprises, and facilitate new business designs using digital connections at the level of databases. Keywords: information supply chain, information system design, information matching and collaboration, distributed database, digital connections scaling one data management regime (e.g., "drilling through" these databases for global scheduling). This oneness reduces the global transaction cost and cycle time of the extended enterprise of the supply chain. Clearly, the integration regime that achieves this oneness needs to be able to reconfigure its connections and respond to new demands, as the oneness is bound to evolve. In the spirit of [Hevner, et. al. 2004 ], we employ the following supply chain scenario to delineate the above design goal: A manufacturer makes different products to supply multiple primes in different industries (including, e.g., Boeing, Cisco, GE, and Wal-Mart). These products share common raw materials, some common parts, and certain common fabrication facilities. All data are controlled under the same enterprise resource planning systems throughout their production cycle; but they are subject to different (simultaneous) supply chain regimes (e.g., data interchange protocols) imposed respectively by these primes. Each prime is also promoting its own goals of (on-demand) collaboration and information sharing throughout its own supply chain, such as e-engineering for design and global coordination of demand-supply schedules. Each chain is in fact recursive since the prime has its own customers (e.g., the prime defense contractors who subcontract to Boeing) who, in turn, have their customers; and the manufacturer has its own suppliers who have their suppliers, too. The situation goes on until it reaches end users and individual production factor providers at the level of persons. The manufacturer needs to reconcile these differing regimes, configure and reconfigure its enterprise databases' roles in these collaboration relations, and minimize the impact of disruptions due to, e.g., any changes and failures in any parts of these concurrent supply chains. Furthermore, the manufacturer wishes to solicit as many new buyers and select from as many new suppliers as possible, from the global market. In all these cases, it wishes to reap the maximum benefits of shared data resources throughout the extended enterprises to coordinate its production and inventory schedules and reach maximum quality and productivity. Thus, there are numerous potential information supply chains just like there are numerous 5 potential supply chains. An open and scalable design for connecting the manufacturer's enterprise databases to any supply chains is required. The ideal is not yet reality. In practice, supply chains tend to use fixed protocols (or, "workarounds") to connect independent databases. This approach is often associated with asymmetrical business relations, where the dominating primes promote asymmetrical sharing of information to their advantages, such as retrieval of on-demand information from supplier databases. While it may also allow the suppliers (e.g., Warner-Lambert) to gain access to select information at the prime (e.g., Wal-Mart's sales forecasting on Listerine), this approach typically presents major obstacles to the suppliers who are subject to multiple concurrent supply chains -e.g., the manufacturer in the above supply chain scenario. More fundamentally, hard-coded designs by nature do not respond well to disruptions such as connection failures, nor facilitate flexibility dictated by, e.g., shifting demands, evolving requirements, and new technology. In addition, application-based proprietary protocols tend to be intrusive and costly to change. Open technologies such as XML, ebXML, and UDDI help to an extent, but their effectiveness is generally dependent on how standardized these databases are in their design and semantics, since interchanging data is not the same as understanding the data (see, e.g., [Levermore and Hsu 2006] for more analysis). Often, as shown in present B2B practices (see, e.g., Alibaba.com, Ariba.com, and PerfectCommerce.com), only basic file transfer (using, e.g., fixed format) is enabled, rather than database queries; which the new design provides. In general, open and scalable connection of databases supports an enterprise to simultaneously participate, on demand, in many collaboration relations across many supply chains, as the manufacturer in the scenario wishes to do. The ability to offer/sell as well as request/buy random information from all participants benefits all parties involved (e.g., gaining cost benefits from flexible processes [Gebauer and Schober 2007] and accumulated data resources [Hsu and Spohrer 2009] , as well as the previously discussed global coordination). Specific to the research objectives of the paper, the capability of on-demand
doi:10.17705/1jais.00233 fatcat:qasd5rsqmbbd7itt3kbp7eiv6m