Home › Hard problems › India

India: hard problems at India's scale

Hyperlocal monsoon and flood forecasts

Solo: strongOne person can run a real test of this in about 8 weeks.

"light rain""light rain""light rain""light rain""light rain"a global model: one forecast for each big square"light rain"a downpour onone districtWhat a district needsheavy rain, here,a few hours aheadWhat it getsgood big-picture forecasts;floods hit with little warningas in Assam and Kerala
Global AI weather models see the big picture, but the rain that floods one district a few hours from now falls inside a single square.

Why it is still open

Global AI weather models are good at large scale, but rain for one district a few hours ahead is still poor, and floods in Assam and Kerala hit with little warning.

An experiment one person can run in 8 weeks

Take open forecasts (GraphCast, or ERA5 data) and IMD station data. Measure the error for 10 districts, then fine-tune a small local rain correction.

Verdict: Solo: strong

Open data and a clear error number, and farmers, insurers and cities care.

The solo test behind the verdict

  1. Can it be measured? There is a number that shows you did better, such as an error rate, a catch rate or a cost per answer.
  2. Is it cheap to try? It runs on a laptop, free open models and open data, or under about ₹50,000 of rented GPU time.
  3. Is there a public yardstick? A benchmark or dataset others use, so your result can be compared and believed.
  4. Would a small win get noticed? A lab, a funder or a customer would care about a 10% improvement, not only a full solution.

Four yeses means Solo: strong. Two or three means Solo: possible. Fewer means Follow, don't build.

PreviousVoice-clone and UPI scams in Indian languagesNextCrop yield and stress for small farms

All 79 hard problems · Working on this, or know of a result we missed? basant@cleandataengineering.ai