Direction 01
A heat companion for outdoor-work supervisors: hourly risk, mandated-break suggestions and symptom triage grounded in heat-action-plan protocols
Sustainable Development Goal 13 / 17
Climate Action
Take urgent action to combat climate change and its impacts. For builders: tools that turn climate risk into local, timely, actionable information for the people who bear it first.
Why this goal matters
Climate change arrives as weather: a cyclone with a name, a heatwave without one, rain that skips a season then falls in a week. Global models exist; what is missing is the last mile — translating planetary signals into what a farmer, a fisher, a ward officer or a school should do this week. That translation layer is software.
The Indian context
India faces intensifying heatwaves, erratic monsoons, coastal exposure and glacier-fed river stress simultaneously. National adaptation programmes and early-warning systems have improved measurably — cyclone mortality has fallen dramatically over two decades — but heat action plans, urban flood response and farm-level advisories still reach far fewer people than they should.
Tamil Nadu, specifically
Tamil Nadu’s 1,000-kilometre coastline faces cyclones and sea-level stress; Chennai alternates between flood and drought; interior districts face heat and water variability that strain both farms and industry. The state publishes climate-action plans and district vulnerability data — raw material for builders who make it usable.
People and communities affected
- Coastal fishing communities
- Outdoor workers exposed to extreme heat
- Rain-dependent farming households
- Flood-prone urban neighbourhoods
- Children and elderly people during heat events
Key challenge areas
- Heat-risk early warning and response
- Hyper-local flood and cyclone communication
- Climate-adaptive farm advisories
- Community resilience planning
- Climate-data translation for local government
The AI Lens
Where AI belongs here — and where it does not.
Where AI can meaningfully help
- Downscaling forecasts into street- and village-level guidance in local languages
- Heat-stress alerts tuned to occupation — construction, delivery, farming — not just temperature
- Fusing satellite, terrain and report data for flood nowcasting
- Summarising climate assessments into ward-level action checklists
Where AI may be inappropriate
- Long-range climate prediction beyond established science — do not out-model the IPCC
- Fatalistic risk messaging without actionable guidance
- Carbon-offset scoring without verifiable measurement
Build Directions
Example problem directions
Starting points, not prescriptions — the strongest submissions narrow one of these into a specific, evidenced local problem.
Direction 02
A fisher’s weather bridge: sea-condition forecasts, warning translation and return-window guidance in Tamil, offline-tolerant
Direction 03
A school climate-drill assistant that localises cyclone and flood preparedness to each campus
Direction 04
A district adaptation dashboard turning published vulnerability data into ranked local actions
Potential users
- Coastal and farming communities
- Worksite supervisors and unions
- District disaster-management authorities
- Schools and local bodies
Possible datasets
- IMD forecasts, warnings and historical weather (public)
- INCOIS ocean-state and fishing-advisory services (public)
- ISRO/Bhuvan and open satellite imagery
- State climate-action plans and district vulnerability assessments (published)
Data-access limitations
- Forecast skill drops sharply at hyper-local scales — communicate uncertainty honestly
- Historical station data has gaps; satellite proxies need validation
- Vulnerability indices aggregate away the street-level differences that matter
Privacy, bias and safety risks
- Missed warnings and false alarms both cost lives and trust — precision-recall trade-offs are ethical decisions here
- Alert fatigue is real; frequency design matters as much as accuracy
- Climate tools that require constant connectivity fail during the disasters they exist for
Responsible-AI considerations
- Defer to official warnings; add localisation and reach, never contradiction
- Design offline-first for disaster conditions
- Test message comprehension with actual users before the season
Suggested impact metrics
- Warning reach and comprehension in pilot communities
- Lead time delivered versus baseline channels
- Documented protective actions taken following alerts
Related challenge areas
Participant resources
Build for Climate Action.
Anchor your team in this goal, define one real problem, and start the three-lock journey.
Register Your Team