Once a company has defined its manufacturing strategy, it has to initiate mechanisms for managing product complexity as well those for managing demand uncertainty in the form of uncertain orders, both perhaps at the same time. The challenge of managing product complexity is to improve productivity (goal of ‘mass production’ strategy) whereas the challenge of market uncertainty is to improve flexibility (goal of ‘mass customization’ strategy) (Meyer 1990).
Product complexity is managed by breaking down a manufacturing task into a number of subtasks and operations. Executing these subtasks and operations in parallel improves productivity. But if different subtasks are performed by different workers (‘division of labour’), then the productivity improvement is restricted by coordination costs, which may otherwise exceed productivity gains. This is also true for expert knowledge, the other resource required to execute operations. Thus, a basic problem in manufacturing is the problem of coordination, which could be stated as:
- After exploding a manufacturing task into thousands of subtasks, how difficult and costly is it to ensure their proper sequencing, scheduling and interaction over a period of time?
- After dividing the task expertise among hundreds of ‘incomplete expert’ workers, how difficult and costly is it to maintain their coordination, motivation and performance?
- As we divide information into millions of tiny bits, hoe difficult and expensive is it to achieve integration, record and update?
That answer to the sequential is that it gets progressively more difficult and more costly.
Therefore, as the complexity and cost of integration and coordination becomes too large, we tend to focus on the question of reintegration. In this context, just-in-time (JIT) efforts aim at the reintegration of physical labour (via flow lines) whereas computer-integrated manufacturing
(CIM) anticipates the reintegration of special expertise organized in functional departments through integrated information process.
Need for Control: For managing task complexity, coordination is required. Likewise to manage market uncertainty, planning and control is required. Management by hierarchical planning and control copes with uncertainties by adaptation to environment and optimization of controller parameters. For instance, a production schedule should be optimized to increase system responsiveness to demand, i.e.
- To keep due dates,
- To reduce total flow-time, and
- To balance factory loads.
This is the planning problem. Against this, the control problem deals with machine-sharing policy, lot is splitting and job sequencing. That is, with:
(i) exploiting resources efficiently, and
(ii) Respecting due dates in the face of uncertainty. In general, the breakdown of long- to shortterm planning decisions indicates levels in the complexity of decisions. This is managed by defining a family of decision problems and generating solutions in a sequential top-down manner.
Fragmented Information infrastructure: Today, the manufacturing industry is still striving for stability of its production system as a major organizational goal. Therefore, in most manufacturing firms, management of change is not yet considered a permanent objective. Whether JIT or CIM, whichever way task coordination is managed, a seamlessly integrated information infrastructure is a must. However, information processing is still very fragmented even in computerized applications (Sahay et al. 1997). Therefore, in many companies, the decisionmaking process is still based on traditional information processing information gathering with ‘paper and pencil’ on request and from inconsistent sources. This process is at its least very time consuming and may yield only insufficient or even unreliable information.
Insufficient Process ability of Available Information: In addition to having a fragmented information infrastructure, most companies are still not organized for fast decision-making processes. Departments are still managed according to their own sub-goals rather than to real enterprise goals. The responsibilities are still structured in one-dimension hierarchies, which mix responsibilities for enterprise assets with those for enterprise operations.