Practical entries

Operations Management connects decisions about continuing service: what may start, what a case needs next, what the available resources can support, what can be promised, and what must change when results fail. These decisions use different results. A capacity calculation can constrain admission; an incident can change priorities; a service promise can remain unresolved after a feasible schedule has been found.

These selected examples show how to combine the methods around a current difficulty; they are not a catalogue or a boundary of the field. Enter where adequate inputs are available and stop at the result the decision needs. The Table of Contents gives direct access to OPS.1–OPS.22 and their nested methods. The Preface, including its conditions on a whole combination, explains why individually useful operating moves can conflict when they use the same resources, authority or evidence.

You can ask an assisting agent to explain a pattern or comment on your case in ordinary language, without framework jargon.

The numbers in these applications are constructed. Their professional premises are supplied for the example; actual clinical, safety, financial and release decisions need the corresponding qualified inputs.

OPS-PUMPWORKS — Find the next supported controller-service decision

Situation: PumpWorks must continue controller test service and incident response while incomplete inputs, rig access, operator recovery and field-release decisions interact.
Question: What can start and finish within the available window, and which other decision must change if the plan changes?
First useful result or blocker: A readiness and capacity basis for a supported operating plan, with deferred demand and the permission, staffing or funding result still needed.
Start with: OPS.8 - Coordinate Queues and Buffers when the immediate difficulty is ready work and incomplete inputs. Reuse an adequate operating account; recover a missing subject or commitment only if it can change this decision.
Stop or return: Return changed feasibility to admission and affected commitments. A supported test-service plan does not supply field-release evidence or authority.

Keep the same subjects across the needed views. OPS.1 - Identify the Operating System, Commitments, and Flow Units bounds the continuing control-service operation. OPS.3 - Distinguish Operating Subjects, Cases, Queues, Resources, and Records separates a release candidate, incident, rig request and test attempt. OPS.2 - Select Work-Management Views and Coordination Methods can use case, recurring-process and queueing views for their different questions. OPS.4 - Keep Current Operating State Recoverable Across Participants makes the relevant evidence, unresolved conditions and next decisions recoverable to the participants; they need not share one screen.
Establish readiness for the next action. In the PumpWorks application, twelve matters appear on boards but only four are ready test packages. OPS.8 identifies what the test can use now. The test’s future output is not its own prerequisite. If laboratory permission is current and the unresolved safety question concerns only field release, the test may proceed while release remains held. A safety condition that also governs testing changes that conclusion.
Distinguish missing inputs from insufficient usable time. Across forty rig-access hours, the example records twenty testing, four setup, four unavailable and twelve without eligible work. OPS.9 - Diagnose and Treat the Current Operating Constraint uses eligibility and service histories to distinguish late prerequisites from a resource deficit in the required window. The historical starvation does not settle the next window: OPS.10 - Qualify Operating Capacity Under Variability finds six usable rig-hours and four packages needing two consecutive hours each. All four require eight hours and cannot fit.
Make the adverse case and other demands part of the plan. Suppose one planned package requires an immediate two-hour repeat and then passes, while the other planned packages pass first time. Two completions plus that repeat use six hours; three plus the repeat need eight. OPS.12 - Protect Human Conditions in Operating Decisions preserves the operator’s required recovery and incident coverage after hour six. Extra rig access needs qualified relief as well as the resource owner’s decision. The full application also compares advance funding and a later slot; an extra window alone does not make the whole extension feasible.
Use the combined result to decide starts and returns. OPS.19 - Reconcile Simultaneous Operating Work Across Cases and Scales retains the feasible two-package plan, defers the third to its qualified later slot and keeps a competing trial closed in the same six-hour window. OPS.5 - Admit Work and Limit Starts gives each considered item its disposition. OPS.7 - Manage Aging, Urgency, and Service Commitments handles the affected local priority or existing promise under the appropriate authority. OPS.11 - Coordinate Interacting Operating Structures checks that access and evidence still concern the required interval and configuration.

If an extra window, qualified relief and the other required conditions later become available, reopen the affected plan and offer. Keep the useful readiness rule and unchanged evidence. Within an already admitted incident, OPS.6 - Continue Cases and Handle Exceptions can instead apply a sufficient current rule directly; report progression only when the performed work and result support it.

OPS-SERVICE-WINDOW — Offer service that rooms, people and commitments can support together

Situation: A recipient requests additional service. Visible space appears sufficient, but qualified staff, existing commitments, support work and recovery share the same window.
Question: What can the parties credibly agree without moving the shortage or burden out of view?
First useful result or blocker: A supported smaller offer, another feasible resource or timing arrangement, a refusal, or the particular professional or authority result still missing.
Start with: OPS.13 - Align Commitments, Resources, and Service Outcomes, distinguishing the requested target, forecast and actual undertaking.
Stop or return: Obtain the parties’ decision on a changed promise. Internal replanning does not discharge an existing commitment; an unresolved professional condition remains with its responsible practitioner.

Use the resource that actually limits this service. In the constructed hospital application, two rooms each have four available hours, but one qualified team has four usable hours. Supplied clinical and operating assumptions require thirty minutes per routine visit, including turnover, and reserve one team-hour for the stated urgent-case scenario. OPS.10 therefore supports six routine visits in total, not sixteen from the room count.
Subtract what is already promised. Two visits are committed and six additional visits are requested. The supported proposal is four additional routine visits under the stated reserve. Serving all six additional requests needs another qualified arrangement or an agreed later time. Eligibility, clinical priority and protected conditions remain supplied professional decisions.
Make the schedule usable for everyone who supplies it. OPS.12 follows the support team’s other duty and recovery. Where the supplied windows permit two groups of three visits, that arrangement can preserve both service and support duties. If the reserve changes from one to two team-hours, only four routine visits fit: after the two existing commitments, the additional offer falls to two. Return that changed resource result to OPS.13 before confirming the larger offer.
Use an account that retains unfinished obligations. OPS.15 - Build a Decision-Specific Operating Account distinguishes the historical due-request cohort from completed visits. Twenty-seven timely, three late-completed and ten still-open requests give timely service of 27/40 = 67.5%. The completed-only 27/30 = 90% answers another question. Use the due-cohort result for overdue service, while the clinical evidence determines clinical action.

OPS.14 - Relate Throughput, Cash, and Operating Consequences enters when paid cover, displaced service or funding changes the choice. It compares actual differing payments and receipts over the decision horizon. An allocated share of an unchanged room or salary cost is not an avoided payment, and a later receipt cannot fund an earlier expense without an available financing arrangement.

OPS-ACCEPTED-RESULT — Faster generation still leaves the promised result unavailable

Situation: An AI-assisted software operation produces more drafts at lower unit cost, but qualified review, acceptance and deployed-service conditions limit what the recipient can rely on.
Question: Which operating change improves the needed result, rather than merely increasing attempts or moving work to reviewers?
First useful result or blocker: A supported admission or service change, a comparison of feasible alternatives, or a quality response with the evidence needed for continuation.
Start with: OPS.15 to recover the requested result and its population, then OPS.10 when the actual acceptance-resource bound decides the immediate question.
Stop or return: Retain the distinction between drafts, review decisions, accepted changes and authorized deployments. A settled resource bound can answer the current request without an additional model or experiment.

Follow generation to acceptance. In the constructed software application, sixty drafts can be generated per day but qualified staffing supports six individual review decisions. Each accepted output needs its own decision. Ten accepted outputs tomorrow are therefore excluded under the current arrangement. Recent days with four to six accepted outputs do not supply a probability for tomorrow, or a guarantee that all six reviews will accept.
Compare changes at the receiving result. OPS.17 - Compare and Refresh Operations Methods compares limiting starts, obtaining qualified review capacity and changing preparation or acceptance. OPS.12 follows review, rework, interruptions and recovery; OPS.13 uses the resulting capacity and conditions to form a later or narrower offer. If the recipient accepts a review service instead of accepted output, make that changed result explicit.
Keep payment and service consequences together. A separate constructed day in the application has four accepted outputs due. Both alternatives generate sixty charged drafts, use six reviews and pay the same salary of 240. The current alternative pays 60 for drafts and accepts four outputs: total payments 300. The cheaper-generation alternative pays 30 plus 10 for rework and accepts two: total 280. Its 20 payment reduction leaves two due outputs unaccepted. OPS.14 compares that service consequence before treating the lower generation price as an improvement.
Use deployed-service evidence for the deployed-service decision. OPS.18 - Control Operating Quality and Reliability uses a separately defined population of one million eligible service requests with a 99.9% success objective. The permitted failure amount is 1,000; 1,500 unsuccessful requests consume 150% of it. Under the example’s authorized response policy, discretionary feature releases pause while permitted urgent recovery and security work continue. Restart needs the agreed service evidence. Accepted code and a fresh reporting period do not establish restored service.

When recovery, review and testing compete for the same people or environment, OPS.19 uses their separate results to change starts or allocations while preserving the required service. Software assurance, security and release authority remain inputs from their responsible practices.

OPS-PRACTICE-CHANGE — Change the way of working without treating publication as successful use

Situation: A new admission or handover method looks promising, but testing it competes with continuing service, and a local success is being used to justify wider adoption.
Question: What can be retained or changed now, and which later-use claim does the evidence actually support?
First useful result or blocker: A supported local method-use decision or revised candidate account, followed only when needed by a separate conclusion about the named practitioners’ use.
Start with: OPS.17 when alternatives need comparison, or OPS.16 - Decide Whether and How to Change an Operating Method when the question concerns one method or proposed way of working.
Stop or return: Keep a sufficient current result. Treat the proposed way as a candidate when its status as a reusable method is still unresolved; publication, attendance and tool use do not establish independent later application.

Select the changed operating contribution. Compare the proposed action and result with current practice and a simpler sufficient correction. Readiness-based admission, protection of an established constraint and extra capacity solve different problems. Use the constraint diagnosis from OPS.9, retaining an unresolved rival explanation when it could change the treatment.
Establish that the trial can coexist with service. OPS.19 can keep a trial closed when two package completions and a possible repeat already occupy all six rig-hours. OPS.16 uses an actual coexistence result, including incident coverage, support, permission, fallback and the return to continuing service, before relying on a later trial window. A plan to obtain those conditions is not their availability.
Limit the local conclusion to what was done. In the later constructed PumpWorks interval, the already admitted method PW-TestAdmission-v2 is tried on two packages. One is correctly admitted; the other is correctly held for missing permission, while service and recovery remain protected. OPS.16 retains that method for the specified package family and conditions, keeps the earlier method for the unaffected family and leaves provider-unavailability use unresolved. This does not establish general superiority or wider independent use.
Ask a different question when another group must use it. OPS.20 - Deliberately Continue and Change Operations Culture bounds the practitioners, practice and interval. In the six-week example, two coordinators use the method correctly in familiar eligible cases; a third starts a case without required provider support. Preserve the supported familiar use and revise the failed branch. The other named shift and longer retention remain unknown. The current conclusion can finish without a new experiment; a later receiving use may justify a feasible protected replay of the unresolved branch.

OPS-RELEASE-AND-WAIT - Change admission without losing the customer’s wait

Situation: The reported time from internal admission to completion falls after a limit on work in progress, but customers receive their orders no earlier.
Question: What improved, and what waiting did the internal measure leave out?
First useful result or blocker: A comparison that keeps the resource arrangement, release rule and customer’s waiting origin together.
Start with: OPS.11.1 - Construct and Reconcile Operating Models Across Scales for the arrangement being changed, and OPS.15.1 - Derive Operating Quantities from Events for the observations used to compare it.
Stop or return: Keep a sufficient comparison. Return a changed service limit to admission and commitments; investigate further only if a remaining uncertainty can change that decision.

Construct the alternatives on the same operation. Four orders arrive at time zero. Each needs one hour at A and then two hours at B. The two stations have independent resources, work in order and start as soon as permitted; there are no setups, failures or returns. OPS.11.1 keeps these premises and the customer’s completion event while OPS.8 supplies the admission and transfer policy. With individual transfer and all orders admitted at zero, B finishes at hours 3, 5, 7 and 9.
Apply the changed admission rule. With at most two orders admitted but unfinished, admit another immediately when a completion frees a place. Admissions occur at 0, 0, 3 and 5; the same resource calculation gives the same four completions. OPS.10.2 - Construct and Revise a Feasible Deadline Schedule places the operations under that admission rule and the unchanged resource conditions. OPS.10 uses the resulting schedule for its capacity and service question. The last order is still ready at hour 9.
Measure both waits from their own events. OPS.15.1 derives a mean internal residence of four hours and a mean customer wait of six. Across the empty-to-empty nine-hour interval, the corresponding count-time areas are 16 and 24 order-hours. Eight order-hours moved outside the admission boundary. OPS.15 presents that distinction to the decision maker; a lower internal average alone does not establish earlier delivery.
Return a changed physical premise to the plan. If one operator must instead perform every A and B operation without overlap, the four orders require twelve operator-hours. The previous nine-hour plan is infeasible. OPS.11.1 revises the shared-resource account; OPS.10.1 - Construct and Compare Operating Capacity Models derives the resource-time obstruction, and OPS.19 reconciles competing work. A promise to deliver all four by hour 10 needs a changed arrangement or a revised commitment through OPS.13.

The Foundational Thinking Suite Reference develops this constructed example through the mathematical formulation and its computation. A real WIP limit can improve service through reduced interference or rework; that conclusion needs the corresponding mechanism and comparison.

OPS-PROTECT-AND-DEVELOP - Prepare or learn while preserving promised service

Situation: A specialist has finished a first pass and could start new work, but an earlier result may return for correction. Some of the waiting interval could be used for preparation or learning.
Question: Which new work can proceed while the promised service remains feasible, and what counts as improvement?
First useful result or blocker: An operating choice that preserves the required correction window, identifies a useful auxiliary unit and retains the customer’s whole waiting interval.
Start with: OPS.11.1 to recover the work and shared resources; then combine OPS.8.1, OPS.8.2 and the capacity, scheduling and quantity methods for the consequences that affect the choice.
Stop or return: Use a sufficient arrangement. Return an impossible promise to its conditions or OPS.13; return the value of the proposed learning or improvement to HCD, DOCA or ME.

Keep the unfinished service in the operating model. In this constructed case, report X arrived at hour −2 and its first pass finished at zero. At hour one, a separate reviewer either accepts it or supplies a correction requiring two uninterrupted specialist-hours. If correction is needed, its completion supplies the promised result, required by hour three. New session Y needs the same specialist for four uninterrupted hours, is ready at zero and must finish by seven. OPS.11.1 retains the shared specialist and feedback event; OPS.15.1 distinguishes first-pass completion from delivery of the required result.
Construct a response that can use the actual feedback. Two hours of correction and four hours for Y fit within seven available hours, but that total does not settle their placement. OPS.10.2 shows that starting Y at zero would postpone correction until 4–6, too late for X. OPS.8.1 - Choose Work-Release Rules When Earlier Work Can Return preserves the 1–3 correction window. If X is accepted at one, run Y at 1–5; if it returns, correct X at 1–3 and run Y at 3–7. This rule uses information available at the choice and meets both stated requirements.
Use only an auxiliary unit that leaves the protected action ready. Suppose a prepared exercise takes half an hour and its closure and restoration of the specialist’s work context take another quarter-hour. No earlier recall is included in this case, and the required resources remain available. OPS.8.2 - Use Protective Capacity While Keeping It Available fits the unit at 0–0.75, before the feedback at one. A longer indivisible lesson could consume the correction window. HCD.2 still has to supply meaningful practice and feedback; the timing calculation alone establishes no learning gain. DOCA.1/.4 or ME.1/.11 can instead develop a preparation opportunity or method trial with its own full burden.
Measure the promised result and compare the cost. OPS.15.1 retains X’s wait from −2: three hours if accepted at one and five if corrected by three. Y waits five or seven hours from its own arrival. The exercise neither erases that waiting nor creates an additional accepted customer result. OPS.14 can compare its incremental financial consequences when that changes the choice; HCD or ME establishes its intended development contribution.
Reopen the affected relation when the condition changes. Move the review to hour two and retain the two-hour correction and deadline three. Even an idle specialist can finish only at four. OPS.10.1 identifies that finite timing obstruction; holding all new work cannot remove it. Obtain earlier usable feedback, a correction method that actually fits, or a revised commitment through OPS.13. Preserve the unaffected resource and waiting definitions.

OPS-PREPARE-AND-RESPOND - Prepare useful work and keep a response possible

Situation: A requested result needs preparation, several participants and a timely decision. More starts or more status messages are consuming the same attention needed to finish it.
Question: What should be prepared, promised and started, and how soon must a consequential change reach someone who can act?
First useful result or blocker: A feasible service arrangement, including the needed preparation and response, or a specific condition that requires a different offer.
Start with: OPS.1 and OPS.13 for the recipient’s result; use OPS.11.1, OPS.10.2, OPS.4 and OPS.14 to connect the work, response and financial consequences.
Stop or return: Use the arrangement while its assumptions hold. Obtain any missing professional contribution and agree changes to the promise with the responsible parties.

Choose a useful whole result before dividing the work. In the constructed report case, the recipient needs an accepted comparison by hour fourteen. Input recovery, analysis and release contribute to one report. Two teams recording their contributions do not create two accepted reports. OPS.1 fixes this result and its completion event; OPS.11.1 connects the contributions. If a smaller comparison would serve the recipient, OPS.13 can offer it explicitly. Omitting a relation that the decision needs would instead leave an unfinished result.
Prepare what the selected work will need. Recover the input, capabilities and resource windows with OPS.13:4.2. A prerequisite may already be available or have a supported arrival before its consuming operation starts. When an unresolved fact could change the offer, use C.11.DUA to compare a useful inquiry with a conditional offer or accepting the stated uncertainty, where the service permits those choices. Count any selected inquiry or trial as work with its own burden. OPS.5 then distinguishes an offer, an agreed commitment and permission to start. OPS.8.1 updates release choices when earlier work returns, and OPS.8.2 can place preparation in spare time that the planned return permits.
Make the information arrive while a remedy can still work. The report needs its input by ten. At seven, recovery takes two hours after authorization, so authorization is needed by eight. A review at nine loses the opportunity. OPS.13:4.5 constructs the detection, decision and recovery timing; in its example, a response starts by 7.5 and delivery occurs by 13.5. OPS.4 supplies the relevant state to the participant who can decide. If recovery cannot start before ten, the same signal instead leads to a changed remedy or service offer.
Keep the four accounts connected and distinct. The work account counts the accepted report and the operations it requires. The information account follows when the input failure becomes known and usable for a decision. The attention account preserves the analyst’s three uninterrupted hours and includes the coordination and recovery burden; extra messages can disrupt that interval. The financial account follows the payments and receipts changed by the choice. OPS.15.1 supplies the event and interval definitions, OPS.12 the human conditions, and OPS.14 the financial comparison. Sharing a diagram does not give these quantities one conservation law.
Check funding and the next admission against the changed arrangement. Suppose immediate input recovery requires a payment of 120 currency units, only 100 is available, and the report would bring 300 the following day. At the required payment time, available funds fall short by twenty. Obtain that funding in time or change the plan; its eventual positive balance cannot fund an earlier payment. Once recovery occupies a resource, include that occupation before admitting another job. A favorable status color supplies neither that resource nor the money.

OPS-ADVICE-IN-USE - Follow an operating explanation into the user’s work

Situation: Reports and lessons are being produced, but their recipients still cannot choose or adapt the described operating method, or different teams offer incompatible advice.
Question: Is the obstruction in the operating claim, its explanation, the learning arrangement or the later application?
First useful result or blocker: A corrected claim or explanation, usable support for the recipient’s next action, or a clearly bounded learning or transfer question.
Start with: OPS.1 for the service result, then EXD.4, ME.12, HCD.19 or HCD.12 according to the difficulty actually found.
Stop or return: Stop when the intended use is supported. Publishing a correct text, understanding it with help and applying it independently later are different results; pursue the one the service promises.

Follow the recipient’s question. In the admission example, mean internal time is four hours while mean customer waiting remains six. Suppose a reader asks why. EXD.4 uses that question to explain the two starting events: the two-hour average before admission remains part of the customer’s wait. The useful result is a correctly bounded comparison, from which the reader can reject an apparent delivery improvement caused only by moving the counting boundary.
Repair the claim or the explanation where it fails. If a second team’s advice equates arrival-to-delivery time with processing time in this same queued operation, ME.12 locates the contradicted claim and its dependent calculation. If the teams simply measure different intervals, retain their definitions and relate them through OPS.15.1 instead. If the account is correct but the reader cannot recover the omitted waiting interval, EXD.4 supplies the missing relation. HCD.19 distinguishes what the published material supplied from what a teacher later explained, so a successful helped response does not hide a missing explanation in the material.
Test the intended further use. For a claim about transfer, HCD.12 uses a materially unfamiliar condition whose consequence is already understood by the assessor. Sharing a previously independent resource can invalidate the old completion schedule; the reader must reconstruct the relevant occupations with OPS.10.2. A worked variant already shown to that reader tests supported reuse, so choose another unrehearsed condition when independent adaptation is the claim. These constructed examples explain the test; they report no observed learner result.
Coordinate production and receiving support. Count publication when the agreed publication result is delivered. If the service also includes explanation, practice or later application, OPS.11.1 includes their work, resources and completion conditions; OPS.10.1 compares the support demand with available capacity. A growing queue of requests for clarification may call for one material repair under HCD.19 rather than more publications. OPS.14 compares the consequential costs. Preserve a useful completed publication without claiming that it has already changed anyone’s capability.

The long applications add case-specific conditions and returns. Use the common Operations methods whose results the current decision needs, together with the qualified contributions of the relevant professional practice.

OPS-AIMS-AND-PERFORMANCE — Set a meaningful aim and interpret what follows

Situation: A target gives little guidance, a peer ranking hides different work, or a missed aim prompts judgement before the changed conditions and joint result are examined.
Question: What contribution is worth pursuing, how will it be compared, and what should the later result change?
First useful result or blocker: A meaningful aim and next action, or a reasoned operating assessment and authorized response with its actual limits.
Start with: OPS.21 when the aim or comparison is missing; OPS.22 when the result needs interpretation. Use an adequate supplied basis directly.
Stop or return: A target does not authorize a service promise or spending. Return a changed premise to the aim, forecast, provision, commitment or consequence rule it affects.

In the repair cooperative example, twelve known useful opportunities justify considering an ambition of twelve while bookings and warranted further demand support a forecast of eight. OPS.21 constructs that distinction and chooses what evidence would show a useful result. MA.5–6 preserve the outlook; the operating and financial methods expose staffing and cash conditions before OPS.13 establishes promises. When work becomes harder, OPS.13 and OPS.19 obtain an agreed response rather than merely lowering a number. OPS.22 then retains the factual miss, examines difficulty and shared consequences, and distinguishes the useful response from an unsupported team ranking. An effective change in the local review rule makes the adverse forecast usable for action; changing its label alone would leave the conflict intact.