OPS.Reference:1.3 - Distinguish the grounds of operating knowledge

Operating knowledge joins professional knowledge of what actions can obtain a result with knowledge of their combined behavior over time. Observations establish what happened under particular conditions. Models expose consequences of selected relations and assumptions. Controlled comparisons, experiments and other suitable inquiries can test an explanation or proposed improvement. Commitments and professional rules establish obligations and permissible actions through their own authority. Each contribution can inform a decision without supplying the others.

Some operating conclusions follow from identities or bounds. For a fixed population with well-defined entry and exit, the number present at the end equals the number present at the start plus entries minus exits. If items split, join, are rejected or re-enter, the account must represent those events under the chosen identity rule. A resource cannot supply more occupied time than its available interval permits. Such relations can disprove a proposed arrangement before detailed scheduling. They cannot establish that the permitted work will be performed correctly, that the recipient will accept it or that a numerical objective is worth pursuing.

Other conclusions depend on a behavioral or statistical model. A service-time distribution, failure model or forecast must fit the work class and conditions to which it is applied. The long-run relation commonly written as Little’s law connects consistent averages; a finite reporting interval with unfinished work needs its own boundary treatment. A mean load can conceal a peak that defeats a response promise. OPS.10 and OPS.15.1 develop the appropriate capacity, event and interval reasoning. Choose the representation from the question: an arithmetic bound can settle an impossibility, a schedule can exhibit one feasible order, and a stochastic model may be needed for a probability of timely service.

Practical knowledge grows when observations challenge these grounds. An unexpected wait can expose an omitted approval, a shared operator or a changed demand mix. A disappointing trial can reveal an unsuitable method, a failure to perform it, or a change in the environment. First recover which claim the evidence can change. Correcting the event account, changing an allocation, developing another method and revising an explanatory model are different responses. OPS.15, OPS.16, OPS.18 and OPS.22 give those questions separate treatment. Their results return to the operating account that relied on the defeated premise.

Operations Management connects explanations of operating behavior with reusable ways of arranging and managing production or service. Its knowledge explains why a way can work and where it can fail; its methods make that knowledge usable in practice; the results of practice can revise either. Replacing a queueing approximation or a dispatch policy can preserve that field while changing an important part of its knowledge or practice.