Jump to text
On this page
Copy and source
Published source confirmed at last check

Source changed 2026-11-18 22:51:26 UTC · snapshot created 2026-11-18 22:54:45 UTC · last check 2026-11-19 02:45:10 UTC

SYSE.22 - Coevolve Engineering Problems and System-Family Options

Status: Stable

SYSE.22:0 - Use This When

Use this pattern when an engineering decision concerns alternatives for an actual or proposed product family and a changed condition calls the problem or the options into question. An observation on one unit can change the choice for the family when its shared design, admissible differences or effectivity matter. If the decision concerns only how one system can meet its own requirements, use SYSE.2.

A new operating condition, an effect on another System, a supplier limit, a technology or new evidence can invalidate the old problem–solution pair even while its backlog closes.

Begin with one observation or claim that can change the current engineering decision. Put the selected problem record beside the System option it currently constrains. State the receiving decision and the claim within it that would change—for example, an admission rule, comparison coordinate, affected-System consequence, evidence need, or choice. If none changes, keep the observation without reopening this decision.

The first useful result contains these four parts:

  1. the problem records selected for current attention and a separately retained problem archive;
  2. actual Systems in current configurations and possible-future System-family specifications with their architecture, membership, and effectivity bases;
  3. supported problem–option correspondences and unresolved mismatches; and
  4. one replayable ChoiceResult: choose an option or tie-set, reject the current set, request one worthwhile probe, or refer the question to the decision and authority that can answer it.

An Agent makes the bounded choice under authority. A DecisionSubject states whose choice is represented. Identify the performer and applicable authorization for any later experiment or implementation Work separately; the deciding Agent may also perform that Work.

When project focus, the operating environment or the affected Systems are unclear, use an applicable result from SYSE.1, SYSE.16 or SYSE.17, respectively, or obtain the result you need. Use SYSE.2 when the immediate question is a linked use-and-System concept, and SYSE.6 when the problem and option set are stable and only an architecture choice remains. Use C.18 when the question concerns only an archive or Front. A contested problem formulation may require a Decision Design Method; a changed Method may require Method Engineering.

SYSE.22:0.1 - Terms and Distinctions

Name in this patternWhat it denotes
problem situationThe world-side condition or relation that can frustrate an interest or intended use.
problem recordA ProblemCard episteme describing a problem situation or unresolved claim for one decision and period.
problem archiveRetained problem records, weak signals, alternative formulations, and stepping stones kept for possible later inquiry under a stated retention rule. This is an open content set governed by that rule.
selected problem portfolioThe finite set of problem records receiving attention, resources, or experiments in the current decision horizon.
actual System configurationA world-side System and its obtaining configuration during a stated interval.
possible-future System-family optionA specification of a candidate member, variant, or configuration under a named family architecture, membership rule, and effectivity basis.
System familyA family relation among member Systems or possible-future specifications under explicit membership and effectivity rules. The family is identified through that relation rather than reclassified as another System.
exploration archiveRetained candidate descriptions and lineage kept under a stated exploration-value rule.
FrontA non-dominated set identified for one candidate set, admission rule, comparator or dominance relation, characteristic set, and date.
decision OptionSetThe finite set eligible for consideration in one current C.11 choice under that decision’s inclusion rule. Inclusion for a development or probe decision does not establish admission for operational use. Membership in an archive or Front supplies no automatic membership here.
problem–option correspondenceA supported claim stating how one option changes admission, comparison, consequence, evidence need, or next action for one selected problem.
experiment proposalPossible-future content describing a bounded change and distinguishing observation whose expected decision value can be compared with its cost.

SYSE.22:1 - Problem Frame

Engineering problems and System concepts develop together. For example, an architecture can expose a manufacturing constraint; a prototype can change what users consider possible; an operating failure can reveal an affected System absent from the original scope; a new component can make a previously unaffordable option feasible.

Five recurring confusions make that mutual development hard to manage:

  1. a problem situation is confused with its ProblemCard;
  2. a problem archive is confused with the finite portfolio receiving attention now;
  3. an actual System is confused with a possible-future specification;
  4. an exploration archive or Front is confused with the current decision’s OptionSet; and
  5. an experiment proposal is confused with authorized and performed experiment Work.

Keeping these distinctions lets the project preserve lineage while changing only the claims affected by a new observation.

SYSE.22:2 - Problem

The usual failure is a frozen pair: one backlog becomes the problem and one baseline becomes the solution. Local optimization continues after a relevant condition changes—for example, intended use, environment, a protected characteristic, a supplier condition, or the reachable System family.

Another failure retains many alternatives without their identities and rules. A single portfolio mixes records of unlike kinds—for example, signals, problem formulations, architectures, configurations, experiment proposals, and decisions. A moving Pareto Front then hides changes in its candidate set or comparator and appears to demonstrate improvement across incomparable questions.

Reopening everything after every signal can add work without changing the decision. The project needs dependencies and currentness conditions for individual claims so it can revise a problem without restarting unrelated Work.

SYSE.22:3 - Forces

  • Stable identities preserve evidence and decisions, while stale problem or family boundaries make that continuity misleading.
  • Retained variants preserve stepping stones, while a project needs finite current commitments.
  • Shared family architectures reduce repeated Work, while one environment or affected System can require a distinct branch.
  • Several options can remain non-dominated. For the next project move, decide whether to retain the set, choose a member or select a worthwhile probe.
  • Novelty can open a useful region, while admission, safety, evidence, integration, and service constraints remain.
  • Dependencies support selective reopening, while records that change no decision add maintenance cost.
  • A local gain can move cost, risk, Work, or failure exposure to another affected System.

SYSE.22:4 - Solution

Maintain a finite selected problem portfolio and a separately identified set of System-family options. Connect them through decision-bearing correspondence claims, compare them under the current use and evidence, and end the pass with one replayable choice or probe result. Observations from either side can then reopen the other without discarding archives and lineage. A supported current comparison can remain usable while its carrier ages; neither retention nor an unchanged receiving use requires a fresh trial.

SYSE.22:4.1 - Pattern-Use Unfolding

The Method uses eight recurring moves. They organize production of the account; problem-development, System-development, research, realization, integration, operation, and decision Work can overlap.

  1. Decision boundary. The Agent identifies the actual System or intended System referent selected for engineering, or the System family with its membership and effectivity rules. For a creator, retain the products or classes of future use whose demands and shared resources its options must serve. Current use, environment, affected Systems, configuration, horizon, resources, DecisionSubject and authority complete the boundary. Local realization and acceptance do not require calling the creator a new SoI.
  2. Situation and problem records. The Agent separates the world-side occurrence or condition from assertions, observations, and evidence about it. Each selected ProblemCard states detection, problematic relation or unresolved claim, improvement check, constraints, affected Systems, evidence, receiving use, and the premises that make that evidence applicable. Include an expiry only where a qualification, right, resource or promised-use window ends; a review date alone does not invalidate the claim.
  3. Archive and portfolio. Alternative formulations and weak signals remain in the archive under retention rules. The selected portfolio is finite and states why each record receives current attention or resources and what removes or reopens it.
  4. System-family options. Actual Systems and configurations, candidate architectures, family membership and effectivity rules, and possible-future specifications are identified separately. For each option, state what is realizable now and any builder or platform change it requires.
  5. Problem–option correspondences. Every relied-on pair states the changed admission, comparison coordinate, protected characteristic, affected-System consequence, evidence need, or next action. A pair with no supported decision-bearing difference remains an unresolved mismatch.
  6. Compare eligible options. Apply the guards for the same use and operating conditions. Compare the eligible options on the declared result and resource characteristics, including protected losses and consequences for affected Systems. An eligible rival that is at least as good on every chosen characteristic and better on one excludes the option from a Pareto Front. To place an option on that Front, rule out domination by each eligible rival. A known advantage or a bound can settle a comparison even with another measurement missing. First check that at least one combination of the performance values fits the evidence for the stated System or family and use. Conflicting inputs require a return to their basis before they can support this comparison. If every admissible combination gives a proposal a dominating eligible rival, exclude the proposal. The identity of that rival may still be open. Establish which option to pursue when the receiving decision requires it. Identify any comparison that could change front membership. If the decision uses the order of a particular pair, report any uncertainty in that order even when membership is settled. An archive, a Front and the decision OptionSet each retain their own membership rule.
  7. Choice or probe. The current OptionSet is fixed. The deciding Agent performs comparison-and-choice Work, applies C.11.CRC when a finite configuration-relative contribution comparison is needed, and applies C.11 to produce a decision result: a choice, tie-set, rejection, named probe, or reroute. The Agent selects a probe only when its expected contribution warrants its full burden: participant and engineering effort, delay, foregone alternatives and displaced protective Work as well as direct expenditure. The performer, resources, access, authority and useful observation window must be available. An affordable discriminator is not by itself a worthwhile probe; a useful qualified result may stand without new observations.
  8. Later Work and selective reopening. Assigned Agents plan, authorize, perform, and interpret later Work through the applicable relations. New observations reopen only the problem records, correspondences, options, Fronts, configurations, or decisions that relied on the changed claim.
SYSE.22:4.1.1 - Recover the changed engineering question

Use the eight moves above as a reminder. When a problem formulation or alternatives are missing, the following sections explain how the project can develop them. The intended practitioner can interpret the project’s operating and configuration records and obtain the needed domain, production, service, and assurance judgements. For a needed compressor model, safety limit or manufacturing process, use the result from the practice that establishes it.

Start at the observation that made the old pair doubtful. Recover the earlier use claim and the System concept that was supposed to support it. For one consequential operating situation, follow what happens from the triggering event through the proposed behaviour to the effect on the user or another affected System. Use SYSE.2:4.3–4.6 for that linked construction; use an adequate existing result directly. The return needed here is a use–concept pair, its conditions, any material uncertainty and the connection that is supported or has failed.

Identify what the observation changes. It may contradict the predicted effect, expose an omitted condition, change the value of the effect, or make a new means available. These lead to different engineering questions. For example, repeated network outages can change “minimize purchased peak power using cloud forecasts” into “maintain occupied-building comfort during a demand reduction, including loss of network service.” The new formulation retains the useful outside result and adds a condition that the old means did not meet. It does not make reduced comfort acceptable simply because the incumbent cannot deliver it.

A new means can instead initiate the inquiry: a local controller may make coordinated short demand reductions possible where only fixed schedules were previously affordable. Ask which use would benefit, what change in behaviour could supply that benefit, and which operating condition could defeat it. Keep a technically interesting challenge in the archive until there is a reason to spend current project resources on it. C.40.CD:4.1–4.5 supplies this general reciprocal development of a problem and a way; here the returned question must bind an engineering use, family configurations and realization conditions.

Preserve the obligation or interest behind the earlier formulation. Distinguish correcting a mistaken description, negotiating a different wanted result, and exploring a deliberately relaxed condition. A solution to the relaxed question answers the original only if its result can be returned under the original condition. When interests or authority remain contested, return that specific formulation question to the competent decision participants. Do not hide it as a low score for a design variant.

The result of this inquiry is a revised problem record and a reasoned relation to its predecessor: retained need with corrected conditions, a distinct new need, a superseded premise, or an unresolved disagreement. That relation tells later engineers which earlier evidence still applies.

SYSE.22:4.1.2 - Construct family alternatives with a feasible common basis

Develop alternatives around the disputed connection. Change the allocation of behaviour, the physical or computational means, the interface, or the operating arrangement when that could answer the revised question. An alternative can retain the incumbent, vary a current member, add a family branch, change the shared architecture, replace a previously assumed means, or stop pursuing this use. Creativity need not come from one assigned role: production, service and domain practitioners can expose a better candidate through their own constraints and available means.

For each promising alternative, recover what remains common and what must vary. State the retained architectural commitments, permitted differences and dependencies between differences. Use SYSE.13:4.1 for variant, unit, configuration and effectivity identity. It supplies the distinctions and supported mappings; the engineering work here determines which proposed combinations can meet this problem.

For example, local fallback, a different compressor, a larger processor and a network gateway are four design choices, not four independent switches. The new compressor may require a different output board; the fallback algorithm may fit only the larger processor; that board may lack an applicable test result. Build a candidate by carrying these dependencies together. Reject an incompatible combination, retain a conditional combination for inquiry, or propose the missing interface or realization change. Do not count the Cartesian product of independently feasible features as reachable family variants.

Retain more than one alternative when they answer materially different conditions or preserve a useful future choice. Ask which differences change admission, value, burden or later reuse; do not expand every possible parameter. The family boundary can also change. If an option requires different functioning, support and evidence assumptions from the shared architecture, compare a separate branch or family with extending the common basis. Forced commonality can move repeated adaptation and service burden into every member.

This produces candidate specifications with explicit variation and compatibility rules. It does not produce new actual Systems or establish that current installed units realize the specifications.

SYSE.22:4.1.3 - Test the join between use, variant and means of realization

A correspondence becomes useful when an engineer can follow why the variant could answer the problem. For each relied-on pair, connect the outside result to the proposed behaviour, the required configuration, the means of obtaining it, and the evidence the receiving decision needs. Use the supported realization branch from SYSE.3:4.2–4.4, a platform result from SYSE.12:4.1, and the applicable SYSE.10 evidence assessment where these answer the question. Open only the first unsupported result that can change the decision.

The conditions must hold together. A candidate that fits one board, a test that covers another firmware realization, and a service procedure for a third installation do not combine into one supported family option. Likewise, separate statements that two trials can each use the only rig this week do not make their simultaneous commitment feasible. Compare compatible configurations, conditions and receiving decisions, and check shared capacity where the portfolio relies on simultaneous Work. Use SYSE.20 when overlap or acceptance of a partial result is the engineering question.

Keep two different conclusions visible. An option can deserve development attention while its operational admission remains unresolved. A bounded experiment can be feasible even when its subject variant is not yet admissible for ordinary use. State the scope of admission at the receiving decision; evidence sufficient to select a probe does not become evidence that permits deployment.

Use a failed join constructively. If a variant cannot be tested within the existing arrangement, the result may be a precise builder investment question for SYSE.23. If the proposed behaviour cannot supply the required outside effect, return to the use–concept pair. If every candidate fails a protected condition, reject the set or seek an authorized reformulation; ranking failed options cannot supply a successful one.

SYSE.22:4.1.4 - Let one side revise the other without losing the original need

Run the candidate forward far enough to reveal the consequential mismatch. A model, an existing comparable case, a supplier answer or a performed trial can supply that result at its own evidential strength. Relate it back to the problem before generating another variant.

Result that changes the inquiryEngineering revisionWhat remains to be established
A candidate achieves the stated output but an occupant or service System is worse offCorrect the problem’s consequence or protected condition; revisit dependent alternativesWhether a replacement candidate satisfies the corrected use
A new means makes a previously unattainable use plausibleAdd a use–concept pair and compare its worth with the selected portfolioActual demand, feasibility and admission; technical possibility alone supplies none of them
Different family members require incompatible assumptionsSeparate their applicable conditions, introduce a justified branch, or change the family basisShared resources and cross-member consequences under the proposed arrangement
The target variant is plausible but its construction, evidence or support cannot be obtainedKeep its promise conditional and frame the limiting builder relation for SYSE.23 or the obtaining question for SYSE.24A compatible supported means of obtaining the result
A candidate disproves one interpretation of the problem but leaves another usefulRevise or retire the defeated claim and preserve the supported branchWhat the evidence supports about the remaining branch and any question still open

A better means may change which problem is worth current attention. Conversely, a revised problem may change which retained design is useful. Preserve the decision-bearing reason for that movement. This is how an old stepping stone can become useful again without treating its earlier rejection as an error or its archive membership as current endorsement.

SYSE.22:4.1.5 - Compare complete alternatives and stop at a useful decision

Use one comparison situation per consequential operating profile, including the failure condition that reopened the question. Compare complete candidates under the same profile and assumptions. Keep admission constraints separate from preferences among survivors. If an option obtains its benefit only with an unprovided builder change, include that change and burden in the comparison or keep the benefit conditional.

When profiles favor different members, the useful result can be a family rule assigning different variants to different conditions, rather than one universal winner. Examine the shared consequence: supporting two members may require extra configuration, spare parts, qualification or service Work. The cost of the portfolio is not necessarily the sum of isolated variant estimates; common tools can save work while common scarce resources can create conflicts. Do not claim family superiority from two separate local wins.

A revealing probe is worth selecting only if an obtainable outcome could change this comparison enough to justify its full burden. The current result can also be enough to choose, retain a tie, reject, narrow or stop. Use the existing decision account in C.11.

SYSE.22:4.1.6 - Pass the unresolved engineering question at its actual scope

The receiving work determines the next use of the joint account. A compatible use–concept pair can go directly to architecture Work through SYSE.6; a supported member can proceed through its existing realization arrangement. Use SYSE.23 when the unresolved question is which change in the target or its builders is worth making to enable later changes. Send the expected change family, blocked correspondence, current configuration and realization basis, protected conditions, horizon and relevant uncertainty. The receiver constructs its investment alternatives; the SYSE.22 choice does not select an investment for it.

Use SYSE.24 when the needed contribution is clear and the remaining comparison is how to obtain it. If the question instead concerns wider agreement about the problem, return that question to the relevant problem-structuring or strategy practice.

A builder improvement from SYSE.23 returns as a changed feasibility premise only for the family options it affects. An investment decision can support an expectation of improved feasibility. Stronger reliance needs evidence of the capability or a realization result that demonstrates the improvement. Reconsider the comparison to the extent that the new result warrants.

SYSE.22:4.2 - Record the Result

The joint problem–System-family account records these eight items:

Result positionRequired content
decision boundaryProject System or System family, membership and effectivity basis, current use, environment, affected Systems, configuration, horizon, resources, chooser, authority, and receiving decision.
situation and claim basisWorld-side occurrence, condition, or unresolved possibility; separate assertions, observations, evidence, source, interval, and uncertainty.
problem archive and portfolioRetained and selected problem records, their different membership rules, lineage, resources, applicable premises, stops and reopen conditions; actual expiry where a qualification, right, resource or promised-use window ends.
System-family optionsActual Systems and configurations, family identity, architecture, variant and effectivity rules, possible-future specifications, realizability, and builder dependencies.
plural-search resultsExploration archive and each Front with its candidate set, retention or admission rule, comparator, characteristics, date, and currentness.
correspondencesEvery decision-bearing problem–option claim, its source and uncertainty, plus unresolved mismatches.
choice or probeFixed OptionSet, shared comparison basis, choice rule, probe budget and value when relevant, result permitted by the choice rule, reason, and overturn conditions.
continuationRetained alternatives, planned and performed later Work when it occurs, receiving results, and selective reopen relations.

SYSE.22:4.3 - What Changes in Practice

The project stops treating backlog items and design variants as two fixed columns. It can say which world-side change reopened which problem–option pair, why a candidate remains in an archive but not the current decision, and why the choice rule permits a choice, rejection, probe, or reroute as the next result.

SYSE.22:5 - Worked Case: Heat-Pump Controller Family

In this constructed case, a manufacturer maintains heat-pump controllers for occupied apartment buildings. The current decision concerns the controller family and a reversible building pilot. Installed controllers and a laboratory unit are actual Systems; the options below are possible-future specifications under the family’s architecture and effectivity rules.

The earlier selected problem was peak electrical demand during cold mornings. The preferred specification used a schedule optimizer and cloud forecast service. These three field observations now reopen the comparison:

  1. several buildings lose network service during the coldest periods;
  2. compressor cycling increases service calls in one installed configuration; and
  3. a new tariff rewards short demand reductions but penalizes slow recovery that leaves occupants cold.

The current problem portfolio and System-family options produce four correspondence rows:

Problem recordSystem-family optionDecision-bearing correspondence
peak demand, now with recovery-time and occupied-temperature acceptancecloud schedule specificationNetwork loss makes the option inadmissible without a qualified fallback; recovery time enters comparison.
network-loss continuitylocal-fallback specificationLocal forecasting can change continuity and recovery, but processor-load and fallback evidence are missing.
compressor cycling for the affected configurationvariable-speed-control specification and incumbent controlVariable speed can change cycling and energy use while increasing calibration and service burden.
speculative voice-control request, retained only in the archivevoice-gateway specification, retained only in the option archiveNo current use or affected-System consequence gives this pair decision priority.

The final row explains archive retention and current non-selection; it supplies no OptionSet member. The first three rows are the complete correspondence set used by the current decision.

The earlier Front compared energy use and peak demand. The current comparison uses a new candidate set and eight characteristics: continuity, recovery time, occupied-zone temperature, peak power, compressor cycling, processor load, calibration effort and service burden. Establish eligibility and compare the surviving options on this basis before declaring the new Front.

The fixed current OptionSet has three members:

  1. continue developing the cloud-schedule specification;
  2. develop the local-fallback specification; or
  3. develop the variable-speed-control specification.

Safety and occupied-zone comfort are hard guards. The remaining characteristics are compared by a partial order. The cloud-schedule specification without a qualified fallback fails the network-loss guard. Variable-speed control retains the incumbent local control and is admitted in configurations with the matching output board. Local fallback still lacks evidence about processor load and cold recovery in the two configurations proposed for it.

The controller-family council is the deciding Agent. Its current assignment authorizes it to choose the next engineering probe, or decline new probing on the present basis, and allocate no more than 160 engineering hours. A building operations manager authorizes any occupied-building trial; the product-family owner authorizes later adoption. The probe choice grants neither authority.

The council uses one shared comparison basis:

  • The fixed OptionSet remains cloud schedule, local fallback, and variable-speed control.
  • The current BeliefState contains the three field observations, the hard safety and comfort guards, the cloud option’s network-loss failure, the variable-speed option’s admitted local control, and the missing local-fallback processor-load and recovery evidence.
  • The OutcomeModel specifies which observation changes admission. To qualify local fallback for both proposed configurations, it must pass in each. A demonstrated failure in either excludes that two-configuration proposal. Keep the observation for each configuration, including any narrower use it supports. Compare local fallback and variable-speed control for the same buildings and conditions where both are applicable. The calibration trial changes estimates of variable-speed burden while leaving the local-fallback admission question open.
Candidate next probeBounded burdenDistinguishing observationsEffect on the current decision
Scripted network-loss and cold-recovery trial96 engineering hours, two hardware-in-the-loop bench days, one three-day reversible building-pilot window, and about one week of decision delay.In each of two representative building configurations: no safety violation; occupied-zone recovery inside 20 minutes; and controller processor load below 70%. The outcome is contrary if either configuration has an established safety or comfort guard failure or a processor load at or above 70%. Incomplete, uncertain or otherwise non-decisive observations remain unresolved only when no such failure has been established.A pass in both admits local fallback for the two tested configurations. Where variable-speed control is also applicable, both options can now be compared and their ordering is open. A demonstrated failure rules out the joint proposal. Record which configuration failed and which claims still hold for the other. If the evidence is inconclusive, report which admission question remains open. The council can finish with that result or consider further work when the investment decision warrants it.
Variable-speed calibration trial128 engineering hours, four calibration-rig days, the same single building-pilot allocation, and about two weeks of decision delay.Calibration effort and service burden may fall or rise within the currently supported range; the trial does not observe network-loss recovery or local-fallback processor load.Either bounded outcome refines the burden comparison for variable-speed control within its supported configurations. The local-fallback admission question remains open.
No probeNo immediate trial resource use or delay; preserves the scarce pilot allocation and engineering capacity.No new observation.Retain the qualified current comparison: cloud-only fails the guard, variable-speed control remains admitted in its supported configurations, and admission of local fallback for both proposed configurations remains open. Later development or adoption still belongs to its competent owner.

The ChoiceRule compares obtainable probes with using the evidence already available. The network-loss trial can settle whether the council may include local fallback in the next development comparison. In the first resource situation of this case, the council judges that resolving this question before the next family investment is worth the 96 hours, scarce pilot allocation and one-week delay. Assume that capable performers and the windows can be obtained, and that this allocation does not displace more valuable protective or development Work. Those benefits, costs and conditions support selection within the 160-hour ceiling. The calibration trial cannot resolve that question. The deciding Agent applies C.11 and records ChoiceResult-HPF-1 = probe_again for the network-loss and cold-recovery trial.

In the paired resource situation, the same discriminator would consume the only pilot window needed for already supported protective maintenance, or no competent performer can use it before the investment decision. The council can decline that probe and retain the qualified comparison above. Local fallback stays unresolved; no observation or safety assurance is invented. A later owner can make a supported bounded development choice among admitted options.

In the selected-probe situation, after that ChoiceResult, a planning Agent must prepare the trial plan, the building operations manager must authorize the occupied-building trial, and assigned Agents must perform and interpret the Work. Family adoption remains a separate decision by the product-family owner. If pilot authority is withdrawn before the trial Work, a new decision pass records reroute and the missing authority without rewriting ChoiceResult-HPF-1. For a later observation, reconsider any of the three current problem–option correspondences and family results that depend on what changed. The voice-control archive entry remains unchanged.

For currentness, first keep the same configurations, relied-on observations, calibration, rights and supported use while only the source export date changes: the comparison remains usable. Now change a controller configuration so that the relied-on processor-load evidence no longer covers it, or let a calibration or pilot permission window end. Reopen or suspend the uses that relied on the changed condition. Keep the needed qualification with the comparison for later receivers. The case’s 20-minute and 70% limits are receiving acceptance conditions, not universal thresholds supplied by this pattern. Their protective basis and any proposed amendment require the relevant engineering judgement and authority; merely choosing or declining a probe changes neither.

SYSE.22:5.1 - Construct the family account behind the probe and use its return

The heat-pump case starts with observations, not three fully formed alternatives. To obtain the specifications above, the engineers first retain the outside need for acceptable heating and change the problem from peak reduction alone to peak reduction with continuity and timely recovery. Following SYSE.2:4.3, they trace the cloud command through the connection, local controller and compressor to the occupied-zone effect. The trace exposes the missing behaviour during disconnection. Continuing with only a lower peak-power target would leave that failure intact.

They then vary the means at the failed connection. One proposal adds an independently functioning local mode; another changes compressor control while retaining the existing local protection. The cloud-only proposal is kept in the comparison long enough to show why it fails the revised condition. Voice control does not repair that connection and stays outside the selected portfolio.

For this constructed account, assume the larger processor is available only in two controller configurations, and that the variable-speed compressor needs its matching output board. The option descriptions therefore name those combinations and the effectivity proposal; they do not offer every feature on every installed unit. The local-fallback option’s conditional status is precise: a plausible configuration exists, but the current decision lacks the load and recovery evidence that would permit its admission. The rig and occupied-building window in §5 explain how that uncertainty could be resolved.

Now suppose the authorized trial passes in both configurations with the larger processor. Local fallback qualifies there. Variable-speed control qualifies where its matching output board is available. For a building where both options qualify, a comparison showing that each has an advantage on a different declared characteristic would keep both on the Front for this two-option comparison. The trial alone leaves that ordering open. These results support the family owner’s comparison for the applicable buildings. Release and use on other boards require their own grounds. The problem portfolio still has three selected problems, and the voice-control request remains in the archive.

The next supplier-change inquiry reveals that producing and assuring more combinations with rig R4 requires repeated interface work. The returned SYSE.23 question is specific: for the named controller family and next supplier replacement, which target, rig and evidence changes would reduce that repeated work while preserving the existing comfort and configuration limits? Local-fallback evidence supplies one applicable input; it does not settle the supplier interface or evidence-reuse questions in SYSE.23:5.

Changed condition. Suppose a later building requires the smaller processor, on which the fallback algorithm would exceed the same load limit. The passed trial remains valid for its two configurations. The revised family account keeps the supported larger-processor member where applicable. The same fallback specification fails the smaller-processor condition and is excluded for that use; it is not merely awaiting more evidence. Incumbent local control can retain its previously supported use while a different algorithm or board is considered for the unmet demand. The owner compares a separate branch’s supply, configuration and service burden with replacing the small board. Reclassifying the small-board building as “not a customer” would change the problem scope and require its own grounds and authority; it would not demonstrate that the fallback solves the original use.

SYSE.22 changes which problem–variant pairs are worth pursuing; SYSE.23 can change the means and burden of reaching them. Either result can stand for its own decision while the other remains unresolved.

SYSE.22:6 - Bias Annotation

Watch for these six recurring biases:

Recurring biasLikely driftRepair
requirements-freeze biasThe approved backlog becomes the permanently current problem.Recover the world-side situation and the receiving claim’s applicable premises, including real time limits; reopen only decisions affected by their change.
solution-fixation biasEvery observation becomes a modification of the incumbent design.Reopen both problem formulations and family alternatives; admit a branch, replacement, or stop when supported.
archive-as-portfolio biasEverything worth remembering receives current attention and budget.Keep retention membership separate from current portfolio membership.
one-Front biasPoints from changed candidates or comparators are plotted as one improvement curve.Identify each Front and its basis; make any cross-basis comparison separately.
novelty biasA distant or AI-generated option is treated as useful or admissible by novelty alone.State the reference set, admission rules, consequences, and evidence.
measurement maximalismThe project waits for field-wide prevalence or conclusive causal attribution before a reversible probe.Use proportionate evidence with explicit uncertainty and limits.

SYSE.22:7 - Conformance Checklist

  • The decision boundary identifies the actual System or intended System referent selected for engineering, or the System family; current use, environment, affected Systems, configuration, horizon, chooser and authority. For a creator, it retains the product-use grounds and bounded acceptance.
  • World-side problem situations and ProblemCard epistemes remain separate.
  • The problem archive and finite selected problem portfolio have separate membership rules.
  • Actual Systems and configurations remain separate from possible-future specifications.
  • Every option identifies its family architecture, membership or variant rule, and effectivity basis.
  • Exploration archive, Fronts, and current OptionSet retain their own candidate and admission rules.
  • Every relied-on correspondence changes a decision-bearing claim or remains visibly unresolved.
  • The comparison identifies eligible options, the evidence for each claimed Front membership and any open comparison that could change it. It preserves protected losses and consequences for affected Systems.
  • The choice or probe result states its basis, rule, full burden and value when relevant, actual obtaining conditions, authority, and reopen condition. Retaining a supported result is not conditional on a new probe.
  • Proposal, planning, authorization, performed Work, observation, and later adoption retain separate results.

SYSE.22:8 - Common Failures and Repairs

Recurring failureRepair
Backlog equals problem portfolioRepair the selected ProblemCard values and state why each receives current attention.
Candidate specification equals future SystemKeep the specification possible-future; state realization and later-test Work.
Addresses problem fills every correspondence cellState the changed admission, coordinate, consequence, evidence need, or next action.
Archive equals FrontPreserve archive membership under its retention rule and Front membership under its comparator.
A changed Front is presented as improvement on the old FrontIdentify both bases and make any cross-basis claim separately.
A probe choice is presented as family adoptionIdentify later planning, authorization, Work, evidence, and adoption decision.
Every new source restarts the projectTrace the source claim to its receiving use and reopen only dependent results.

SYSE.22:9 - Consequences

The project can change both what it is solving and what Systems it considers without discarding lineage. Useful stepping stones remain recoverable, stale comparisons become visible, and one bounded probe can be chosen while the problem and System family remain revisable.

The cost is maintenance of identities and decision-bearing correspondences. Some correspondences may remain unresolved. The best current result may be a probe, rejection, or referral to another decision.

SYSE.22:10 - Rationale

FPF supplies problem records, plural search, archives, Fronts, currentness, comparison, and choice. Systems Engineering adds the action-changing specialization: selected problem formulations must be compared with architecture- and configuration-identified System-family options under project use, operational environment, affected-System consequences, realization limits, and engineering evidence.

The terms problem factory and solution factory describe different but corresponding Work and Method structures. The described Work can take place concurrently. Identify the Systems involved separately; use SYSE.20 when overlap or required order becomes the engineering problem.

SYSE.22:11 - SoTA and Source Use

Engineering can revise problem formulations and System-family options together, including the Work and Methods used to develop them. Problem archives and portfolios retain material for later choices; comparison and acceptance use their stated bases. When the way of developing a System becomes the obstacle, the same inquiry can extend to its builders.

NASA’s Design Solution Definition guidance, §4.4.1.2 connects iterative concept development, operating concepts, stakeholder expectations, technology maturity and enabling products. Use a sufficient result from that practice directly. This pattern’s additional work is the engineering connection between a changing problem portfolio and configuration-qualified family alternatives: explaining which combination is reachable, which use it can answer, and which changed premise reopens the comparison.

A competent product-line practice can provide a variability model and a way of producing its members. For example, the SEI production-plan guidance describes how core assets are used to build a particular product. Retain that supported basis rather than rebuilding it here. The remaining question in this pattern is whether changed use and realization conditions require a different family option, problem formulation or current commitment.

SourceRetained contributionUse boundary
Dorst and Cross (2001), Creativity in the design process: co-evolution of problem–solutionMutual development of problem and solution spaces in protocol studies of experienced industrial designers.Use as bounded design-process evidence; establish the current System family and configuration separately.
Liker et al. (1996) and Sobek, Ward, and Liker (1999) on set-based concurrent engineeringCommunication about design sets, delayed commitment, feasibility, and narrowing in automotive product development.Transfer the set-based moves only where the receiving profile’s constraints and evidence support them.
Castle, Stock, and Gorochowski (2024), Engineering is evolutionVariation, expression, evaluation, selection, exploration, and exploitation as an engineering perspective.Use the evolutionary analogy as a hypothesis source; ground the current engineering Method and cultural claims separately.
Zhang et al. (2025), Darwin Gödel MachineAn AI-engineering case retaining a lineage-bearing archive of coding-agent variants for later improvement.Treat coding benchmarks as one application case; test transfer to physical Systems and causal claims separately.

Reconsider how you use a source when changed evidence alters its practical contribution or the scope of the conclusions you can draw from it.

SYSE.22:12 - Relations

  • SYSE.1, SYSE.16, and SYSE.17 supply project focus, operational environment, and affected-System consequences for compatible uses.
  • SYSE.2, SYSE.6, SYSE.7, SYSE.10, and SYSE.13 supply linked concepts, architecture decisions, descriptions, evidence, and family/configuration/effectivity identity.
  • SYSE.15 can supply a compatible engineering-Method repertoire. SYSE.21 can supply bounded observations of an enacted variant and its population; each claim keeps its own evidence and transfer boundary.
  • C.40.CD:4.1–4.6 supplies the general reciprocal development of a question and a way; SYSE.22:4.1.1–4.1.6 binds its use here to family compatibility, realization and receiving decisions. SYSE.2:4.3–4.6 supplies the linked use–concept construction, and SYSE.3:4.2–4.4 the realization branch.
  • C.22.2 governs ProblemCard epistemes and C.22.PFR any world-side ProblematicForRelation. C.17, C.18, C.19, G.5, and G.11 govern characterization, archives and Fronts, live pools, selected sets, and use-qualified currentness. SYSE.19 and SYSE.23 retain changed-premise and claim-specific receiving rules.
  • C.11.CRC governs finite configuration-relative contribution comparisons and C.11 the bounded choice.
  • Supply the project focus, selected problem portfolio, System-family option set, correspondences, unresolved mismatches, and current ChoiceResult from SYSE.22 to the Agent using SYSE.23 for an investment or reconfiguration decision.
  • Feedback reopens SYSE.2, SYSE.6, or SYSE.13 only when the changed result crosses their stated reopen condition.
  • Decision Design, Strategy, Method Engineering, and Operations Management retain their domain Methods even when their Work uses the same archives or decisions.

SYSE.22:End

Referenced in the corpus

12 literal mentions in other sections. Read their context to establish the relation.