{"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 delivery workload dependencies rework overtime burnout empirical study","project portfolio overload multi-project work delays employee stress empirical research","Understanding the causes of work overload in multi-project organizations 2026 Delisle","system dynamics project management rework cycle workload delay model"],"source_ids":["S1","S2","S6"],"no_result_note":"The underlying overload, delay, rework, and strain problem is supported, but no retained source directly demonstrates collapse after a modest demand increase while all single-team alarms remain nominal."},"closest_prior_art":{"queries":["dynamic mode decomposition project management organization","eigenmode project management workload","VAR eigenvalue organizational performance workload panel model randomized intervention","state-space project management rework workload"],"source_ids":["S2","S3","S4","S5"],"no_result_note":"No retained source applies randomized control-to-mode projection and a locked modal policy to aggregate cross-functional program-week delivery and worker-strain data."},"historical_terminology":{"queries":["older terminology project overload rework cycle schedule pressure overtime quality erosion project management","firefighting workload rework project delays organization empirical","vicious cycle schedule pressure rework overtime project management","MIT project management 90% syndrome rework cycle Brooks law"],"source_ids":["S2","S6"],"no_result_note":null},"products_practices_standards":{"queries":["official Kanban Guide work in progress limit flow metrics cycle time 2025","Azure DevOps cumulative flow WIP cycle time lead time official","ISO 45003 workload psychosocial risk official guidance","PMI program sponsor authority program management governance"],"source_ids":["S4","S5","S7","S8"],"no_result_note":null},"non_english_regional":{"queries":["Projektüberlastung Mehrprojektmanagement Arbeitsintensität psychische Belastung","Überlastung Multiprojektmanagement Abhängigkeiten Wartezeiten Nacharbeit","sobrecarga multiproyecto retrasos retrabajo dependencias gestión proyectos","gestión multiproyecto sobrecarga retrasos estrés"],"source_ids":["S6"],"no_result_note":"German official literature uses Arbeitsintensität and related job-demands, workload, and work-overload terminology. Spanish-language practitioner results were reviewed but not retained because they were weaker than the official and primary sources."},"composition_subproblems":{"queries":["cluster randomized trial work in progress limits project delivery teams workload","randomized experiment protected focus time dependency resolution software teams delivery","factorial randomized intake cap quality capacity cross-functional programs","randomized intervention workload cap project teams cycle time overtime"],"source_ids":["S3","S4","S5","S7"],"no_result_note":"Searches found the component methods and interventions separately, but not the proposed combination of a regularized endogenous transition matrix, bootstrap-stable modes, randomized B estimation, modal projection, subgroup harm gates, and a subsequent locked-policy trial."}},"sources":[{"source_id":"S1","title":"Understanding the causes of work overload in multi-project organizations: A multilevel perspective","url":"https://www.sciencedirect.com/science/article/pii/S026378632600075X","publisher":"Elsevier / International Journal of Project Management","date_or_year":"2026","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["An interpretive case study identifies work overload in a multi-project organization.","Lack of human-resource slack, overlapping and unpredictable deadlines, and blurred work-life boundaries interact across organizational levels.","The authors argue that overload should be addressed through organizational arrangements rather than assigned solely to individuals."]},{"source_id":"S2","title":"System Dynamics Modeling for Project Management","url":"https://web.mit.edu/jsterman/www/SDG/project.html","publisher":"MIT System Dynamics Group","date_or_year":"1992","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Large projects can appear healthy until previously hidden errors generate rework, overtime, hiring, schedule slippage, or quality and scope reductions.","Design changes and dependencies can create ripple effects across an organization.","System-dynamics project models and policy analyses were already used across multiple industries, making coupled delivery feedback an old research problem."]},{"source_id":"S3","title":"Dynamic mode decomposition with control","url":"https://arxiv.org/abs/1409.6358","publisher":"arXiv; Proctor, Brunton, and Kutz","date_or_year":"2014","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["DMD with control models x(k+1)=Ax(k)+Bu(k).","It separates estimated unforced dynamics from actuation and quantifies input effects.","It extracts modes and reduced-order input-output models from time-indexed state and control observations, closely matching the proposal's generic A/B decomposition lever."]},{"source_id":"S4","title":"The Kanban Guide","url":"https://kanbanguides.org/the-kanban-guide/","publisher":"Kanban Guides","date_or_year":"2025","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Operational Kanban already requires measurement of WIP, throughput, work-item age, and cycle time.","Flow measures are intended to inform workflow interventions and may be supplemented by context-specific measures.","WIP-centered monitoring and control are established practices, although the guide does not use invariant modes or randomized control identification."]},{"source_id":"S5","title":"Cumulative flow guidance for lead time and cycle time","url":"https://learn.microsoft.com/en-gb/azure/devops/report/dashboards/cumulative-flow-cycle-lead-time-guidance?view=azure-devops","publisher":"Microsoft Learn / Azure DevOps","date_or_year":"2025","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["Azure DevOps product guidance operationalizes WIP, scope, lead time, cycle time, blockages, and reactivated work.","The guidance states that higher WIP is associated with longer lead and cycle times and recommends WIP limits.","Existing dashboards can reveal bottlenecks and delayed external dependencies, providing a strong conventional comparator to modal forecasting."]},{"source_id":"S6","title":"Mental health in the working world — Determining the current state of scientific evidence","url":"https://www.baua.de/EN/Service/Publications/Report/Mental-Health-2.pdf?__blob=publicationFile&v=1","publisher":"German Federal Institute for Occupational Safety and Health (BAuA)","date_or_year":"2017","source_type":"SECONDARY_RESEARCH","language":"English translation of German report","claims_supported":["The official review treats high work intensity as a major stress factor affecting health, well-being, motivation, and performance.","Its work-intensity review included 294 German- and English-language studies.","It identifies job demands, workload, and work overload as overlapping historical terms and emphasizes quantity, time, speed, quality, and recovery."]},{"source_id":"S7","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?browse=tc","publisher":"International Organization for Standardization","date_or_year":"2021","source_type":"OFFICIAL_STANDARD","language":"English and Spanish","claims_supported":["Organizations are expected to manage psychosocial risks within an occupational-health-and-safety management system.","The standard applies across organization sizes and sectors and aims to prevent work-related injury and ill health while promoting well-being.","Its definition of worker includes worker representatives where they exist, supporting consultation and subgroup harm safeguards."]},{"source_id":"S8","title":"Organizational Project Management (OPM) Improvement Program Management","url":"https://www.pmi.org/learning/library/organizational-project-management-opm-improvement-program-management-10741","publisher":"Project Management Institute","date_or_year":"2015","source_type":"TRADE_PROFESSIONAL","language":"English","claims_supported":["A program sponsor is accountable for the program and establishes and leads governance.","The sponsor should possess authority or influence over affected organizational entities.","A program board can include the sponsor, program manager, operational integrator, and subject-matter experts, while end-user voices should be represented."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"Multi-project and cross-functional delivery environments demonstrably experience interacting overload, scarce slack, deadline overlap, switching, rework, schedule slippage, and worker strain. Older project-system-dynamics work also describes delayed discovery and reinforcing ripple effects. However, the specific claim that a reproducible eigenmode jointly loading workload, dependency waits, rework, and strain precedes failure while all silo alarms remain nominal has not been externally demonstrated.","source_ids":["S1","S2","S6"],"uncertainty":"The retained empirical work is case-based or concerns broader work intensity; it does not establish the proposed weekly state vector, a modest-demand threshold, absence of local alarms, recurrence across programs, or modal predictive superiority."},"adopter_evidence":{"status":"SUPPORTED","finding":"A program sponsor or senior operational executive with authority across participating units is an identifiable authorizer, supported by established program-governance practice. Program managers, operational integrators, participating leads, worker representatives, and an occupational-safety reviewer are identifiable implementing or reviewing actors.","source_ids":["S7","S8"],"uncertainty":"Actual authority over staffing, intake, people data, and cross-team controls is organization- and jurisdiction-specific and must be documented before enrollment."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"All major components exist separately: project feedback models for rework and delay, DMDc-style separation of A and B with modal analysis, established WIP and cycle-time monitoring, and psychosocial-risk governance. This supports technical plausibility and clear comparators, but not reliable recovery of stable organizational eigenmodes or causal modal-control effects from the proposed small, reactive, interference-prone panels.","source_ids":["S2","S3","S4","S5","S7"],"uncertainty":"DMDc demonstrations do not validate weekly organizational aggregates; 32 clusters may be inadequate for a three-factor cluster-randomized design after interference, attrition, common shocks, and subgroup safety analyses. Feasibility depends on prospective power, adherence, measurement invariance, conditioning, and stable regimes."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"System-dynamics project rework and schedule-pressure models","source_ids":["S2"],"same_problem":true,"same_causal_lever":false,"overlap":"Models coupled delay, rework, overtime, productivity, scope, and quality feedback and explains why projects can look healthy before late collapse.","remaining_difference":"Uses specified feedback structures and policy simulation rather than empirically estimated invariant modes, randomized B, modal projections, and locked policy evaluation."},{"name":"Dynamic mode decomposition with control","source_ids":["S3"],"same_problem":false,"same_causal_lever":true,"overlap":"Uses the same generic state-transition form, separates A from B, extracts dynamic modes, and quantifies control effects.","remaining_difference":"Does not address cross-functional program delivery, worker strain, randomized identification, aggregate people-data governance, blocked program validation, or subgroup harm endpoints."},{"name":"Kanban and Azure DevOps flow monitoring and WIP control","source_ids":["S4","S5"],"same_problem":true,"same_causal_lever":false,"overlap":"Operationalizes WIP, cycle time, lead time, scope, blockages, rework, and WIP-limiting interventions in routine delivery practice.","remaining_difference":"Acts on individual flow metrics or bottlenecks rather than a coupled endogenous propagation mode and does not randomize controls or distinguish changes in modal amplitude from changes in A's gain."}],"contrastive_claim_remaining":"Across independent cross-functional programs, an explicitly shock-adjusted, regularized A contains a bootstrap-stable coupled workload-debt mode that delivers at least the preregistered 10% blocked held-out predictive lift over individual-metric, common-shock, and single-bottleneck models; randomized protective controls shift its next-week amplitude in the projected direction; and a subsequently locked modal policy reduces late commitments without worsening quality or strain in any preregistered worker group.","contrastive_claim_falsifier":"The claim fails if no stable and well-conditioned mode recurs across held-out programs, predictive lift is below 10%, a common shock or single bottleneck performs as well, randomized assignments do not move the modal coordinate as projected, or the locked policy fails on delivery or crosses any quality or subgroup-strain harm margin.","confidence":"MODERATE","search_limitations":"The bounded search retained exactly eight direct sources and covered all six lanes, but paywalls and indexing limitations may hide relevant work. Searches cannot establish world novelty, patentability, freedom to operate, market size, routine adoption across all sectors, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"Independent empirical, official, and historical sources support the underlying coupled overload, delay, rework, and strain problem, even though the proposed eigenmode is unverified.","source_ids":["S1","S2","S6"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Program sponsors and cross-organizational governance boards are established accountable authorities, with program managers, operational leads, worker representatives, and safety reviewers as identifiable participants.","source_ids":["S7","S8"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Generic mode-with-control analysis and routine WIP management are prior art, but a falsifiable increment remains in applying randomized control-to-mode projections and harm-gated locked policies to cross-functional program-week delivery and strain panels.","source_ids":["S2","S3","S4","S5"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A preregistered aggregate-data feasibility and power phase can test measurement completeness, stability, conditioning, blocked predictive lift, interference, and cluster requirements before any locked-policy comparison; only a prospectively powered randomized B study would follow.","source_ids":["S3","S4","S5"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"No intrinsic stop was identified if data remain aggregate, individual scoring and personnel actions are prohibited, worker consultation and psychosocial-risk review occur, sponsor authority is documented, and quality and subgroup-strain harm rules can halt and roll back controls.","source_ids":["S6","S7","S8"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six required lanes were searched adversarially, including historical, product, standards, German and Spanish terminology, and component combinations. Eight retained direct sources span seven publishers and include primary research, official standards and guidance, and first-party product documentation.","source_ids":["S1","S2","S3","S4","S5","S6","S7","S8"]}},"strict_success":true,"screen_survival":true,"remaining_research_value":"MODERATE","recommended_next_step":"Preregister a non-interventional feasibility and power audit using de-identified aggregate program-week data from candidate programs. Require complete state definitions, shock logging, sufficient effective observations, blocked program holdouts, comparison against individual-metric and bottleneck models, eigenbasis and bootstrap-angle gates, residual diagnostics, interference mapping, and prospective cluster power. Proceed to randomized B identification only if every feasibility gate passes and sponsor, worker-representative, privacy, and independent safety approvals are documented.","world_novelty_boundary":"This bounded search identifies adjacent prior art and a remaining testable contrast; it does not establish world novelty, patentability, freedom to operate, market size, realized impact, or absence of undiscovered prior art."}