How technology can help India cope with natural disasters
From using AI to predict floods to crowdsourcing, India is heavily using technology in its bid to cope with natural disasters such as floods and cyclones
The year 2019 has been one of the worst years in the history for India when it comes to natural disasters. More than 13 Indian states (Kerala, Gujarat, Karnataka, Maharashtra, Madhya Pradesh, Andhra Pradesh, Odisha) have been affected due to floods. Government reports state that this year has been witness to the heaviest monsoon rains to wreck havoc in India in the last 25 years. More than 1,600 people have been killed with millions of people losing their homes and their livelihood.
While this year has been devastating, every year, the same pattern repeats itself. Almost every year, we hear of an Indian state being flooded. According to the National Flood Commission report, around 40 million hectares of land in India is prone to floods.
Can technology help? In India, some pilot projects and initiatives launched by different states can serve to be the right models for other states to follow. Take, for example, the pilot project done by Google in collaboration with the Central Water Commission, to estimate the flood level situation in Patna. Google claims that its AI based model was used to send a map-based alert to people who lived within thousand square kilometers around Patna. The map indicated which areas were likely to get more flooded and which areas had the probability of not being flooded. Google claims that the accuracy of its model was over 90 per cent.
We now take a look at the initiatives of some Indian states that have used information technology
to enhance the safety of their citizens.
Odisha – a beacon of hope
Twenty years ago, when a super cyclone hit the state of Odisha (then known as Orissa), it left the state in tatters. More than 10,000 people died, entire villages were simply washed away by the cyclone and over 3.5 lakh houses were destroyed. More than 25 lakh people were stuck in places due to water logging. Cut back to this year. When a cyclone of nearly the same severity hit Odisha this year, the state managed to curb the number of casualties to only 64. This is a significant achievement for a state and has several lessons for the rest of the country.
The journey to save millions of lives for a state which is prone to disasters is not easy. The state has progressively improved over the past 20 years and heavily uses information technology to improve safety of its citizens.
Odisha is the first state in the country that has implemented an Early Warning Dissemination System (EWDS). This aims at establishing a foolproof communication system to address the existing gap of disseminating disaster warning from the state, district and block levels to communities. It covers 1205 villages in 22 blocks of six coastal districts of the state which are prone to multifarious hydro-meteorological disasters like cyclones, floods and tsunamis. This system integrates technologies such as digital mobile radio, location based alert systems, remotely operated siren systems and universal gateways. The system helps in disseminating warning communication simultaneously from the state, district and block levels in different forms like messages, voice, siren, etc. The early warning dissemination is fail proof and operates on a 24/7 basis in all circumstances.
Today, the state has set up last mile connectivity in the coastal areas of the state. As many as 122 siren towers were set up for early warning communication. States Bishnupada Sethi, Managing Director of Odisha State Disaster Mitigation Authority (OSDMA), “During the extremely severe cyclonic storm ‘Fani’, cyclone related alert and action suggested had been delivered through Location Based Alert System (LBAS). SMS to 1.8 crore BSNL subscribers of the likely affected districts including group based alert messages was disseminated based on threat status to a particular area. All messages were delivered in the local language. Besides this, early warning sirens were activated and voice messages disseminated every hour in the coastal areas. 14 satellite based voice and data terminal had also been activated for emergency communications. All cyclone bulletins and warnings messages of IMD were disseminated with key state level officers, collectors and the media.”
Due to this early warning system, the last person at the remotest corner in the coastal area could be alerted about an impending disaster in minutes. “From the state level, the entire coast of Odisha could be reached at the push of a button. Even people could be alerted about a disaster at the dead of night when they are fast asleep. Overall impact of the disaster in the coastal area considerably reduced because of the EWDS and in strategically located coastal villages,” points out Sethi, on the strategic importance of the early warning system in saving lives.
OSDMA in collaboration with Regional Integrated Multi-Hazard Early Warning System (RIMES) has also developed a web and smartphone-based platform called “SATARK” (System for Assessing, Tracking and Alerting Disaster Risk Information based on Dynamic Risk Knowledge). The application is developed to provide real time watch, alert and warning information for different hazards like heatwave, lightning, agriculture risk (drought), flood monitoring, ocean state information and tsunami risk, earthquake monitoring, cyclone/storm surge for improved disaster management. It uses different level of warnings and issues corresponding advisories based on the event scenario.
SATARK aids in provision of timely early warning information for different hazards such as lightning, heatwaves, floods, agricultural risks (droughts), etc. The system translates generic weather forecast products into user-friendly actionable advisories, based on thresholds drawn from historical patterns, to ensure effective preparedness in place to minimise the risk. The system utilises a “machine learning algorithm” to self-learn from each seasonal cycle of operation, and improving on its own advisory generation process, over season. The advisory clearly underlines the guidelines set by the state government for the user to be aware about what they need to do before, during and after a disaster event. This is outlined both in Odia and English languages. By knowing the exposed parameters, disaster managers are able to carry out effective preparedness measures for the anticipated risk and carry out well planned response activities aftermath of disaster.
The application provides location specific alerts. The system integrates model forecast data, observational data, sensor data and other available data in different format (spatial and non-spatial data) available from various national and international agencies. This application has the capabilities to integrate real-time lightning data from the lightning detection sensors installed over Odisha.
SATARK App is the mobile version of the SATARK web portal which is a one stop hub to monitor and evaluate various risks associated with different disaster forecast for Odisha. The app allows users to keep notified about various forecasts and its potential impacts, along with respective advisories based on the registered and added blocks. The app also acts as a two way communication between user and the State Emergency Operation Center, which allows the user to report regarding the accuracy of forecast and disasters in their area.
Sethi says, “SATARK is a citizen centric one stop app for information related to multiple disasters. The entire system is automated and generates advisories taking into the different models configured earlier. The system utilises a “machine learning algorithm” to self-learn from each seasonal cycle of operation, and improving on its own advisory generation process, over seasons. It is a one stop application to predict, forecast weather information and disseminate warning on multi hazards with user friendly advisories. The app is meant for disaster managers and common people and supports two-way communication. The app is a hub for real-time alerting, forecasting as well as disseminating advisories to the needy.”
Kerala uses crowdsourcing to inspire hope
In the month of August 2018, the state of Kerala was inundated with floods. Affecting over 1.3 million people, this was one of the most devastating floods that the country has ever faced. An effective information system is a vital cog in the management of ongoing flood relief operations. Since the flooding was unexpected, Kerala state did not have any information system to effectively manage the crisis.
It is in this context Kerala State IT Mission has carried out a brainstorming and decided to develop a crisis management system following the crowdsourcing method and to deploy the same in the shortest possible time, following a DevOps methodology. A crisis management platform was developed and hosted and made available to the public available at the URL www.keralarescue.in It was up and running within 12 hours after the first day of flooding. This was made possible through the contributions of techies across the globe who volunteered to code, review and recode to enhance the functionality of the platform as demanded by the need confronting the state.
Its transformation from information portal to that of a service portal with multifarious uses was applauded by all. The mere text based rescue requests posted were enhanced to capture geo-coordinates automatically and the geo-tagged information provided by the people in this portal came handy for the rescue teams during rescue operations. The state’s IT Mission was able to enhance the capabilities of this platform to cover all the relief operations. District needs were updated from time to time on the portal and was being watched by the donors all over the world.
Soon people started posting rescue requests and later on geo tagging functionality was enabled to accurately pin point the stranded location. The flood of requests received in the portal were subjected to artificial intelligence based algorithms to prioritise on the basis of key words such as elderly, urgent, pregnant, sick, ladies only, etc. Based on the prioritisation chart, the geo tags attached to the request were subjected to analysis utilising geo intelligence framework to generate cluster maps and heat maps which were handed over to the naval and disaster management authorities, control rooms who in turn provided the coordinates to naval pilots NDRF/Army/police/fire and rescue personnel who were operating in the field.
When the number of rescue requests peaked, the volunteers started calling and verifying and co-ordinating ground level activities. There were volunteer groups who had setup outbound call centres in Bangalore, Hyderabad and even in other parts of the world like a 250 member volunteering group from USA. Every request had geo-location triangulated with the help of GPS, GSM and internet of the mobile phones and the high smartphone penetration in the region acted as a boon. Using the geolocation, heat maps were generated and rescue teams on the ground was mobilised to the most affected areas. This ensured immediate evacuation and prevention of loss of life. Even in the case of phones getting switched off or the mobile towers failing, this approach helped the rescue operations by prioritising in real-time.
The IT Mission also came up with a system for proper supply chain management system available through the URL www.krsc.kerala.gov.in to organise the logistics associated with the relief material demand and supply. In addition to this there were numerous volunteer driven initiatives which created real time mapping of flood, rescue efforts, relief camps, etc.
The Government of Kerala entrusted the Kerala State IT Mission (KSITM) to set up an ICT Platform comprising of Web based backend and a mobile app-based field survey application to document the flood related damage caused to houses and commercial establishments in affected districts. KSITM has developed a volunteer registration and management system wherein the volunteers can register themselves by choosing the days and area where they can be available for the survey. The mobile app developed for the purpose (RebuildKerala), was used in the field survey of assessing damages of houses and establishments and the same gets geo-tags and captures the photos from the field which gets populated to the backend work flow in real-time if internet connectivity is available at the field or saved offline.
A real-time governance dashboard (www.rebuild. lsgd.kerala. gov.in) provided online monitoring of these activities and gives all required insights needed for the district administration. The involvement of large number of volunteers and technology platform enabling the survey provided an efficient and transparent way of assessing the damages within the stipulated timelines and to provision timely aid to the needy populace. The Kerala Government estimates that the ICT enabled platform could save over a lakh lives by saving precious time by way of receiving accurate information about the number of people stranded and their exact locations.
The Kerala State IT Mission, in association with IIITMK, also developed a unique platform to recover the lost certificates to help flood affected citizens by leveraging digital records. Equipped with computers printers and the internet, the government setup ‘certificate adalats’, and helped in accessing digital records such as driving licenses, ration cards, marriage certificates and school records. An Aadhar enabled digilocker was created to store the documents for future requirements. Further, the Government also created a service that allowed citizens to submit their applications for reissuing documents through a dedicated WhatsApp number.
M Sivasankar, IAS, Secretary, Department of Electronics and Information Technology, Government of Kerala has publicly lauded the efforts of every volunteer in collaborating and contributing to building the technology solution for Kerala.
Drones and social media
In 2015, the social media platform, Twitter, was used by a number of government groups and people to share vital information (helpline phone numbers, train schedules, relief counters, weather forecasts, etc) about the Chennai floods on Twitter. This became a test case for Twitter, and showed government agencies on how social media platforms could be leveraged for effective communication related to natural disasters. During the 2013 Uttarakhand floods, drones were used to locate missing people and scan the terrain to provide relevant updated information to the authorities. Recently, students from IIT Madras developed an AI-enabled drone that can help authorities provide vital information on people trapped in disaster-hit areas.
Tamil Nadu has built a web GIS based system called TNSMART. This application, which is developed in collaboration with ISRO, has modules related to thresholds, hazard forecast, disaster impact forecast, advisory, response planning, etc. Similarly, Karnataka has a GPS enabled system for near real-time monitoring and communication of disasters in the state. In India, the Government has encouraged the use of digital technologies in ensuring help during disasters. For example, the Digital India Action Group (DIAG) recently released a whitepaper on using IoT for effective disaster management.
India as a country is prone to natural disasters. However, as the state of Odisha has shown, continued and sustained improvements accompanied by use of relevant technologies can help in saving lives. The Government of India has also decided to setup an integrated control room for emergency response to minimise information gaps between federal and state agencies during disasters.
With each disaster, there are numerous lessons to be learnt. India needs to document best practices that were observed during each disaster and ensure that these are captured in the form of a knowledge management platform, coupled with modern e-learning tools, so that each state can learn from each other and put in place systems that are ready for handling disasters of the future!