{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"pareto_frontier_navigation__chemistry_materials","arm":"CONSTRAINED_HIGH","candidate_id":"pareto_frontier_navigation__chemistry_materials__CONSTRAINED_HIGH","decision":"PROPOSAL","abstention_reason":null,"proposal":{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"pareto_frontier_navigation__chemistry_materials","arm":"CONSTRAINED_HIGH","candidate_id":"pareto_frontier_navigation__chemistry_materials__CONSTRAINED_HIGH","proposal_index":1,"version":0,"title":"Density-Gradient Frontier Tray for Recovering ABS from Mixed Plastic Flakes","problem":"A recycler selecting a density cutoff for recovering acrylonitrile-butadiene-styrene (ABS) flakes from mixed shredded plastics can mistake a poor cutoff for an unavoidable purity-versus-recovery tradeoff. A single sink-float trial may discard recoverable ABS while admitting contaminants, even though another feasible cut could be no worse in ABS recovery, contamination, liquid carryover, and fragment damage.","actors":["Mixed-plastic flake feed","ABS and non-ABS polymer fragments","Density-gradient liquid","Gradient column and nested collection ports","Recovered physical cut lots","Materials laboratory technician","Recycling process owner","Laboratory safety authority"],"observable_state":"Under a one-density sink-float baseline, changing the cutoff alters both recovered ABS mass and non-ABS contamination, but only one or two bulk fractions are available for comparison. Filled polymers, entrained air, incomplete wetting, and overlapping densities may cause a familiar cutoff to lose ABS without a compensating improvement in the other measured objectives.","consequence":"The process owner may commit to a physically dominated cutoff, causing avoidable rejection of target material, contaminant carryover, additional reprocessing, or unnecessary liquid handling.","affected_objective":"Identify a feasible physical cut that is non-dominated across ABS recovery, non-ABS contamination, retained separation liquid, and flake damage, subject to containment and product-purity guardrails.","intervention":"Place a bounded, representative batch of mixed flakes into a sealed, temperature-controlled density-gradient column after identical wetting and degassing. Buoyancy physically stratifies fragments along the column. Nested ports collect adjacent density bands into separate trays, and cumulative combinations of those bands instantiate multiple cutoff options from the same exposure. Weigh each lot, measure polymer identity on blinded subsamples with direct materials spectroscopy against physical reference specimens, measure retained liquid and fragment-size change, and exclude infeasible lots. Compare the resulting measurements without a composite score: a cumulative cut is removed when another cut is at least as good within measurement uncertainty on every objective and better on at least one. Keep the non-dominated physical lots as the frontier tray. For the bounded trial, apply a predeclared contamination ceiling and select, among frontier lots satisfying it, the cut with the greatest measured ABS recovery; retain the other frontier lots and rationale for rechecking.","structural_mapping":[{"archetype_element":"Option Set","domain_realization":"The cumulative combinations of adjacent density-band lots, each corresponding to a physically realizable outlet boundary."},{"archetype_element":"Objective Dimensions","domain_realization":"ABS recovery, non-ABS contamination, liquid carryover, and change in flake-size distribution."},{"archetype_element":"Feasibility Constraints","domain_realization":"Contained compatible liquid, stable gradient, acceptable product-contamination ceiling, no prohibited additive release, and no material damage beyond the predeclared limit."},{"archetype_element":"Dominance Criterion","domain_realization":"Cut A is dominated if another feasible cut is no worse within measurement uncertainty on all four objectives and is better on at least one."},{"archetype_element":"Frontier Map","domain_realization":"A rack of retained non-dominated physical cut lots linked to their direct mass, spectroscopy, carryover, and size measurements."},{"archetype_element":"Preference or Priority Rule","domain_realization":"First enforce the contamination ceiling; then choose the frontier cut with the greatest measured ABS recovery rather than collapsing all objectives into a hidden score."},{"archetype_element":"Tradeoff Rationale","domain_realization":"The selected cut records the additional target mass retained and the contamination, liquid, or damage accepted relative to neighboring frontier cuts."},{"archetype_element":"Recheck Trigger","domain_realization":"Repeat the comparison when feed composition, fragment-size distribution, temperature, wetting behavior, liquid composition, or product specifications change."}],"mechanism_mapping":[{"mechanism_slug":"dominance_screening","role":"Pairwise comparison removes cutoff lots that sacrifice at least one measured objective without improving any other relevant objective.","counterfactual_removal":"Without dominance screening, the column still separates material, but inferior cutoffs remain in contention and the archetype's false-tradeoff reduction is lost."},{"mechanism_slug":"pareto_frontier_analysis","role":"The non-dominated cumulative cut lots constitute the attainable frontier for the tested feed and physical apparatus.","counterfactual_removal":"Without retaining the non-dominated set, the trial degenerates into choosing one arbitrary density cutoff and cannot distinguish technical screening from the final preference choice."},{"mechanism_slug":"scenario_sensitivity_sweep","role":"Replicate aliquots and bounded changes in temperature or feed composition test whether apparent dominance survives material variability.","counterfactual_removal":"The core physical separation and initial frontier remain, but confidence that the frontier is stable across feed variation is lost."}],"causal_chain":["A stable liquid-density gradient creates a spatial buoyancy field without algorithmic routing.","Wetted and degassed flakes migrate to positions determined by their effective density and fluid interactions.","Nested ports convert spatial bands into distinct physical lots from one matched exposure.","Cumulative combinations of adjacent lots instantiate several feasible cutoff options.","Direct material measurements establish each option's target recovery, contamination, liquid carryover, and damage without a weighted score.","Feasibility guardrails exclude unacceptable lots before comparison.","Pairwise dominance checks remove cuts that worsen an objective without compensating improvement.","The remaining physical lots expose the genuine frontier, after which the explicit contamination-first preference rule selects one cut and acknowledges its opportunity cost."],"baseline":"A single-density batch sink-float test that produces one floating and one sinking fraction, followed by adjustment of the cutoff through separate trials. Because trials use separately conditioned aliquots and expose few cut points, they may confound feed variation with cutoff performance and leave dominated settings unrecognized.","nearest_rivals":["A series of independent sink-float baths at different fixed densities; it can sample multiple cuts but uses separate exposures and transfers that may change wetting, losses, and damage.","Per-flake spectroscopic sorting; it targets polymer identity directly but relies on item-by-item sensing and routing rather than producing a physical density-cut frontier.","Selective dissolution and reprecipitation; it may separate polymers chemically but changes solvent burden, polymer condition, and feasibility objectives.","Manual sorting aided by reference specimens; it preserves inspectability but is vulnerable to visually indistinguishable polymers and operator variation."],"remaining_contrastive_claim":"Relative to a single fixed-density cut, the proposed column instantiates several matched, physically recoverable cutoff options in one exposure and permits strictly dominated cuts to be identified before applying the contamination-first choice rule. Whether it yields a useful or stable frontier is untested.","authority_safety":{"decision_authority":"The materials laboratory manager may authorize the bench experiment; the recycling process owner may select a candidate cut only after laboratory safety review and product-specification review.","authorized_first_step":"Run one contained bench comparison on no more than two kilograms of characterized mixed-plastic feed, split into randomized baseline and gradient-column aliquots, using an already approved separation liquid and existing waste-containment provisions.","excluded_actions":["Production-scale deployment","Release or disposal of separation liquid outside approved laboratory waste handling","Use of unapproved flammable, reactive, or toxic density modifiers","Processing feed suspected to contain uncharacterized hazardous residues","Automatic routing of production material based on the exploratory result","Representing the selected cut as universally optimal"],"halt_rollback":"Stop for leakage, unexpected heating or vapor, unstable stratification, visible additive release, uncontainable microplastic escape, or measurement-control failure. Isolate the column, return recoverable liquid to its labeled containment vessel, retain all flakes as controlled laboratory waste or feedstock, and invalidate the run rather than selecting a cutoff."},"negative_tests":{"strongest_counterevidence":"Effective-density overlap, trapped gas, poor wetting, or additive variability may make band position a weak proxy for polymer identity; the apparent frontier could then be noise or an artifact of sample preparation.","problem_falsifier":"The problem is falsified if replicated baseline cutoffs are already non-dominated across all declared objectives within measurement uncertainty, or if the allegedly sacrificed ABS has no distinct feasible density band under the approved conditions.","intervention_falsifier":"The intervention is falsified if replicated gradient runs do not reproduce band ordering, if cumulative cuts cannot be collected without material mixing or loss, or if no cutoff dominance relationship survives blinded composition assays and uncertainty bounds.","risks":["Exposure to separation liquids or vapors","Leakage and microplastic release","Leaching of additives from shredded plastics","False dominance from spectroscopy or mass-balance error","Gradient drift with temperature","Flake agglomeration, trapped air, or incomplete wetting","Damage or contamination that makes recovered material unusable","Overgeneralization from one feed batch"]},"next_evidence_step":"Predeclare the four objectives, feasibility limits, dominance rule, measurement uncertainty treatment, and contamination-first preference rule; then conduct the authorized two-kilogram split-sample trial. Collect every density band, construct all contiguous cumulative cuts, assay blinded subsamples, and determine whether at least one apparent baseline option is reproducibly dominated and whether the retained frontier persists across replicate aliquots.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Not assessed against other proposals because runtime isolation prohibits inspecting them; this candidate was derived solely from the supplied archetype and chemistry/materials domain card.","revision_record":{"parent_version":null,"progress_targets_addressed":["Construct one causally explicit chemistry/materials candidate","Preserve feasibility filtering, dominance screening, frontier retention, and explicit final preference","Make the essential intervention physical rather than computational or procedural","Specify bounded evidence, authority, safeguards, falsifiers, and substrate counterfactual"],"conceptual_changes":["Translated abstract option navigation into physically instantiated cumulative density cuts of a mixed-polymer feed.","Separated physical production of frontier lots from the contamination-first value choice among those lots."],"operational_changes":["Specified a sealed gradient column, matched exposure, nested collection ports, direct material assays, and a two-kilogram first trial."],"evidence_changes":["Set prior art to unsearched and defined replicate, blinded, uncertainty-aware falsification checks without asserting effect size."],"claim_changes":["Limited the claim to matched physical option generation and the ability to test dominance; made usefulness and stability explicitly untested."]},"substrate_contract":{"primary_allowed_process":"PHYSICAL_MATERIAL","counterfactual_independence":"The essential effect is buoyancy-driven spatial segregation followed by physical collection of distinct material lots. If all software, algorithmic inference, dashboards, reports, incentives, authorization rules, and procedural enforcement are removed, the approved liquid gradient still stratifies the flakes and the ports still produce different recoverable cutoff lots. Direct weighing and materials spectroscopy are measurement support; they do not create or route the separation. Human preference is needed only to choose among the already produced frontier candidates, not to cause their physical formation.","forbidden_channel_audit":"No model, database, recommender, dashboard, or software feedback loop controls flake motion or outlet routing. No sensor triggers downstream sorting. Pumps, if needed, maintain a fixed physical condition rather than executing an inferred policy. The contamination ceiling, safety authorization, and selection record are support wrappers; removing them does not eliminate buoyancy, stratification, or collection, although actual experimentation must still obey laboratory authority and safety requirements."}}}