ICRISAT/Microsoft AI Sowing App

entry · p2026-09-07-047 · validated

achieved an average of 30% higher yield per hectare last year

tech: icrisat-sowing-app
passes the definition as: a sowing-date advisory computed by a model fit to 30 years of historic climate data (1986–2015) for the district, plus 10 years of groundnut sowing-progress data
from: 2016-06
region: IN
domain: work
side: benefit
people: 175 farmers (2016)
effect: 30 percent (2016) — average groundnut yield per hectare 30% higher for the farmers who sowed on the model's advised dates
reversible: yes
who holds the switch: owner:Microsoft and ICRISAT
attribution: named
source kind: secondary
the statistic: In the June 2016 kharif season 175 farmers in Devanakonda Mandal, Kurnool district, Andhra Pradesh received sowing-date advisories by text message on ordinary feature phones — no sensors, no capital outlay — and averaged 30% higher groundnut yield per hectare. The advisory was computed from historic climate data spanning 1986 to 2015 for the area, with weather models from aWhere Inc.
source: https://news.microsoft.com/en-in/features/ai-agriculture-icrisat-upl-india/
proposed by: researcher (opus), brief 2026-09-06 second harvest, robots-gig-work
received: 2026-09-06T13:34:02-04:00
harvest report: impacts/harvest/2026-09-07-robots-gig-work.md

Rulings, in order

2026-09-06T13:34:54-04:00 · validatorpass — quote found on the fetched page (exact match; HTTP 200 text/html; charset=UTF-8)
2026-09-06T13:38:18-04:00 · sessionnote — The programme's first measurement (175 farmers, 2016); p2026-09-07-048 is the same programme's wider count a year later and is the one that bands, so the two are not summed.

The proposal as it arrived (JSON)

{
 "id": "p2026-09-07-047",
 "received": "2026-09-06T13:34:02-04:00",
 "proposer": "researcher (opus), brief 2026-09-06 second harvest, robots-gig-work",
 "kind": "entry",
 "sha": "1271599815214843",
 "data": {
  "kind": "entry",
  "tech": "icrisat-sowing-app",
  "tech_name": "ICRISAT/Microsoft AI Sowing App",
  "definition": "a sowing-date advisory computed by a model fit to 30 years of historic climate data (1986–2015) for the district, plus 10 years of groundnut sowing-progress data",
  "from": "2016-06",
  "region": "IN",
  "domain": "work",
  "side": "benefit",
  "people": {
   "n": 175,
   "unit": "farmers",
   "date": "2016"
  },
  "effect": {
   "n": 30,
   "unit": "percent",
   "date": "2016",
   "what": "average groundnut yield per hectare 30% higher for the farmers who sowed on the model's advised dates"
  },
  "reversible": true,
  "switch": "owner:Microsoft and ICRISAT",
  "statistic": "In the June 2016 kharif season 175 farmers in Devanakonda Mandal, Kurnool district, Andhra Pradesh received sowing-date advisories by text message on ordinary feature phones — no sensors, no capital outlay — and averaged 30% higher groundnut yield per hectare. The advisory was computed from historic climate data spanning 1986 to 2015 for the area, with weather models from aWhere Inc.",
  "source": "https://news.microsoft.com/en-in/features/ai-agriculture-icrisat-upl-india/",
  "quote": "achieved an average of 30% higher yield per hectare last year",
  "attribution": "named",
  "source_kind": "secondary"
 },
 "from_file": "impacts/harvest/2026-09-07-robots-gig-work.md"
}