When Adam Smith discussed the production of needles in the mid 1700's ("An Inquiry Into the Nature and Causes of the Wealth of Nations"), manufacturing was typically done by individual craftsmen who performed all the processes themselves. They acquired the process knowledge by passing through an apprenticeship program. The Industrial Revolution would soon change this, but the concept of "information" as an entity would not arise until 200 years later.

Frederick Taylor recognized that the knowledge about a process could be documented and organized, ("Principles of Scientific Management") even if he did not see that knowledge as "information" in the way we do now.

Henry Ford, implementing the concepts of Taylor's Scientific Management, created the assembly line with the idea that operations could be defined and standardized to the point that minimal training would be required for any worker. The place in the line determined the task to be done and the line could be staffed with practically anyone off the streets. The critical information about what was to be done in what sequence was virtually "hard-coded" into the factory floor.

As products became more complex and varied, the rigid assembly line was modified to accommodate the changes. It became the job of the supervisors or shop foremen to direct the activity of the labor force as production requirements changed. These supervisors became the delivery mechanism for the process-directing information.

In today's environment many of our products are customized, and the configuration requirements are delivered to the shop floor via computer systems. The process knowledge is no longer built into the assembly line; it is dynamic data that has its own delivery mechanism. Without this information, the manufacturing process would stop, regardless of the availability of materials and labor. And yet that information and its delivery systems are considered overhead, a necessary evil whose costs are to be minimized. The costs are not tied directly to the company's products. How did this situation come about?

Computing first found corporate life through accounting and has been viewed, like accounting, as an overhead function, providing a trail of information to document what has already occurred. That use has changed over time but it has changed so gradually that no one has taken conscious note. Computing systems now communicate what will be done in an organization, from product design and order processing through JIT, process scheduling and upgrading of delivered products. The information being delivered actually directs the operations of the organization. The information delivery systems constitute a control mechanism that can dynamically reconfigure the manufacturing process in real time.

The importance of computer-delivered information can be gauged by the reactions of our users. We are being asked to deliver more systems, not fewer; support more complex processes, not simpler ones; provide quicker response time, not slower. They see these systems as mission-critical, not "frills." The fact that users are concerned about response time, let alone availability, gives further credence to this view.

How often have we heard that the slow response time from a system is jeopardizing the success of a project or new corporate initiative. Is it logical to claim that slow response on an overhead item could have that much impact?

I propose that information systems be classified as either directive or non-directive. A directive system is one which is used to direct the processes or activities in a proactive sense, identifying what is to be done. A non-directive system is one that tracks activity or resources and reports on it after the fact. While I believe that information is as much a "basic ingredient" as raw material and labor, we don't currently have a method of quantifying information in a way that would allow us to treat it as a direct cost. A more realistic proposal is to classify directive systems as factory overhead, while non-directive systems would still be classed as pure overhead. When information from a non-directive system is used as direct input to a directive system, the non-directive system should be reclassified as directive. The information flow is the determining factor.

In addition, the value of directive systems could be depreciated over the expected life of the system in the same way that machinery costs are depreciated. This would provide visibility to the fact that information systems have a finite life expectancy and the company must plan for their maintenance and eventual replacement.

As an aside, identifying the life expectancy as part of the cost-benefit analysis would also provide valuable insights into the viability of a project and encourage long-term thinking. A more dynamic environment would dictate a shorter design/build cycle in order to maximize the window of opportunity. It would be inappropriate to spend three years developing a system when the business environment life cycle in that area is only five years.

However, the challenges we face are constantly increasing. We are moving into an age of customized products from an age of standardized ones. Market forces are demanding quicker response and more flexibility. These challenges can only be met via directive information systems, but we must be able to accurately associate the costs of these systems with our production environments.