{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","research_id":"eoa_inverse_innovation_exp11_external_scrutiny_20260804","cell_id":"invariant_mode_decomposition_design__organizational_management","opaque_id":"invariant_mode_decomposition_design__organizational_management__B","search_lanes":{"direct_problem":{"queries":["cross-functional project dependencies workload rework delays worker burnout study","project management system dynamics rework cycle schedule pressure overtime project failure","site:who.int ILO WHO long working hours health burden official 2021","German project overload dependencies rework delays Arbeitsbelastung Projekt Studie"],"source_ids":["SRC3","SRC4","SRC5"],"no_result_note":null},"closest_prior_art":{"queries":["\"eigenvalues\" \"project management\" system dynamics workload rework","\"vector autoregression\" project performance workload rework","\"dynamic mode decomposition\" organizational performance management","\"modal analysis\" project management organizational workload"],"source_ids":["SRC1","SRC2","SRC3","SRC7"],"no_result_note":null},"historical_terminology":{"queries":["project management system dynamics rework cycle schedule pressure overtime project failure","\"A control system project development model derived from System Dynamics\" PDF","\"A.S. White\" \"control system project development model\""],"source_ids":["SRC1","SRC3"],"no_result_note":null},"products_practices_standards":{"queries":["site:iso.org ISO 45003 workload work pace psychosocial risks official","official Kanban Guide work in progress WIP limits PDF","Atlassian work in progress limits dependencies cycle time first party","PMI pulse profession project failure causes scope creep dependencies 2024"],"source_ids":["SRC6","SRC7"],"no_result_note":null},"non_english_regional":{"queries":["German project overload dependencies rework delays Arbeitsbelastung Projekt Studie","Die Einflüsse von Arbeitsbelastung auf die Arbeitsqualität agiler Software-Entwicklungsteams","site:who.int ILO WHO long working hours health burden official 2021"],"source_ids":["SRC4","SRC5"],"no_result_note":"German-language searching found direct workload-quality terminology and research but no close German-language implementation of the proposed empirical modal-control package. The WHO search also surfaced an Arabic-language version of its long-hours evidence."},"composition_subproblems":{"queries":["vector autoregression eigenvalue modes organizational project delivery workload system identification","\"dynamic mode decomposition with control\" external forcing actuation","cluster randomized trial organizations workplace intervention protocol methodology","cluster randomized trial work in progress limits protected quality capacity project teams"],"source_ids":["SRC1","SRC2","SRC7","SRC8"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"A control system project development model derived from System Dynamics","url":"https://www.sciencedirect.com/science/article/abs/pii/S0263786310001134","publisher":"International Journal of Project Management / Elsevier","date_or_year":"2011","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Project-development dynamics have previously been represented with state, control, and output matrices and analyzed for stability, controllability, observability, and dominant eigenvalues.","The model links dominant eigenvalues to hiring delay, assimilation delay, and employment duration, making it close domain-specific prior art.","The work is model-derived and compared with NASA project data, rather than estimating a regularized panel transition from multiple programs with randomized protective controls."]},{"source_id":"SRC2","title":"Dynamic Mode Decomposition with Control","url":"https://epubs.siam.org/doi/abs/10.1137/15M1013857","publisher":"Society for Industrial and Applied Mathematics","date_or_year":"2016","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Dynamic mode decomposition with control separates approximated endogenous dynamics from actuation using snapshots of observables and control inputs.","External forcing can corrupt uncontrolled modal estimates, supporting the proposal's separation of A from controls and observed shocks.","The method is general dynamical-systems prior art and does not establish the proposed organizational workload-debt mode or its worker-safety outcomes."]},{"source_id":"SRC3","title":"The rework cycle: how it really works ... and reworks ...","url":"https://www.pmi.org/learning/library/2019/04/07/15/16/project-success-rework-cycle-launch-5008","publisher":"Project Management Institute","date_or_year":"1993","source_type":"TRADE_PROFESSIONAL","language":"English","claims_supported":["Undiscovered rework can propagate into dependent tasks, causing productivity loss, delay, and further rework.","Complex development projects have long been modeled as interacting work-flow and rework structures rather than isolated metrics.","The source supports the broader coupled-delivery problem but does not demonstrate a reproducible empirical eigenmode hidden from silo dashboards."]},{"source_id":"SRC4","title":"Die Einflüsse von Arbeitsbelastung auf die Arbeitsqualität agiler Software-Entwicklungsteams","url":"https://arxiv.org/abs/2204.10911","publisher":"arXiv","date_or_year":"2022","source_type":"PRIMARY_RESEARCH","language":"German","claims_supported":["German-language research recognizes excessive workload in agile software teams as a potential cause of sickness absence, reduced well-being, reduced productivity, and poorer work quality.","The source supports measuring worker strain and quality alongside delivery rather than treating throughput as the sole objective.","It does not establish the proposed multivariate propagation mode or a randomized modal intervention rule."]},{"source_id":"SRC5","title":"Long working hours increasing deaths from heart disease and stroke: WHO, ILO","url":"https://www.who.int/news/item/17-05-2021-long-working-hours-increasing-deaths-from-heart-disease-and-stroke-who-ilo","publisher":"World Health Organization and International Labour Organization","date_or_year":"2021","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Long working hours are a material occupational-health hazard rather than merely a delivery metric.","WHO and ILO report higher stroke and ischemic-heart-disease risks for work of at least 55 hours per week and recommend agreed working-time limits.","Overtime interventions therefore require worker-health safeguards and cannot be justified solely by delivery gains."]},{"source_id":"SRC6","title":"ISO 45003:2021 — Occupational health and safety management — Psychological health and safety at work — Guidelines for managing psychosocial risks","url":"https://www.iso.org/standard/64283.html","publisher":"International Organization for Standardization","date_or_year":"2021","source_type":"OFFICIAL_STANDARD","language":"English and Spanish","claims_supported":["Organizations have an established occupational-health framework for managing psychosocial risks and preventing work-related injury and ill health.","Worker representatives are included in the standard's conception of workers, supporting consultation and governance requirements.","The standard supports safety oversight but does not prescribe modal workload analysis."]},{"source_id":"SRC7","title":"The Official Kanban Guide","url":"https://kanban.university/wp-content/uploads/2023/04/The-Official-Kanban-Guide_A4.pdf","publisher":"Kanban University / Mauvius Group","date_or_year":"2022","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Work-in-progress limits, pull signals, lead time, delivery rate, and WIP are established workflow-management practices.","The guide states that limiting admitted work can reduce delay and context switching and balance demand with capability.","Intake caps are therefore established practice, but selecting and combining them through an empirically identified risky mode is not described."]},{"source_id":"SRC8","title":"Outcomes of a Proximal Workplace Intervention Against Workplace Bullying and Harassment: A Protocol for a Cluster Randomized Controlled Trial Among Norwegian Industrial Workers","url":"https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2020.02013/full","publisher":"Frontiers in Psychology","date_or_year":"2020","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Organization-level workplace interventions can be evaluated with cluster-randomized designs and repeated measurements.","The protocol identifies management, HR, union representatives, and health-and-safety representatives as implementation and oversight actors.","It demonstrates the feasibility of a bounded organizational trial while also emphasizing context, implementation process, proximal outcomes, and worker consultation."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"Public evidence supports the broader problem package: rework propagates through dependent project work, schedule pressure and excessive workload can worsen quality and worker health, and established flow practice treats WIP, delay, and capacity as interacting quantities. The specific signature—collapse after modest demand growth, no silo alarm, and a recurring bootstrap-stable workload-debt eigenmode—has not been externally demonstrated and remains a testable hypothesis.","source_ids":["SRC3","SRC4","SRC5","SRC7"],"uncertainty":"The cited evidence spans development projects, agile software teams, general occupational health, and Kanban practice. It does not quantify how often the exact hidden coupled state occurs across independently managed cross-functional programs."},"adopter_evidence":{"status":"SUPPORTED","finding":"An identifiable authorizing coalition exists: program sponsors and participating program leads can authorize workflow controls, while HR or people-data governance, worker representatives, and occupational-health or safety reviewers govern strain measurement and worker protection. Comparable workplace trials explicitly involve department managers, HR, unions, and health-and-safety representatives.","source_ids":["SRC6","SRC8"],"uncertainty":"Actual authority will vary by organization, collective agreement, jurisdiction, and whether controls affect shared teams or customer obligations."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"The analytical ingredients and intervention components are implementable separately: project control models use state/control matrices and eigenvalues; DMD with control separates dynamics and actuation; Kanban operationalizes WIP limits and flow metrics; and workplace cluster-randomized interventions with repeated measurements have been conducted. No source validates the proposed full composition, its sample-size rule, weekly eight-state panel, randomized B projection, or locked-policy comparison.","source_ids":["SRC1","SRC2","SRC7","SRC8"],"uncertainty":"Thirty-two programs and 24 weeks may still be inadequate after accounting for regularization, common shocks, program heterogeneity, interference, missingness, adherence, and near-degenerate modes; prospective simulation-based power analysis is necessary."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"White's control-system project development model","source_ids":["SRC1"],"same_problem":true,"same_causal_lever":true,"overlap":"Applies linear control-system concepts to project development, including state and control matrices, stability, controllability, observability, dominant eigenvalues, rework, staffing, and external disturbance framing.","remaining_difference":"It simplifies a pre-specified system-dynamics model and compares it with project data; it does not estimate a cross-program weekly transition panel, separate randomized protective-control effects, require held-out modal prediction, or test delivery and disaggregated strain under a locked policy."},{"name":"Dynamic Mode Decomposition with Control","source_ids":["SRC2"],"same_problem":false,"same_causal_lever":true,"overlap":"Provides the general data-driven mechanism for extracting modes while separating endogenous dynamics from actuation.","remaining_difference":"It is not an organizational-management intervention and supplies no project-specific state vector, worker-strain safeguards, clustered identification design, or evidence that a workload-debt mode exists."},{"name":"Project rework-cycle system dynamics","source_ids":["SRC3"],"same_problem":true,"same_causal_lever":false,"overlap":"Models coupled propagation among work, undiscovered errors, dependent tasks, productivity, rework, schedule, and quality rather than relying on isolated project metrics.","remaining_difference":"It uses an explicit feedback structure rather than discovering bootstrap-stable invariant modes from program-week observations and does not identify projected effects of randomized protective controls."},{"name":"Kanban WIP limits and flow management","source_ids":["SRC7"],"same_problem":true,"same_causal_lever":false,"overlap":"Uses WIP limits, pull, lead time, delivery rate, and capacity balancing to reduce delay, context switching, and quality loss.","remaining_difference":"It is a general workflow practice, not a rule that retains a coupled endogenous mode only after stability, spectral-gap, residual, predictive-lift, controllability, and subgroup-harm gates."}],"contrastive_claim_remaining":"Across held-out independently managed programs, a bootstrap-stable coupled mode of WIP, dependency wait, rework, cycle time, delivery, overtime, and absence recurs and improves next-week collapse prediction by the preregistered margin over individual-metric, shock, and single-bottleneck models; randomized protective controls then shift that modal coordinate in predicted directions, and a locked mode-based policy reduces late commitments without worsening quality or strain in any preregistered worker group.","contrastive_claim_falsifier":"The claim fails if no stable well-conditioned mode survives blocked validation and spectral-gap gates; if it adds less than the preregistered 10% predictive lift over individual, shock, or bottleneck models; if assigned controls do not move the modal coordinate as predicted; or if the locked policy misses delivery outcomes or crosses any quality or subgroup-strain harm margin.","confidence":"HIGH","search_limitations":"The bounded search covered direct problem language, control-system and modal prior art, historical rework terminology, current standards and products, German-language terminology, and component combinations. It cannot rule out unindexed, paywalled, proprietary consulting, internal enterprise, patent, thesis, or additional non-English implementations. It does not establish world novelty, patentability, freedom to operate, market size, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"The coupled project-delivery problem is externally supported at the level needed to justify research: dependency-linked rework, workload, delay, quality, and worker-health consequences are documented. The proposed hidden eigenmode is appropriately left as the hypothesis to test.","source_ids":["SRC3","SRC4","SRC5","SRC7"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Program sponsors and leads are identifiable operational authorizers, with HR or people-data governance, worker representatives, and health-and-safety oversight identifiable for worker-affecting controls and aggregate strain data.","source_ids":["SRC6","SRC8"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Close prior art covers project eigenvalue/control analysis, generic controlled modal decomposition, rework feedback, and WIP limits, but the empirical cross-program composition with randomized B projections, held-out modal lift, a locked control rule, and disaggregated no-harm outcomes remains distinct and falsifiable.","source_ids":["SRC1","SRC2","SRC3","SRC7"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A preregistered, prospectively powered cluster-randomized identification study with fixed aggregate measures, blocked validation, explicit mode-rejection gates, adherence and contamination logging, and no automatic personnel action is bounded. Workplace cluster trials provide an implementation analogue, although power must be established before enrollment.","source_ids":["SRC2","SRC8"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first step uses protective controls and aggregate audit data, excludes discipline, surveillance, forced overtime, and staffing cuts, and includes consultation, independent review, subgroup harm margins, and rollback. Occupational-health evidence makes these safeguards mandatory but does not create an irresolvable stop.","source_ids":["SRC5","SRC6","SRC8"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six required lanes were searched adversarially, including historical control and rework terminology, standards and established flow products, German-language workload terminology, and combinations of system identification, modes, controls, and cluster trials. Exactly eight direct sources from independent publishers were retained, including multiple primary studies, official guidance, and an international standard.","source_ids":["SRC1","SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"]}},"strict_success":true,"screen_survival":true,"remaining_research_value":"HIGH","recommended_next_step":"Run a preregistered simulation-based design and feasibility phase before the full trial: use de-identified historical aggregate program-week data to verify measurement completeness, cluster independence, attainable effective sample size, regularization behavior, mode conditioning, blocked-validation procedures, and power under common shocks and contamination. Proceed to randomized identification only if governance approval, worker consultation, power, and all pretrial validity gates are satisfied; reserve the locked-policy comparison for a separate powered stage after B and the mode pass.","world_novelty_boundary":"This result identifies adjacent public prior art and a remaining contrastive research claim only. The bounded search cannot establish world novelty, patentability, freedom to operate, absence of proprietary practice, market size, or realized organizational impact."}