Answers to the five climate due-diligence questions with metrics, KPIs and the data science behind them. Figures marked model are engineering estimates from pilot data and published factors; confirm items need the founder's records before submission.
AIOTF's lever is input and residue optimisation on smallholder farms: the Climate Neural OS fuses satellite (NDVI/ND-BSI), soil, weather and on-farm sensors, then issues per-plot nudges that cut synthetic nitrogen, paddy methane and diesel, and route crop residue away from open burning into biochar or bioenergy. Every action is logged to a hash-chained MRV ledger so the reduction can be measured, not just claimed. Resilience comes from earlier stress alerts (water, heat, disease) that protect yield.
| Lever | Abatement | Cost ₹/tCO₂e | Cost $/tCO₂e | Efficiency |
|---|---|---|---|---|
| Precision N + bio-fertiliser substitution (N₂O) | 0.32 tCO₂e/acre/yr | ₹1,900 | $23 | −22% synthetic N per acre |
| AWD water scheduling for paddy (CH₄) | 0.9 tCO₂e/acre/yr | ₹1,450 | $17 | −31% irrigation water, −30–45% CH₄ |
| Crop-residue diversion from open burning → biochar | 0.7 tCO₂e/t straw | ₹4,600 | $55 | ≈ 0.28 t biochar per t straw, ~65% stable C |
| Solar-pump dispatch replacing diesel | 1.1 tCO₂e/pump/yr | ₹3,200 | $38 | ≈ 420 L diesel avoided per pump-yr |
Durability differs: farm avoidance (N₂O, CH₄, diesel) is permanent once avoided; soil carbon is reversible (20–100 yr); biochar is ~100+ yr; DAC with geological storage is 1,000+ yr. AIOTF should be compared with avoidance and short-to-medium-durability removals, not DAC.
| Component | TRL | Test conditions | Independent validation |
|---|---|---|---|
| Satellite + sensor fusion (ND-BSI, NDVI, THI) | 6 | Pilot plots, Latur & 3 states, Kharif/Rabi seasons | None yet — target ICAR / state agri university trial |
| Per-plot nudge engine (SLM + GraphRAG) | 5 | Field pilot, farmer-reported adoption | None yet |
| Spectral circularity grading (FTIR/NIR/HPLC/Raman) | 4 | Synthetic spectra; lab calibration pending | Needs ISO 17025 lab, RMSEP/R² report |
| Quantum optimisation (QUBO/VQC) | 3–4 | Classical emulation only | No hardware claim |
| Hash-chained MRV ledger | 6 | Prototype, tamper test passes | Needs third-party audit (VVB) |
confirm Replace with signed trial reports where they exist.