{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp09_archetype_breadth150_20260804","research_id":"eoa_inverse_innovation_exp09_light_prior_art_20260804","cell_id":"load_shedding__chemistry_materials","search_lanes":{"direct_problem_and_intervention":{"queries":["autonomous laboratory experiment scheduling reject experiments capacity overload materials characterization","autonomous laboratory admission control experiments","laboratory scheduler cancel queued experiment priority capacity","experiment queue reject cancel overload laboratory automation"],"source_ids":["SRC1","SRC2"],"no_result_note":"No direct source was found implementing the complete combination of forecast-triggered, tier-selective rejection, protected-capacity reservation, fairness accounting, and hysteretic recovery in an autonomous materials laboratory."},"synonyms_and_historical_terms":{"queries":["self-driving laboratory scheduler resource constraints experiments materials science","laboratory information management system scheduling priority cancel queued experiments capacity","dynamic scheduling laboratory workflow resource conflicts","experiment queue cancel pending laboratory automation"],"source_ids":["SRC1","SRC2","SRC4"],"no_result_note":null},"products_practices_and_standards":{"queries":["SiLA 2 scheduler laboratory resource reservation cancellation queue standard","laboratory workflow management priority scheduling resource reservation","autonomous laboratory orchestration scheduling simulation priorities","sample scheduler laboratory capacity priority instruments"],"source_ids":["SRC1","SRC3","SRC4"],"no_result_note":null},"component_combination":{"queries":["autonomous materials laboratory multiple campaigns resource limits queues bottlenecks experiments scheduling paper","self-driving lab shared instruments queue bottleneck materials synthesis characterization capacity","autonomous lab experiment backlog resource constraints priority safety controls","A-Lab autonomous materials laboratory sample bottleneck characterization queue"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"AlabOS: A Python-based Reconfigurable Workflow Management Framework for Autonomous Laboratories","publisher":"Royal Society of Chemistry / arXiv","url":"https://arxiv.org/abs/2405.13930","source_type":"PRIMARY_RESEARCH","claims_supported":["Autonomous materials laboratories require workflow orchestration and resource management as their complexity grows.","AlabOS represents experiments as modular task workflows and uses resource reservation to prevent conflicts among concurrent synthesis and characterization tasks.","The associated implementation provides queued-work monitoring, operator cancellation of unfinished experiments or tasks, request priorities, and simulation before physical deployment.","AlabOS was demonstrated in A-Lab, which processed about 3,500 samples over 1.5 years."]},{"source_id":"SRC2","title":"Scalable Multi-Agent Lab Framework for Lab Optimization","publisher":"National Institute of Standards and Technology","url":"https://www.nist.gov/publications/scalable-multi-agent-lab-framework-lab-optimization","source_type":"PRIMARY_RESEARCH","claims_supported":["Scaling autonomous materials research across facilities creates a resource-coordination challenge.","MULTITASK explicitly models realistic equipment limits, multiple research-campaign agents, agent-instrument interactions, and facility-wide simulation.","The work establishes that shared resource limits and competing campaigns are visible design concerns for autonomous materials laboratories."]},{"source_id":"SRC3","title":"Scheduling — Experiment Orchestration System","publisher":"EOS Developers / UNC Robotics","url":"https://unc-robotics.github.io/eos/user-guide/scheduling.html","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["EOS jointly schedules laboratory protocol runs subject to dependencies, shared devices, resources, expected durations, and priorities.","EOS supports resource holds that preserve an allocation for successor tasks within a protocol run.","EOS provides an offline discrete-event simulator reporting utilization, concurrency, scheduling decisions, and bottlenecks.","Its documented schedulers reorder and reserve work but do not document overload-triggered rejection of selected unstarted protocol runs."]},{"source_id":"SRC4","title":"SiLA2 Manager Documentation, Release 01.02.2021","publisher":"Technical University of Munich SiLA2 Manager Project","url":"https://sila2-manager.readthedocs.io/_/downloads/en/latest/pdf/","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["The manager schedules experiments with specified execution times and required services and prevents creation when a required service is unavailable for the requested interval.","Scheduled experiments create service bookings, and the interface distinguishes unstarted, running, finished, and error states.","The documentation warns that deleting automatically created bookings can bypass a safety mechanism and permit conflicting access to state-sensitive laboratory devices.","This demonstrates that releasing reservations cannot be treated as a purely administrative operation; device-safety invariants must remain authoritative."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The sources make the underlying overload problem researchable: autonomous materials laboratories execute concurrent, multi-step campaigns against finite instrument and resource pools; resource limits, queues, priorities, reservations, contention, and characterization bottlenecks are explicitly modeled. However, the bounded search did not locate an empirical report showing this exact failure chain—optional admitted recipes causing mandatory hazard checks, controls, or unstable-sample characterization windows to be missed—so the proposal's strongest consequence remains plausible rather than directly demonstrated.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"closest_prior_art":[{"name":"AlabOS experiment and resource management","source_ids":["SRC1"],"overlap":"Autonomous materials-laboratory orchestration; modular synthesis and characterization workflows; queued-state visibility; resource reservation; priorities; cancellation of unfinished work; logging; and predeployment simulation.","remaining_difference":"Cancellation is operator-invoked and priority governs resource requests. The source does not disclose an automatically forecast-triggered policy that rejects only declared lower-tier unstarted recipes, protects specified obligations, records campaign-level fairness, and resumes through a lower hysteretic threshold."},{"name":"EOS joint scheduling, priorities, holds, and simulation","source_ids":["SRC3"],"overlap":"Forecastable durations, joint scheduling across protocol runs, task priorities, scarce-resource allocation, protected resource holds, bottleneck simulation, and offline comparison of scheduling policies.","remaining_difference":"EOS optimizes start order and resource assignment while retaining protocol runs. Its documentation does not disclose selective terminal rejection or expiration under an overload red-line, reason-coded resubmission, fairness allocation, or hysteretic readmission."},{"name":"MULTITASK facility-wide autonomous laboratory simulation","source_ids":["SRC2"],"overlap":"Multiple autonomous research campaigns competing for realistic equipment limits, queues, and shared facility resources, with simulation before deployment.","remaining_difference":"The framework establishes and simulates the contention setting but does not disclose the proposed protected-capacity shedding policy or its rejection and recovery semantics."},{"name":"SiLA2 Manager booking and conflict prevention","source_ids":["SRC4"],"overlap":"Admission is blocked when required services are unavailable, experiments and reservations are visible, and service bookings protect state-sensitive devices from conflicting access.","remaining_difference":"This is availability-based booking at experiment creation, not forecast-triggered shedding of previously admitted lower-tier recipes. It lacks scientific tiering, protected obligations, campaign fairness, explicit resubmission handling, and hysteresis."}],"prior_art_disposition":"ADJACENT_PRIOR_ART","contrastive_claim_remaining":"Relative to priority scheduling, resource reservation, conflict-free admission, and operator cancellation already shown in the retained art, the remaining falsifiable claim is that an autonomous materials laboratory can use a validated coupled-resource forecast to automatically give explicit terminal rejection to preclassified, unstarted optional recipes—while reserving feasible capacity for active-batch follow-ups, controls, hazard work, and irreversible characterization windows—and that fairness accounting plus a lower multi-cycle recovery threshold prevents hidden backlog growth, concentrated sacrifice, and oscillatory readmission.","contrastive_claim_falsifier":"The claim is falsified if an existing disclosed system already performs that complete policy, or if replay on timestamped campaigns shows that selective rejection does not improve protected-work feasibility over priority scheduling and reservation alone, sheds any protected or scientifically indispensable recipe, converts rejection into an increasing resubmission backlog, violates the declared fairness bound, or repeatedly toggles admission around the threshold.","gates":{"adequate_source_search":{"status":"PASS","rationale":"The bounded search covered direct formulations, self-driving/autonomous-laboratory terminology, older queue and dynamic-scheduling language, first-party orchestration and booking systems, resource standards/practices, and combinations involving competing campaigns, resource limits, priority, cancellation, reservation, simulation, and safety. Four opened direct sources span four publisher identities and include primary research and first-party documentation.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"supported_problem":{"status":"PASS","rationale":"The exact severe consequence was not directly observed, but primary and first-party sources support the essential problem structure: concurrent autonomous workflows, shared finite equipment, queues, priorities, resource conflicts, and bottleneck identification. This is sufficient for PARTLY_SUPPORTED problem evidence and a replayable local question.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"distinct_testable_claim":{"status":"PASS","rationale":"The remaining claim is narrower than the prior art: automated forecast-triggered terminal shedding of eligible unstarted recipes, protected-capacity guarantees, fairness accounting, and hysteretic recovery. Each element can be checked against timestamped scheduler records and explicit classification errors.","source_ids":["SRC1","SRC2","SRC3"]},"bounded_next_test":{"status":"PASS","rationale":"A non-operative replay or shadow simulation on one completed campaign is bounded and technically aligned with demonstrated autonomous-lab and EOS simulation practices. It can compare protected-schedule feasibility, wrong-shed classifications, backlog boundedness, fairness distribution, and threshold oscillation without controlling hardware or rejecting live work.","source_ids":["SRC1","SRC2","SRC3"]},"no_obvious_safety_or_authority_stop":{"status":"PASS","rationale":"The authorized first step is offline replay followed by non-operative shadow mode, so it neither starts, aborts, rejects, nor disposes of experiments. The proposal preserves human approval and halt authority and excludes active reactions, controls, calibrations, hazard checks, and regulatory work. SiLA2 documentation confirms that booking integrity is safety-relevant, reinforcing the need to test reservation release without bypassing device conflict protections; no unavoidable stop applies to the proposed non-operative test.","source_ids":["SRC3","SRC4"]}},"screen_survival":true,"world_novelty_boundary":"This bounded four-source screen establishes only coarse researchability and an adjacent-prior-art boundary. It cannot establish world novelty, patentability, freedom to operate, market size, expert acceptance, completeness of the literature, or realized scientific, operational, or safety value."}