Separating Reality from Fiction: Feasible Production Scheduling Using Finite Capacity Scheduling Systems
Does the continuing use of infinite capacity for scheduling by a high percentage of companies mean that it is still a good idea? The habit of calculating conclusions based on invalid assumptions will yield invalid results; however, many manufacturers have been doing just that for 30 years.
ERP vendors and most industry consultants continue to support the use of infinite capacity scheduling. And why not? ERP software is based on it. This trend is likely to continue until the user community stops accepting infinite scheduling technology.
Material Resource Planning (MRP) the MRP II and, more recently, ERP systems have been the predominant tools for planning and scheduling. MRP systems were originally designed as material management systems; it soon became clear that capacity management would also be required.
The MRP developers implemented a capacity planning approach that assumed infinite capacity. The assumption of infinite capacity implies that any task can be processed as soon as it arrives at the work center. Infinite scheduling continues to be applied in ERP systems even though the infinite approach to planning and scheduling results in long product cycle times, high levels of work in process, unpredictable deliveries, infeasible schedules and numerous other problems.
Today's demands have increased to the point that capacity management calculations produced by infinite scheduling systems are no longer adequate to remain competitive. A logical solution is to interface the ERP material systems with finite capacity scheduling (FCS) systems; together they solve the material and capacity management problem.
What is FCS Scheduling?
To qualify as an FCS system, the system must schedule all tasks through all work centers and never exceed the capacity available at any work center at any moment in time. This statement is inconceivable in the context of infinite capacity.
Manufacturing companies typically produce multiple products that compete for resources at multiple work centers with varying due dates. This scenario creates complex scheduling problems. Many feasible solutions exist and the objective is to determine the best feasible scheduling solution without violating any constraints within the system. The term "feasible" is important. FCS systems result in feasible solutions; infinite scheduling systems never produce feasible solutions. The assumption of infinite capacity deals a fatal blow to scheduling. Regardless of how many scheduling iterations are made, the infinite scheduling method never produce feasible solutions.
Of the FCS vendors commercially offering systems, there are two predominant methods used for calculating the sequence of events. Although the methods and results vary, legitimate FCS scheduling systems all produce feasible solutions. If they do not produce a feasible solution, then they do not qualify in meeting the minimum requirements to be considered FCS. Both 'Event based' and 'Job based' methods significantly improve productivity, increases scheduling stability and predictability over traditional infinite capacity scheduling methods.
Production environments run the gamut from straight forward to very complex. They are typically described as flow shop or job shop. While these designations tend to indicate how the production facility functions, it says nothing about the scheduling difficulty. Scheduling can be viewed as a spectrum with easy scheduling problems at the flow shop end of the spectrum and difficult scheduling problems at the job shop end of the spectrum. Most companies today fit somewhere between the two ends of the spectrum.
Flow shops that have assembly production lines producing identical product over a relatively long time period present little or no scheduling problem. Material availability is the determining factor for production in flow shops. Resource availability is usually fixed. You start the line and produce the product. Examples of this type of production are represented by industries producing the same items continuously. ERP systems perform reasonably well in this environment.
In a Job Shop, orders are very different. Unlike the flow shop, the pattern of progress from work center to work center in a job shop is random. A true job shop has a high degree of variability in routing patterns through the shop. This randomness of the job shop environment creates scheduling problems that require advanced scheduling methods. There are many possible degrees of scheduling difficulty between the two extremes of flow shop and job shop.
Until recently, many manufacturing company's fit the flow shop designation. However, as the trend toward reduced order size and customization has progressed, these distinctions have begun to blur. Traditional flow shop applications are moving toward the job shop end of the spectrum. The main cause of this trend is the demand for shorter cycle times, smaller inventories, small lot sizes and custom production runs. The result is that manufacturers at both ends of the spectrum are demanding advanced scheduling technology. Scheduling has become increasingly more important! Companies that ignore this fact will loose ground competitively.
What Scheduling Isn't
All production systems, regardless of products produced, require material management and capacity management solutions.
Material management systems evaluate what materials will be necessary to meet production demand, and to assure purchasing or procuring the right materials in time for production and assembly of the final product. This is the 'where used' function and it is handled by the material management system. The problem not solved by ERP systems is the 'when used' function, which requires detailed scheduling. The combination of materials and capacity management systems produces effective coordination.
Scheduling is the process of accurately sequencing multiple work orders that are competing for the same resources with varying routings. The objective is to complete jobs as needed in the most efficient manner and meet established objectives.
Scheduling a manufacturing enterprise is time dependent and requires a series of complex decisions to logically sequence work through the resources. MRP and ERP systems use the term Master Production "Schedule" (MPS) to decide what orders to complete by what date. In an attempt to make sure materials and components are available when required for production, a Rough-Cut Capacity Plan is used. One function is to establish a dispatch list, and it is just that -- a rough-cut approximation.
In other words, the Master Production Schedule (MPS) details how many end items will be produced within specified periods of time. This is a plan by which production decisions are made about what products to complete and when they are to be completed. The Master Schedule contains due dates for orders, but it has no detailed schedule for jobs and tasks through work centers.
Another scheduling issue is how to handle unpredictable events, such as unpredictable down time. Unpredictable events cannot be scheduled. ERP systems estimate average down time over the scheduling cycle and reduce overall capacity available by some historical percent in an attempt to account for the down time. This is another illogical adjustment creating additional error for an already inaccurate system.
Though the scheduling tool may and should be used to evaluate scheduling options, a schedule, once finalized, should be a commitment to sequence work through the plant in a specific way for a specified time