- Documenting the Need for New Technology
- Distributed Information Systems Everywhere
- The Global Communication Spaghetti Pot
- Electronic Archeology: Layers upon Layers
- The Gathering Storm of New Activities on the Web
- Global Electronic Trade
- Agile Systems
- Cyber Warfare and the Open Electronic Society
- Summary: Staying Ahead of Chao
1.5 Global Electronic Trade
The Internet is fast becoming the ether by which enterprises transact all forms of business. This is the next stage of Internet development. It goes beyond the electronic retailing and trading that we have all come to know, and love or hate, as the dot-com industry.
Global electronic trade involves collaboration between enterprises to transact business electronically on a scale unforeseen a few years ago. This is the electronic collaboration model of doing business. The stakes are highreduced costs, lower inventory levels, faster time to marketall adding up to increased profitability and increased competitiveness.
The various visions and predictions for how this will happen can all be captured in the idea of the global eCommerce Web, shown in Figure 1.5. It is a grand scheme.
Figure 1.5 The global eCommerce vision
At present special Web sites called trading hubs, or eMarketplaces, are being developed for particular industries, such as retailing, automotive manufacturing, aerospace, financial products and services, telecom provisioning, and the like. Each of these trading hubs provide services that let enterprises in that industry collaborate by integrating various business processesfor example, supply chain processes between buyers of automobile parts and the suppliers.
A trading hub is set up to enable electronic commerce between enterprises in a particular industry. It provides standard formats for electronic trading documents used in an industry, and a range of services to support eCommerce between enterprises in that one industry. Services include, for example, demand forecasting for the industry, inventory management, partner directories, transaction settlement services, and so on. The role of the trading hub is to serve as an integration medium for the electronic management processes of enterprises that wish to collaborate in business.
In electronic commerce, trading documents are contained in messages that drive the steps in a trading process. There are standards for representing the trading documents within the industry. So the processes in each enterprise can easily understand documents from another process and respond intelligently by taking the next steps in a mutual trading process. The current trend is to base these standards on Extensible Markup Language (XML). Various services available on the hub may be supplied by di'erent individual members. Then there are content managers, which are applications such as a distributed master catalog. Each participant manages its own content, and the master catalog keeps a constantly updated view of all the inventories of all participants in the hub. Anyone can get a globally consistent view of the "content" of the hub. Global content management is a critical facilitator to encouraging electronic trading agreements on the hub. The highest-level services are transaction managers of electronic business processes that automate the trading agreements between enterprises. These applications are process workflow engines that automatically initiate the next step in a trading process each time a message completing the previous step is received. Electronic trading across a hub can take many forms, from collaborative integration of individual business processes to auctions and exchanges of goods (electronic barter). The hub provides aggregation and settlement services.
The advantage of trading hubs is reduction in the costs of doing business. For example, procurement costs of a Fortune 100 manufacturer can run annually into billions of dollars. Just-in-time purchasing on the hub saves a significant percentage of these costs.
Now, the grand vision is that the trading hubs for all the various industries will be linked together across the Internet into the global eCommerce Webthat is, a global trading hub of all hubs. It will be event driven, of course. As one visionary puts it:
The traditional linear, one step at a time, supply chain is dead. It will be replaced by parallel, asynchronous, realtime marketplace decision making. Take manufacturing capacity as an example. Enterprises can bid their excess production capacity on the world eCom-merce hub. O'ers to buy capacity trigger requests from the seller for parts bids to suppliers who in turn put out requests to other suppliers, and this whole process will all converge in a matter of minutes.1
To what extent will this vision ever happen? It implies an awful lot of very complex high-level business events flying around, causally dependent upon sets of other high-level events. Will these global trading processes "all converge"? Or will there be a lot of timeouts and thrashing about with no convergence? This kind of real-time marketplace decision making will be driven by events. It is clearly going to need a new technology, beyond master catalogs and workflow engines, that enables us to detect event patterns that signify market situations of interest. These patterns will need to specify sets of business- or trading-level events and their data, timing, and causal relationships. What kinds of "situations of interest"? Here are some examples.
Risk mitigation. Certainly, in the real-time marketplace, automatically recognizing and avoiding risky situations is going to be necessary. For example, an enterprise may want to abort a multistep purchase process in midstream if events happening in the hub's current context signify purchases by other parties that are likely to drive up the overall cost by the time the process has completed.
Policy enforcement. Enterprises will need to maintain strict limitations on real-time trading processes because of the speed with which they could get out of hand. Each enterprise will have its own private policies. Policies could range from constraints on the data in events that are driving the trading process, such as "no parts manufactured in country XYZ," to avoidance of processes that can be indirectly influenced by competitors. The former kind of policy is a constraint on the data in events, while the latter may require detecting patterns of events happening in other trading alliances.
Opportunistic trading. The flip side of risk mitigation is recognizing when trading processes should be speeded up or changed in midstream to take advantage of favorable situations. Again, such situations may often be signified by patterns of events, not directly related to the executing process, but happening in the global trading context.
Regulatory monitoring. This is the same kind of issue as policy enforcement, but on a global scale as agreed to by all participants. Electronic trading will be subject to regulations placing limitations on trading processes, perhaps local to one industry's hub, or perhaps the result of international trade agreements across all hubs. It is well beyond the scope of this book to try predicting what the regulations of the global eCommerce hub might be. But what can be predicted is that the technology to monitor compliance will be event based, will involve recognition of complex patterns of events happening in the total global context, and will need to be flexible enough to be changed frequently without disrupting the monitoring functions.
Real-time marketplace decision making may well become a cybergame of cat-and-mouse, where the advantage goes to those with the best event monitoring and event pattern recognition capabilities. Players in such a game will need to analyze and adjust trading processes frequently. Tools to help analyze multilevel trading processes will need to track causal relationships between events, both horizontally and vertically. So we can expect that analysis tools incorporating causal tracking and complex event pattern detection will be used to monitor the business event layer and the layers below. This kind of monitoring and viewing activity (we will discuss "viewing" later) will need to have the same real-time scalability at higher levels that network-level tools have today.