Riskmap: Real-time Flood Information Sharing Platform

Riskmap is a real-time platform for sharing time-critical flood information between residents and government agencies.

Globally, over 360 million residents live in coastal areas that are less than 10 meters above sea level leaving them vulnerable to annual flooding and storm surges. With sea level rise and more frequent extreme weather events, people living in these coastal regions will face increasing flood impacts in the coming years. In coastal cities that are growing rapidly, flood conditions can change dramatically and infrastructure failures can make it challenging to predict flood outcomes.

Many coastal areas also rely on tourism as one of their primary economic drivers and yet it is challenging to provide critical information to transient visitors. In an era of uncertainty, how do we best protect people in these coastal communities?

In any flooding or extreme weather event, the community is the first line of support and information is the most critical asset. Riskmap harnesses the power of citizen reporting and social media to map flood information without needing to install any new applications or training. Currently operating in three countries - Indonesia, India and the United States - this platform connects residents, who often have the best localized information, with emergency managers to drastically cut down on response times. Residents can share time-critical information such as flood location, depth, a photo and description using an automated chatbot. These reports are then mapped on a live public web-map informing fellow residents about dynamically changing situations in the city and help each other navigate to safety. Riskmap is language agnostic and residents can share reports using Twitter, Facebook, and Telegram. The public map also visualizes real-time sensor information, such as flood gauges, monitoring wells and pumping stations. We work with local government agencies to develop a custom-built interface in-line with existing response protocols to add alerts and updates on the map. By connecting the residents and emergency management officials, this platform helps in civic co-management of disaster, allowing a two-way dialogue to generate a more detailed standard operating picture.

Petabencana.id - Indonesian instance of the platform was widely used during high-intensity rainfall event in Jakarta on February 21, 2017, when over 300,000 residents used the website to navigate the flooded city. The public map was also embedded in the Uber app for drivers to help them travel safely through the city. Similarly, in Chennai during a high-intensity single day rainfall event on November 2, 2017, the Riskmap.in website received 1152 page views a minute at its peak and 111,808 page views in 24 hours.

We are integrating real-time weather data and leveraging machine learning to help us accurately predict the affected areas and expand the reporting through automated outreach. Additionally, we are developing locational alerts to help inform residents to prepare for the flooding or storm. We firmly believe that by combining social and machine intelligence, riskmap can effectively plug the gaps between traditional command and control models of disaster response. Ultimately, all the fine-grained data gathered from this platform will aid in better decision making before, during and after flood events.

Mobile communication technologies and social media have brought in fundamental change as to how communities organize and coordinate in emergencies and natural disasters. However, disaster management agencies and local government have been slow to adapt, to engage with, and support, local modes of resilience and response. Riskmap platform is an open source framework for time-critical urban data. It is developed to improve information sharing to democratize decision-support, using the power of digital maps and social media; it enables open access to time-critical information so that everyone can make safety decisions and coordinate both individual and collective responses during disasters

During disasters, communities are the first line of defense and internet-connected smartphones, and social media networks effectively act as a catalyst for the creation of informal networks for coordination in response to natural disasters. By combining social and machine intelligence, Riskmap allows such networks to be a part of a formal disaster response by providing information via automated chatbots on facebook, twitter, and direct messaging platforms. Our web-based geospatial platform lowers the barrier for entry, allowing anyone with a basic smartphone to share time-critical information without any need for specialized training or installation of an app.  

Riskmap is being developed as an open-source platform, over next 12 months, with a focus on scaling the project in Japan and Mexico. The team is focusing on extending the functionality of the platform to evacuation preparedness/ shelter navigation, power outages as well as post-disaster assessment after hurricanes. In the next 12 months, the project aims to cover rapidly developing megacities in South Asia such as - Mumbai, Bangalore and Kolkata. Furthermore, we are actively developing smart geofencing algorithms to drive more real-time data collection and use social media for effective disaster outreach. 

We are developing Riskmap as an open source project. Within the next couple of years, we hope to build a dedicated global community of developers around the solution. For example, in the case of Petabencana.id – Risk map instance in Indonesia  - we are focusing on capacity building by creating a local nonprofit - Yayasan PetaBencana - who can then take forward the day to day running and maintenance of the platform. We aim to emulate similar strategy and develop an operating model that allows growth of context-specific solutions built around a common framework by interdisciplinary teams.

Informal community networks - operating on top of social media and instant messaging applications - have been extremely effective at hazard reporting, and response coordination during multiple events and different natural disaster types. Riskmap relies on the network effect of the social media as well as local partners during disaster events. Instead of developing a standalone application, Riskmap utilizes chatbots within an existing application such as Facebook Messenger, Telegram, and Twitter to reach the intended audience and leverage the stickiness of these platforms. We also use organic and automated outreach through social as well as traditional media to spread awareness.

Riskmap is currently operational in 4 cities in Indonesia. It was piloted in Chennai, India and Broward County, Florida. During its deployment over past one and half years, about 5600 confirmed reports have been submitted - a majority of them during intense rainfall events. During the same period, over 462,304 unique users have used the map to check the latest flood information. By integrating localized knowledge from a variety of sources into a single, robust platform, Riskmap provides a comprehensive overview of disaster events, enabling residents, humanitarian agencies, and government authorities to make more informed decisions during emergencies.

At present, Riskmap is operationally available to about a population of over 50 million in three countries. Over the next year, we are planning to scale and expand the project to include coastal megacities in India and Japan. The reach of the solution is directly dependent on the penetration of social media as well as the severity of the event. During a severe flooding event, on an average, the platform receives about 1000 confirmed reports and over 300,000 unique map views within 24 hours.

Riskmap is being developed by Urban Risk Lab - an interdisciplinary organization of researchers and designers at Massachusetts Institute of Technology. We develop methods and technologies to embed risk reduction and preparedness into the design of cities and regions to increase the resilience of local communities. As designers, we see the ability to synthesize contributions from many different disciplines as one of our most important strengths. Riskmap platform requires knowledge from many different areas: urban planning, UX design, sociology, software development and emergency management. With our current interdisciplinary team, we also draw from the incredibly innovative ecosystem at MIT. 

Our primary motivation is to develop cutting-edge disaster response tools that save lives and make them available to residents and governments. It is of utmost importance to us that these tools be transparent, scalable, reduce ownership costs and be open to democratic scrutiny. Our current development of the Riskmap platform is supported through grants under GPL v3 open source license. 

In the long-term, each country-level instance of the riskmap will be supported by creating a local non-profit - who can then take forward the day to day running of the platform. This nonprofit will be funded through maintenance agreements and grants through the government agencies as well as generate revenue by developing context-specific solutions on top of riskmap data for stakeholders in insurance, planning, and infrastructure. With a distributed network of partners, any new feature developed for one part of the world will be available in other deployments, making it easier to share strategies for risk reduction.

With its global community of private, public, and nonprofit leaders, Solve will help us accelerate development and growth of the Riskmap to impact coastal communities across the globe. By enabling partnerships with public and private stakeholders, we aim to reach a wider audience and build an active community around open and scalable solutions for disaster response. Funding from partner foundations will help us to scale our solutions to resource-constrained areas of the world that are already being impacted by changing climate, sea level rise and intense rainfall events.

Traditionally, disaster response and recovery is a domain of governments, and lack of protocols to integrate citizen-generated information into the response frameworks make it challenging to initiate partnerships. Important sources of funding are typically only available post-disaster and must also be spent in a short window of time. Lack of resources available to disaster management organizations and fragmented responsibility among agencies are some barriers for adoption. Solve will be instrumental in connecting the team with stakeholders in public as well as private sectors, to the larger open source development community and highlight the importance of citizen-data in disaster response.