A supplied screening queue model: more coverage raises load, extra capacity reduces waiting

Supplied research · Reviewed findings · 30 October 2026

Incoprea Research · Supplied findings II · 30 October 2026

Evidence status: Reproduced conditional queue simulation; no false-positive, detection, safety or real staffing validation.

In a supplied queue simulation, selecting more arriving items for inspection increases waiting when staffing stays fixed. At 99% coverage, increasing parallel servers from 40 to 45 sharply reduces the simulated delay. The result is reproducible, but the source’s claim that 99% coverage becomes stable only above 40 servers is incorrect: 40 already gives a nominal load below one.

The hypothesis and the tested part

Screening Trades Flow for Safety links inspection coverage, false positives and clearance time. The supplied code tests only coverage and queue capacity. Each arrival is selected randomly with a fixed probability; there is no evidence threshold, risk score, true threat status, false-positive rate or detection outcome.

The numerical finding therefore concerns waiting for a generic inspection. It cannot establish the original false-positive mechanism or show that a real biosecurity system is safe, effective, or adequately staffed.

What the supplied simulation implements

The source specifies 600 arrivals per hour and four-minute service at each identical parallel server. It generates independent Poisson arrival counts for each minute, then places all of that minute’s items at the same minute timestamp. This is a minute-batched implementation, not continuous, individually timed Poisson arrivals. Selected items are scheduled with the earliest available server; unselected items bypass inspection.

Each scenario covers 480 hours, or 20 days, starting with every server free. The supplied sweep contains 24 combinations of four server counts and six coverage levels. A separate capacity section adds seven scenarios at 99% coverage, making 31 scenarios in total. They use successive draws from one generator seeded 20261030; they are not 31 replications of one condition.

Selected-item waiting, not total clearance

The reported mean and 95th percentile include all selected items, including those with zero queue delay. They exclude unselected items. They measure arrival-to-start waiting, not completion time: every selected item still requires four minutes of service afterward.

The counter named n_inspected records selected items scheduled for service. Some final scheduled starts or completions can lie beyond the 20-day arrival window. It is not a verified count of inspections completed by the deadline. A 400,000-selected-item stop guard exists, but no supplied scenario reaches it; the largest count is 286,044.

ServersCoverageNominal loadMean wait before service95th-percentile wait
4090%0.900.223 min1 min
4095%0.950.698 min3 min
4099%0.994.858 min17 min
4599%0.880.129 min1 min

These figures are from the supplied main sweep. Adding five servers increases nominal service capacity by 12.5%. Waiting becomes much smaller, but it does not disappear. Including inspection time, the corresponding 99%-coverage mean time for selected items is about 8.858 minutes with 40 servers and 4.129 with 45.

Correcting the stability arithmetic

With coverage c, expected selected arrivals are 600c per hour. Each server can perform 15 four-minute inspections per hour. Nominal utilization is therefore 600c/(15N)=40c/N.

At c=0.99 the offered workload is 39.6 fully occupied servers. Forty gives utilization 0.99; 45 gives 0.88. Forty already satisfies the below-unit-load condition. Additional headroom can improve delay performance, but “stable” and “short waits” are different claims. The supplied capacity section also reports a mean below its stated ten-minute target at 40 servers, so it does not establish a requirement for more than 40.

Finite-run limits

No warmup period is discarded and no steady-state convergence check or uncertainty interval is supplied. Near saturation, the initial empty system and finite window matter. The package itself contains two separately generated 40-server, 99%-coverage scenarios: one reports mean wait 4.858 minutes and a 17-minute 95th percentile, while the other gives 3.836 and 10 minutes. Neither is a universal stationary value.

Minute batching, deterministic service, constant arrival intensity, unlimited waiting space, random coverage and absence of breaks, failures, priorities or departures further limit transfer to practice. A reported zero 95th percentile elsewhere in the sweep also does not imply every item has zero delay.

Context and provenance

Larson and Odoni’s queueing definitions distinguish queue waiting from total system time. Dai and He’s parallel-server analysis provides established context for capacity, utilization and service quality; its analytical formulas are not substituted for this different batched model.

The supplied Incoprea package credits DeepSeek Harness with the analysis, code and drafting. Codex audited it, ran the unchanged script in an isolated copy, obtained a byte-identical result file, and prepared this corrected text. This reproduces supplied computation, not a new study or external peer review. The original hypothesis remains unchanged at Incoprea Research.

Inspect and reproduce the supplied result

Python 3 and NumPy. The script writes queue.json beside itself. Save the complete code as reproduce.py in a new folder and run python3 reproduce.py. This is a check of a supplied study, not a newly designed test.

Show the complete runnable reproduction
# Portable publisher copy: only ROOT/output pathname adapted; algorithm,
# parameter grid, seed and order are identical to the supplied original.
# Requires NumPy. Run: python3 reproduce.py ; writes queue.json beside script.
"""V20 #28 'Screening Trades Flow for Safety': coverage, capacity and clearance time."""
import json, os
import numpy as np
ROOT=os.path.dirname(os.path.abspath(__file__))
rng=np.random.default_rng(20261030)
LAM=600.0; SERVICE=4/60.0; HOURS=24*20
COVERAGES=[0.5,0.7,0.8,0.9,0.95,0.99]

def simulate(cov,N):
    # Poisson arrivals by minute; deterministic 4-min inspection on N parallel servers
    n_min=int(HOURS*60); arrivals=rng.poisson(LAM/60.0,n_min)
    queue=[]; servers=np.zeros(N); waits=[]; done=0
    t=0.0
    for m in range(n_min):
        for _ in range(arrivals[m]):
            if rng.random()<cov:
                # schedule: earliest free server
                j=int(np.argmin(servers))
                start=max(servers[j],m*1.0)
                wait=start-m*1.0
                servers[j]=start+SERVICE*60
                waits.append(wait)
        if len(waits)>400000: break
    waits=np.array(waits)
    return dict(coverage=cov,N=N,n_inspected=int(len(waits)),
                mean_wait_min=float(waits.mean()) if len(waits) else None,
                p95_wait_min=float(np.percentile(waits,95)) if len(waits) else None,
                load=float(LAM*cov*SERVICE/N))

out={'design':{'arrivals_per_hour':LAM,'service_min':SERVICE*60,'hours':HOURS,'inspectors':None},'sweep':[],'capacity':[]}
for N in [40,45,50,60]:
    for c in COVERAGES:
        r=simulate(c,N); out['sweep'].append(r)
        print(f"N={N} cov={c} load={r['load']:.2f} mean_wait={r['mean_wait_min']:.1f}min p95={r['p95_wait_min']:.1f}")
# minimum N for 99% coverage and mean wait under 10 min
for N in [40,41,42,44,46,48,50]:
    r=simulate(0.99,N); out['capacity'].append(r)
    print('capacity',r)
json.dump(out,open(os.path.join(ROOT,'queue.json'),'w'),indent=2)
Show the published results
{
  "design": {
    "arrivals_per_hour": 600.0,
    "service_min": 4.0,
    "hours": 480,
    "inspectors": null
  },
  "sweep": [
    {
      "coverage": 0.5,
      "N": 40,
      "n_inspected": 144138,
      "mean_wait_min": 6.937795723542716e-06,
      "p95_wait_min": 0.0,
      "load": 0.5
    },
    {
      "coverage": 0.7,
      "N": 40,
      "n_inspected": 202128,
      "mean_wait_min": 0.005664727301511913,
      "p95_wait_min": 0.0,
      "load": 0.7
    },
    {
      "coverage": 0.8,
      "N": 40,
      "n_inspected": 230094,
      "mean_wait_min": 0.034703208254017924,
      "p95_wait_min": 0.0,
      "load": 0.8
    },
    {
      "coverage": 0.9,
      "N": 40,
      "n_inspected": 259318,
      "mean_wait_min": 0.22296176894777842,
      "p95_wait_min": 1.0,
      "load": 0.9
    },
    {
      "coverage": 0.95,
      "N": 40,
      "n_inspected": 273868,
      "mean_wait_min": 0.6979749368308821,
      "p95_wait_min": 3.0,
      "load": 0.95
    },
    {
      "coverage": 0.99,
      "N": 40,
      "n_inspected": 284841,
      "mean_wait_min": 4.85812084636692,
      "p95_wait_min": 17.0,
      "load": 0.99
    },
    {
      "coverage": 0.5,
      "N": 45,
      "n_inspected": 144259,
      "mean_wait_min": 0.0,
      "p95_wait_min": 0.0,
      "load": 0.4444444444444444
    },
    {
      "coverage": 0.7,
      "N": 45,
      "n_inspected": 201998,
      "mean_wait_min": 0.0003564391726650759,
      "p95_wait_min": 0.0,
      "load": 0.6222222222222222
    },
    {
      "coverage": 0.8,
      "N": 45,
      "n_inspected": 230650,
      "mean_wait_min": 0.004617385649252114,
      "p95_wait_min": 0.0,
      "load": 0.7111111111111111
    },
    {
      "coverage": 0.9,
      "N": 45,
      "n_inspected": 259397,
      "mean_wait_min": 0.028735875896791406,
      "p95_wait_min": 0.0,
      "load": 0.8
    },
    {
      "coverage": 0.95,
      "N": 45,
      "n_inspected": 273192,
      "mean_wait_min": 0.059126914404521363,
      "p95_wait_min": 1.0,
      "load": 0.8444444444444444
    },
    {
      "coverage": 0.99,
      "N": 45,
      "n_inspected": 284925,
      "mean_wait_min": 0.12930420286040187,
      "p95_wait_min": 1.0,
      "load": 0.88
    },
    {
      "coverage": 0.5,
      "N": 50,
      "n_inspected": 143810,
      "mean_wait_min": 0.0,
      "p95_wait_min": 0.0,
      "load": 0.4
    },
    {
      "coverage": 0.7,
      "N": 50,
      "n_inspected": 201936,
      "mean_wait_min": 2.9712384121701924e-05,
      "p95_wait_min": 0.0,
      "load": 0.56
    },
    {
      "coverage": 0.8,
      "N": 50,
      "n_inspected": 231226,
      "mean_wait_min": 0.00026813593627014265,
      "p95_wait_min": 0.0,
      "load": 0.64
    },
    {
      "coverage": 0.9,
      "N": 50,
      "n_inspected": 258964,
      "mean_wait_min": 0.004282448525663798,
      "p95_wait_min": 0.0,
      "load": 0.72
    },
    {
      "coverage": 0.95,
      "N": 50,
      "n_inspected": 274556,
      "mean_wait_min": 0.009790352423549293,
      "p95_wait_min": 0.0,
      "load": 0.76
    },
    {
      "coverage": 0.99,
      "N": 50,
      "n_inspected": 284713,
      "mean_wait_min": 0.017947898409977767,
      "p95_wait_min": 0.0,
      "load": 0.792
    },
    {
      "coverage": 0.5,
      "N": 60,
      "n_inspected": 144583,
      "mean_wait_min": 0.0,
      "p95_wait_min": 0.0,
      "load": 0.3333333333333333
    },
    {
      "coverage": 0.7,
      "N": 60,
      "n_inspected": 202166,
      "mean_wait_min": 0.0,
      "p95_wait_min": 0.0,
      "load": 0.4666666666666667
    },
    {
      "coverage": 0.8,
      "N": 60,
      "n_inspected": 230539,
      "mean_wait_min": 0.0,
      "p95_wait_min": 0.0,
      "load": 0.5333333333333333
    },
    {
      "coverage": 0.9,
      "N": 60,
      "n_inspected": 258698,
      "mean_wait_min": 1.5462044546150336e-05,
      "p95_wait_min": 0.0,
      "load": 0.6
    },
    {
      "coverage": 0.95,
      "N": 60,
      "n_inspected": 273353,
      "mean_wait_min": 5.12158271538999e-05,
      "p95_wait_min": 0.0,
      "load": 0.6333333333333333
    },
    {
      "coverage": 0.99,
      "N": 60,
      "n_inspected": 285210,
      "mean_wait_min": 0.00026647032011500296,
      "p95_wait_min": 0.0,
      "load": 0.66
    }
  ],
  "capacity": [
    {
      "coverage": 0.99,
      "N": 40,
      "n_inspected": 284772,
      "mean_wait_min": 3.836451617434298,
      "p95_wait_min": 10.0,
      "load": 0.99
    },
    {
      "coverage": 0.99,
      "N": 41,
      "n_inspected": 286044,
      "mean_wait_min": 1.2945106347275244,
      "p95_wait_min": 4.0,
      "load": 0.9658536585365854
    },
    {
      "coverage": 0.99,
      "N": 42,
      "n_inspected": 285114,
      "mean_wait_min": 0.5619331214882468,
      "p95_wait_min": 2.0,
      "load": 0.9428571428571428
    },
    {
      "coverage": 0.99,
      "N": 44,
      "n_inspected": 285272,
      "mean_wait_min": 0.17994755882105498,
      "p95_wait_min": 1.0,
      "load": 0.9
    },
    {
      "coverage": 0.99,
      "N": 46,
      "n_inspected": 285032,
      "mean_wait_min": 0.08470277021527407,
      "p95_wait_min": 1.0,
      "load": 0.8608695652173913
    },
    {
      "coverage": 0.99,
      "N": 48,
      "n_inspected": 284156,
      "mean_wait_min": 0.03578316136206872,
      "p95_wait_min": 0.0,
      "load": 0.8250000000000001
    },
    {
      "coverage": 0.99,
      "N": 50,
      "n_inspected": 284491,
      "mean_wait_min": 0.017441676538097865,
      "p95_wait_min": 0.0,
      "load": 0.792
    }
  ]
}
Show provenance and review boundaries
{
  "source_folder": "Incoprea Hypothesis Tests II - 2026-10-30",
  "source_code_sha256": "2af237963a5f56b1630a45b083307b68cc21396376961c2d4fbdb9f8186fe03f",
  "source_result_sha256": "358379d01463ff48add8420e7e819c1ceb027ecb1a75d85d60c3acc940f2feae",
  "review": "Unchanged supplied code regenerated a byte-identical result; portable copy changes output location only",
  "scope": "Conditional finite queue simulation, not a measured inspection or safety system",
  "runtime": "Review used Python3.14 and NumPy2.5.4",
  "limits": "Minute-batched arrivals; waiting only among selected scheduled items; no warmup or repeated-run interval; no false-positive or detection data"
}

Supplied analysis, code and initial drafting: DeepSeek Harness, as identified in the package. Consistency review, reproduction and corrected publication text: Codex. Internal computational review is not external peer review. Incoprea Research · Original posted question or hypothesis.

Comments

Post a Comment

Popular posts from this blog

Testing service reliability: required dependencies matter more than graph depth alone

Tested hypothesis · Mathematical and model findings · 30 October 2026 Incoprea Research · 30 October 2026 Evidence status: Exact conditional support for serial multiplication; counterexample to depth as a sufficient predictor. Reliability model and synthetic checks, not measured outage evidence. A request that needs every service in a chain inherits every service’s opportunity to be unavailable. That mechanism in Depth Beats Count is correct under independence. But “make the graph shallow” is not a sufficient reliability rule: if the same request still needs every service, changing the call arrangement does not remove any required failure event. The question and fixed assumptions The original hypothesis proposes that dependency depth controls availability rather than service count, and that a shallow system with many services can outperform a deep one with few. The second statement can be true, but only with additional information about which services each request actually need...
Read more

Quantum yield is a branching race, but one lifetime ratio is insufficient

Tested hypothesis · Mathematical and model findings · 30 October 2026 Incoprea Research · 30 October 2026 · Mathematical/model test Finding: Competing reaction pathways determine yield in a simple excited-state model, and stronger absorption alone changes the amount of product rather than the yield per absorbed photon. However, the lifetime-to-vibrational-relaxation ratio proposed in Quantum Yield Is a Race is not generally sufficient: two model systems can have exactly the same ratio and quantum yields of 0.1 and 0.8. This is an exact kinetic derivation and synthetic stochastic check. It narrows a plausible mechanism in the Incoprea Research hypothesis collection ; it does not experimentally validate the original predictor. Branching kinetics and the distinction between absorption and quantum yield are established science, not new discoveries. This note has no external peer review. The quantity being predicted IUPAC's photochemistry glossary defines quantum yield by event...
Read more

When rare events overwhelm a detector: an exact Bayes test

Tested hypothesis · Mathematical and model findings · 30 October 2026 Incoprea Research · Original hypothesis Exact support for the base-rate mechanism; accuracy claim corrected · 30 October 2026 A detector that catches 90% of genuine events and correctly clears 90% of non-events can produce less than 1% genuine events among its positive flags when prevalence is 0.1%. Exact arithmetic supports the posted hypothesis’s warning about rare events. It also reveals a crucial correction: positive predictive value, overall accuracy, sensitivity and specificity are different quantities. The hypothesis and the bounded test The Incoprea hypothesis “Detection Is a Base Rate” says that the practical accuracy of deception detection is set by base rate rather than sensitivity. Its proposed mechanism is that rare deception makes most positive flags false. We tested that probability claim using a hypothetical detector. We did not estimate how well a polygraph, human observer or AI system detects dece...
Read more