Enterprise Systems Planning 

There is an information revolution going on in the business world currently. The crux of that revolution is that control of information systems is passing into the hands of the users of the information. For more than 35 years, business information processing and control belonged to MIS departments in their "glass houses." Two advances in technology in the past 10 years have caused the balance of control to shift. The first was the development of the personal computer (PC). The second was the concept of shared data. Today we call this the client-server environment that connects many users via local communication networks. This new environment has provided users with more control over what goes in and comes out of their systems than ever before. As with all advances in technology there is "no free lunch." No sooner had PCs and local area networks begun to proliferate than new information system problems appeared. Data were being kept in more than one place and were inconsistent; data protection standards were weak or missing; there was little validation of the quality of the data that was being entered; and data were now fragmented over many host machines.

These problems began to cause serious data integrity problems as well as causing management to lose sight of the big picture because the parts were not easily assembled to reflect the needed view. The choices were clear: stop the expansion of local control and return to centralized MIS shops or find a way to maintain

control over the disparate pieces of the business information systems and provide management with what is needed for strategic decision making.

If we think of traditional mainframe and mini database environments, we think of having all the applications in one place. That's a comforting thought. We also think of everyone trying to access those data, often at the same time, and that can effect system performance. Also, if the system develops a fault and goes down, all the users are cut off from using the database. Not a comforting thought (see Figure 1). Then there is the other issue of two distinct databases that need to communicate with each other. As an example we might think of the situation where design engineering has constructed the bill of material and the drawings. And those data reside in the Engineering database on machine X. In the same company we have the material planners who need that data to develop material plans. They are on machine Y. The problem and solution seem obvious, but they're not.

It's not just that it's double the effort to reenter data from another system into ours. We still must deal with the accuracy problem of the second data entry, and the time lag between the synchronization of the two databases. Also, if we are transferring data electronically, is it timely, is it secured, and is it accurate too? We can extend this example to Purchasing and Accounts Payable, Order Entry and Master Scheduling, and of course, Production and Cost Accounting. Often, all these pairs of functions are on different machines and databases. In the past we have hedged these problems by duplicating certain portions of one database in the other. One example we are all familiar with is having two vendor files. One for Purchasing and one for Accounts Payable.

We are well aware of the issue of trying to keep those two databases in sync with each other.

Information system strategists began to explore alternatives. There were several pieces to the puzzle that needed to fit together in some sort of integrative fashion if progress were to be made. The major ones were:

• a database that could keep track of itself when it resided on more than one host.

• a database that would eliminate redundant data so that a piece of information only appeared once.

• a set of tools that permitted both users and MIS to maintain and manipulate the data in almost any manner they chose.

• a method to enforce organizational policy and procedures consistently with respect to business transactions.

How does the client-server environment overcome these issues, and how does it lead to this new enterprise view?

The recipe: take one relational database, a user/operating system interface, several display terminals, good application software and local storage for each user. When you put all these compo­nents together, it sounds and looks something like that entity we call client-server.

To be Continued

Lean Six Sigma Training   Thinking Out of the Box   
Balanced Scorecards  Strategic Tactical Planning  
Supply Chain Inventory Management  Total Quality Management Principles
Lean Manufacturing Implementation  Lean Manufacturing Principles