Supply Chain Uncertainties; Minimizing the Impact on Operations

Even the best managed supply chains have uncertainties. Suppliers are producers themselves and their operations have unexpected interruptions. No one’s promise dates are achieved with 100% certainty. So your “Just-In-Time” operation is sometimes Not-In-Time.
To minimize the impact of these supply chain uncertainties, most producers create buffer inventories of supplies. But the bigger the buffers, the higher the costs and, of course, these buffer inventories defeat the whole idea of Just-in-time production.
There are other ways to reduce the impacts of the Supply Chain uncertainties. Suppose our operations management approach is designed around the concept of plan-revision rather than plan-generation. That is, we accept the notion that replanning is the norm and so we design our operations management tools tailored accordingly. Such tools would have the following characteristics:

• The ability to determine the scope of the impact of a change to one or more planned activities
•  The ability to revise part of a schedule without regenerating all of it
• The ability to unscheduled all activities that satisfy some specified condition
• The ability to schedule into the holes in an existing schedule
• The ability to find feasible possibilities, that is to find alternatives by using substitute resources
•  The ability to keep a change log to provide a history of the evolution of a schedule

The requirements for a replanning system differ in degree, if not totally, from those of a planning system. But planning – starting with a blank timeline – is just a special case of replanning with a partially populated timeline. So there is no loss in generality or utility if the tools we select to support operations planning and scheduling are designed instead for the replanning process.
And the better we can replan, the more we can minimize the impacts of supply chain uncertainty.
There are technical challenges associated with designing a replanning system. First, we would like to be able to describe each production activity with alternative ways to perform it. For example, we could use any of several equivalent machines, or we could use any number of qualified technicians to perform a step in a production routing. When alternatives are a part of an activity description, these options become available when replanning is necessary. However, once these alternatives are allowed, the combinations of production possibilities become large. This is often called the “combinatorial explosion”. So a good replanning system should allow the alternatives but must manage the combinatorics so that it remains practical.
A second technical challenge in designing a replanning system is to determine the exact scope of the effects of a change. For example, if a process takes longer than expected, and it uses several resources that are also used in later processes, how does the time overrun affect downstream activities? Most planning systems would require us to unscheduled everything downstream and rebuild the schedule from the time of the overrun. However, the impact may not really affect all activities downstream. A good replanning system would be able to find all of the affected activities but none of the unaffected ones. We shouldn’t change anything that we don’t need to. The logic needed is a recursive chase of the resource linkages. While the logic is definable, few planning systems have it. If we wish to minimize the disruption from Supply Chain Uncertainties, we should seek replanning systems that can identify the exact scope of the impact of a change.
These ideas and other pragmatic aspects of a good replanning system are discussed in a new blog at www.SchedulingDoneRight.com.