AI in Disaster Management: From Nepal to India

Key Highlights & Summary 

• Nepal example: During the recent Nepal disaster, an AI-powered web portal built by IT engineer Niraj Bhushal of Nepal\'s Finance Ministry matched crowdsourced data on missing persons with official lists of the dead and injured. The tool also helped map damaged buildings. Drone operator Manish Maharjan used thermal cameras to detect human-shaped heat signatures in debris and guide rescue teams.

• Growing actors: Disaster management is no longer limited to designated authorities. Technology firms such as Google, telecom providers and satellite companies are now integral players. AI can process large volumes of data through natural language, speech and video on an almost real-time basis, which is vital when time is critical.

 • Early warning: AI models trained on vast past weather data learn how the atmosphere behaves and produce forecasts much faster than physics-equation-based methods. This makes hyperlocal forecasts possible in almost real time, as IIT Bombay is attempting for heavy rainfall events in Mumbai. 

• Flood forecasting: Google\'s Flood Hub uses a prediction model first applied to floods in India in 2018 and can issue advisories up to seven days earlier. Forecasts must be combined with hazard maps to assess risk, where AI can integrate multiple datasets for threat assessment. 

• Response and post-event: Smartphones and drones act as force multipliers, generating large volumes of data. AI can parse unstructured data, even in local languages. Specialised tools help locate disconnected habitations, assess road conditions, identify helicopter landing sites and prioritise food and relief distribution. 

• Human dimension: Success still depends on human expertise, strong institutions and governance frameworks. As the UN Office for Disaster Risk Reduction notes, the real measure is whether AI enables timely, equitable and trusted protection for the most vulnerable. 

Essential Definitions

 • Disaster management: The cycle of preparedness, mitigation, response, relief and recovery aimed at reducing disaster risk and impact. 

• Early warning system: A set of tools and processes that detect hazards and communicate timely alerts to people at risk. 

• Hazard mapping: Identifying and mapping areas exposed to hazards such as floods, earthquakes and landslides. 

• Force multiplier: A tool that increases the effectiveness of existing resources without increasing their numbers. 

• Hyperlocal forecast: A weather prediction for a very small area and short time span.

 Legal and Constitutional Framework

 • Disaster Management Act, 2005: Provides for NDMA, SDMAs and DDMAs, and a legal framework for planning, mitigation and response. 

• Article 21: The right to life includes the right to safety, and the State\'s duty to protect citizens from disasters. 

• Entry 23, Concurrent List: Social security and welfare, and Entry 29 on prevention of the extension of infectious or contagious diseases, relevant to relief measures. 

• Article 355: The Union has a duty to protect states against external aggression and internal disturbance.

 • National Disaster Management Plan, 2019: Aligned with the Sendai Framework for Disaster Risk Reduction 2015-2030. 

• Digital Personal Data Protection Act, 2023: Governs the processing of personal data, relevant to crowdsourced and missing-person data.

 • Information Technology Act, 2000: Provides the base for data and cyber-security regulation. 

• Sendai Framework: Emphasises early warning systems and risk-informed governance. 

Conclusion AI can make disaster management faster, more anticipatory and more inclusive, but it cannot replace institutions and community knowledge. India needs reliable data systems, data privacy safeguards, local-language models and capacity building at district level. Human oversight and accountability should remain central to technology-led resilience.

 UPSC Relevance 

• Prelims: NDMA, Sendai Framework, Flood Hub, early warning systems, DPDP Act 2023, Cyclone and flood forecasting agencies. 

• Mains GS-III: Disaster management, science and technology, applications of AI, and cyber and data security.

• Mains GS-II: Governance, institutions and role of the private sector in public service delivery. 

• Essay/Ethics: Technology with a human face, and equity in technology-driven governance.