aprender-simulate 0.30.0

Unified Simulation Engine for the Sovereign AI Stack
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95

# TPS Test Case 1: Push vs Pull (CONWIP) Effectiveness
# EDD-compliant experiment specification

experiment_version: "1.0"
experiment_id: "TPS-TC-001"

metadata:
  name: "Push vs Pull Effectiveness"
  description: |
    Validate CONWIP superiority via Little's Law.
    This is the foundational TPS test case demonstrating that
    WIP-limited (Pull) systems outperform unrestricted (Push) systems.
  author: "PAIML Engineering"
  created: "2025-12-11"
  tags: ["tps", "operations", "push-pull", "conwip", "wip-control"]

equation_model_card:
  emc_ref: "operations/littles_law"

hypothesis:
  null_hypothesis: |
    H₀: There is no statistically significant difference in Throughput (TH)
    or Cycle Time (CT) between Push and Pull systems when resource capacity
    and average demand are identical.
  alternative_hypothesis: |
    H₁: Pull systems achieve lower cycle time with equivalent throughput.
  expected_outcome: "reject"  # We expect to FALSIFY H₀

reproducibility:
  seed: 42
  ieee_strict: true

simulation:
  topology:
    type: "tandem_line"
    stations: 5

  scenarios:
    - name: "push"
      wip_limit: null  # Unlimited
      description: "MRP-style unlimited WIP release"

    - name: "pull_conwip"
      wip_limit: 10  # System-wide cap
      description: "CONWIP with WIP cap of 10"

  arrivals:
    distribution: "poisson"
    rate: 4.5  # units/hour

  processing:
    distribution: "lognormal"
    mean: 1.0
    cv: 1.5  # High variability

  duration:
    warmup: 100  # hours
    simulation: 1000  # hours
    replications: 30

falsification:
  import_from_emc: true
  criteria:
    - id: "TC1-CT"
      name: "Cycle time reduction"
      condition: "pull_ct < push_ct * 0.6"
      severity: "major"
      confidence: 0.95

    - id: "TC1-TH"
      name: "Throughput maintained"
      condition: "|pull_th - push_th| / push_th < 0.05"
      severity: "major"

    - id: "TC1-VAR"
      name: "Variance reduction"
      condition: "std(pull_ct) < std(push_ct) * 0.5"
      severity: "minor"

  jidoka:
    enabled: true
    stop_on_severity: "critical"

statistics:
  comparison_test: "welch_t_test"
  alpha: 0.05
  effect_size: "cohens_d"

reporting:
  format: "markdown"
  output: "reports/tps/tc1_push_vs_pull.md"
  include:
    - "hypothesis_test_results"
    - "ct_distribution_comparison"
    - "wip_time_series"