Extreme Heatwave Early Warning and Human Thermal Stress Index

The receipts 🧾

Every red flag on this statement, in full. These are the four places it bites.

1. Exhibit A

   The statement also demands a mortality risk index and heat-attributable death data is not published at local granularity in India, so that half rests on either published epidemiological relationships applied from elsewhere or nothing at all — say which you did

2. It gets worse

   Wet-bulb globe temperature needs radiation and mean radiant temperature, which are frequently approximated badly; use a documented approximation and cite it rather than inventing a shortcut

3. Still reading?

   Ward-level output implies a spatial resolution that coarse forecast fields do not have, so be explicit that you are attributing a coarser field to wards via vulnerability weighting rather than claiming ward-resolution meteorology

4. And the finisher

   Predicting mortality is a claim with real consequences and weak evidence at this scale — presenting a relative risk grading is defensible, presenting projected death counts is not

The damage report 📉

Every score this statement earned, and what each one actually costs you.

• Feasibility

  3/5

  Buildable. Not comfortably. There is a week in here you have not planned for yet.

  The thermal stress half is entirely achievable — the established indices have published formulations and every input variable comes free from standard forecast fields — but the mortality risk index the statement also demands needs historical heat-attributable death records at local granularity, and India does not publish those, so half the ask has no data behind it.

• Innovation scope

  3/5

  Mildly interesting. The novelty will not carry the room; the build has to.

  The candidate indices are named, the lead time, the granularity and the delivery channels are specified, but how you combine thermal stress with demographic vulnerability into a single actionable grade — which is the actual product decision — is left entirely open.

• Clarity

  4/5

  The ask is unambiguous, which quietly removes your favourite excuse.

  The physiological argument is made precisely and correctly with a worked contrast between the same temperature at different humidities, the candidate indices are named, and the lead time, granularity, delivery channels and municipal triggers are all specified.

• Acceptance potential

  4/5

  Strong footing before you have written a line. Try not to waste it.

  The gap this addresses is real — warnings genuinely are issued on dry-bulb temperature and the physiology genuinely is ignored — the computation is free and well founded, the demo makes its own case visually, and heat is a rising hazard that far fewer teams choose than flooding or cyclones.

• Effort

  Heavy

  Heavy. Somebody on this team is not sleeping in week three. Pick who, on purpose.

  Index computation is straightforward but assembling ward-level demographic vulnerability layers, downscaling forecast fields to ward granularity, building the GIS dashboard and wiring the advisory and alerting paths is four workstreams with the vulnerability data the awkward one.

• Demo-ability

  Easy

  Easy to demo — and so is everyone else's. Working is the floor here, not the achievement.

  Two wards at the same temperature receiving different alert grades makes the entire argument of the statement visible in a single frame, and it is the rare demo that teaches the judge why the product needs to exist while demonstrating that it works.

• Data

  None supplied

  No dataset comes with this one, so every accuracy figure you quote is a number about labels you invented.

  Nothing is provided with the statement. You are sourcing, cleaning and labelling it yourself, and that work is invisible in the demo but very visible in the questions.

The demo they will have already seen 🪞

Somewhere around 140–330 teams are heading here, and the description is doing the choosing for most of them. They will read the same brief, reach the same architecture, and build a version of the same demo you are planning. Being correct is the floor. If your five minutes could be swapped with the team before you and nobody in the room would notice, you have not picked badly — you have built predictably, which costs exactly the same and hurts more.

What survives 🛡️

The ground worth standing on when the questions start.

• The core computation is free, well founded and standardised — the heat stress indices have published formulations and every input arrives in ordinary forecast output, so there is no data or modelling risk in the half that matters
• The two-wards-same-temperature demo teaches the judge why the statement exists and proves the system works in the same thirty seconds, which is a rare property
• Established heat action plans in Indian cities give you real, documented intervention thresholds to map your alert grades onto, so the advisories can be grounded rather than invented