Engineering.Reference:0.6 - The elegance mantra and reducing repeated firefighting

When engineering or operating work repeatedly needs urgent correction, or a local efficiency gain leaves the receiving work no better off, use FPF C.42, Reduce Avoidable Work While Preserving the Needed Result, for the general method. Its elegance mantra connects the questions below. This section explains their engineering application. Here elegance concerns reducing avoidable work while preserving the needed result and its necessary conditions. Reducing repeated firefighting means changing what recreates a difficulty, alongside any immediate response needed to protect current work.

Keep the needed result in view; remove unneeded work, prevent avoidable rework, distinguish waiting from processing, change the cause worth changing, and carry the change into the next performance.

These questions connect through the result and the effort needed to obtain it. Their order helps explain the reasoning; work can begin at a consequential question whose answer is already needed. Supported partial answers remain useful. An unknown cause can leave one proposed change undecided while another adequately supported response remains available.

Establish what is needed in use. A faster way of producing an unneeded result consumes resources without supplying the missing contribution. FPF A.1.STM:4.1–4.3.1 connects intended use to what must function and how it can be obtained. Removing work requires checking what would be lost, including protection, maintenance or needed knowledge. An experiment that rules out a plausible design can supply such knowledge; its negative result is not by itself avoidable rework. If the need changes, reconsider the work justified by it, with the people who can change the affected commitments.

Prevent the particular rework. A mismatch between the required configuration and the one actually made can force correction, repeated manufacture and renewed checking. SYSE.13 establishes the configuration identity and applicability needed by the decision. OPS.18:5.5 shows the connection: a customer approves proof P2, but the work station supplies P1. The available authorized correction retrieves the approved proof, checks its edition at the station and requires first-article acceptance. Catching P1 before printing protects the order without spoiled booklets. A small check can therefore reduce total work. Compare additional control with the remaining exposure and the work it would add. The full case distinguishes applying this protection from establishing which failure caused the mismatch and from observing recurrence over later orders.

Separate waiting from processing. Elapsed time includes both. Faster processing can leave the receiving result late when it still waits for a ready input, resource or decision. OPS.8 and OPS.9 recover those mechanisms and compare changes against the accepted result. OPS.9:5.4 shows why the deadline for a first result and sustained output can favour different changes. Removing a required acceptance decision, or moving a queue outside the measured boundary, does not establish an improvement for the receiving work.

Choose a change that acts on the loss. Ending an unnecessary operation, preventing a configuration error, changing a handoff and improving processing are different possibilities. A technology change may address several causes at once; it need not wait for every other question to close. Compare the obtainable gain over the relevant period with construction, introduction, support and exception handling, including effort shifted to other participants. Keeping an adequate manual method can be the better choice, as the handover case explains. SYSE.39 develops this comparison for repetitive software-platform work. For a recurring operating method, OPS.16 connects the proposed change to a supported trial; changing responsibility or authority requires the corresponding organizational decision.

Return the change to work and work to the explanation. FPF A.3.1.MI distinguishes the changed method, its introduction, actual performance and later use. Make the applicable instruction and support available, develop missing capability where needed, and examine what happens in the next performance. OCE.10 helps when access, workload or participation conditions explain a difficulty. A recurrence can reopen the configuration account; a new queue can reopen the capacity diagnosis; changed receiving needs can reopen which work is worth retaining. Preserve answers whose grounds still hold. Publishing a revised instruction alone establishes none of these performance results.

This connected account concerns avoidable effort for a needed contribution. Development can instead create new knowledge, purposes and capabilities worth pursuing. The system-thinking account connects a changed target to the wider project; the method-development entry leads to work on the way of obtaining it. Reducing present effort is one possible contribution to that continuation.