Skip to content
SIH Buddyby Ganeev Singh
Dev

🔥 Roast My Pick · SIH26192

Flash Flood Prediction System for Hilly Regions using Multi-Source Data Theme

Ministry of Home Affairs

Brutal84/100

Bold. Let us find out precisely how bold, in the order a panel will find out.

Proceed with caution. A real and severe need, but hyper-local flash-flood prediction depends on IoT sensor networks not deployed in the target villages and validation data that is sparse — build on established rainfall-threshold methods with historical inventories, and be honest that the hyper-local lead-time claim outruns the available data. Roughly 130–310 teams are expected to go here.

The receipts

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

  1. Exhibit A

    Real-time IoT soil-moisture networks are not deployed in the target villages, so the sensor inputs the hyper-local forecast depends on are unavailable

  2. It gets worse

    Flash floods are extremely rapid and hyper-local, so village-level prediction with useful lead time is near the limit of what is achievable

  3. Still reading?

    Validation data for village-level events is sparse, so the lead-time claim is hard to substantiate

  4. And the finisher

    A disaster-authority judge knows how hard this is and will ask what the forecast was validated against

The damage report

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

  • Feasibility

    2/5

    You have picked a fight with physics, procurement, or both. One of them always wins.

    The meteorological and terrain data partly exist, but flash floods are extremely rapid and hyper-local, real-time IoT soil-moisture networks are not deployed in the target villages so those inputs are unavailable, and validating village-level predictions needs event data that is sparse — so the hyper-local lead-time claim rests on data and sensors you do not have.

  • Innovation scope

    3/5

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

    Flash-flood and landslide early warning combining rainfall thresholds with antecedent moisture and slope is established hydrology, so your room is in the multi-source integration and hyper-local resolution rather than in the concept.

  • Clarity

    4/5

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

    The data sources to integrate, the hyper-local village-level target and the lead-time-for-evacuation goal are stated clearly, so the target is well defined even though the sensor data is not available.

  • Acceptance potential

    2/5

    The numbers do not like you. Bring something the numbers cannot see.

    The need is real and severe, but the hyper-local lead-time claim depends on real-time IoT soil-moisture networks that are not deployed in the target villages, validation data is sparse, and flash floods are genuinely near-unpredictable at village resolution — so a disaster-authority judge will ask what the forecast was validated against and the honest answer is limited.

  • Effort

    Massive

    A semester of work wearing a hackathon costume. Something is getting cut; decide what now, not in week five.

    Integrating rainfall, soil moisture, slope stability, historical inventories and IoT feeds into a hyper-local predictive model with warning logic is a broad multi-source build.

  • Demo-ability

    Medium

    Demoable, if you rehearse it. Nobody rehearses it.

    A warning firing ahead of a replayed historical event is a clear story, but hyper-local flash-flood prediction is hard to validate, so the demo leans on a reconstructed event rather than a proven forecast.

  • 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 130–310 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.

  • Rainfall-threshold and antecedent-moisture methods for flash-flood and landslide warning are established hydrology to build on
  • Historical landslide inventories from GSI and Bhukosh give you real event data for a study region
  • A warning firing ahead of a replayed historical event is a clear, on-mission demo

None of that means do not pick it. It means do not walk into that room having heard any of this for the first time from a judge.

Full analysis ↗Roast another

The framing is a joke. The findings are not — they are the same analysis on the statement page, and every line above is attached to a score or a fact in the record. It is one opinion with its reasoning attached, so argue with it before you trust it.