Methodology & Reset
jalarakshak-engine-ts-1.0 (port of reference_engine.py) · demo-seed-2026-10-12
Equations
ΔP = discharge_bar − suction_barhydraulic_kW = flow_m³h × ΔP / 36η_w2w = hydraulic_kW / electric_kW- Reject reading if flow < 50 m³/h, ΔP < 0.5 bar, kW < 2, field missing, or η outside 0..1.
- Baseline = median η of first 7 days; recent = median η of last 7 days; ≥5 valid days each.
deterioration % = max(0, (baseline − recent)/baseline × 100)- Pump index = 10 + min(30, 1.5 × deterioration) + vibration (+18 watch / +32 high) + bearing temp (+8 / +18) + repeat low-flow (+10) + open issue (+8), capped at 100.
- Mitigation points = index × criticality / 5 × mitigation fraction; inspections = 0.
- Maximise Σ(mitigation + λ × information) s.t. budget, crew hours, outage, one action per asset, verified evidence gates. Exact exhaustive enumeration.
Assumptions & warnings
- All assets, readings, costs, dates and thresholds are fictitious. No BWSSB, SCADA or GIS connection.
- The condition index is an explainable demonstration index, not a probability of failure.
- Thresholds are illustrative, not industrial-standard limits or BWSSB-approved criteria.
- Inspections improve information only; they do not physically reduce failure risk.
- Composite improvement is not proven failure prevention.
- The ₹1,30,000 scenario budget is unrelated to the Government First ₹25 lakh pilot budget.
- No equipment actuation, SCADA write or work scheduling occurs.
Numerical self-test: 14/14 passed
| Test | Expected | Computed | Result |
|---|---|---|---|
| P-101 assessment | 12.1 / conf 0.95 | 12.1 / conf 0.95 | PASS |
| P-102 assessment | 58 / conf 0.95 | 58 / conf 0.95 | PASS |
| P-103 assessment | 44 / conf 0.95 | 44 / conf 0.95 | PASS |
| P-104 assessment | null / conf 0.1 | null / conf 0.1 | PASS |
| V-201 assessment | 78 / conf 0.9 | 78 / conf 0.9 | PASS |
| V-202 assessment | 12 / conf 0.62 | 12 / conf 0.62 | PASS |
| VFD-301 assessment | 38 / conf 0.85 | 38 / conf 0.85 | PASS |
| MCC-401 assessment | 10 / conf 0.85 | 10 / conf 0.85 | PASS |
| Base optimised | A-01,A-06,A-08,A-10,A-11 · ₹98500 · 18h · 90m · 107.940 | A-01,A-06,A-08,A-10,A-11 · ₹98500 · 18h · 90m · 107.940 | PASS |
| Base risk-first greedy | A-06,A-01,A-04,A-09 · ₹102500 · 17h · 90m · 98.414 | A-06,A-01,A-04,A-09 · ₹102500 · 17h · 90m · 98.414 | PASS |
| After P-102 strainer_blocked | A-03,A-06,A-09,A-11 · ₹103500 · 18h · 105m · 109.524 | A-03,A-06,A-09,A-11 · ₹103500 · 18h · 105m · 109.524 | PASS |
| Restricted budget | A-06,A-08,A-11 · ₹77000 · 13h · 90m · 83.640 | A-06,A-08,A-11 · ₹77000 · 13h · 90m · 83.640 | PASS |
| A-02 stays gated after strainer finding | blocked | blocked | PASS |
| Inspections have zero physical mitigation | 0 | 0 | PASS |
Reset demo
Session holds 0 findings, 0 receipts and 0 audit events in this browser only. Reset restores the frozen seed state.