Create a solution in response to wildfires while keeping sensitive data safe
Wildfires are increasing in both intensity and frequency. How can we use smart technologies to track them and alert people in at-risk areas?
Our planet has entered an era in which natural disasters and humanitarian crises are inevitable occurrences on every continent. In 2018 alone, there were major disasters from the deadliest wildfire in California’s modern history, Hurricane Maria in Puerto Rico and Hurricane Michael in Florida, tsunamis in Indonesia, and a severe flood in Nigeria.
Dealing with the complexity of community needs and coordinating layers of emergency response simply cannot be solved with a single solution, but rather with a portfolio of integrated solutions. These combined solutions can work together to identify disasters and victims, respond to requests for help, cope with the aftermath of a disaster, and finally begin infrastructure and community recovery. Without a vision of integrated solutions between smart technologies and artificial intelligence (AI), disaster responders will face disjointed individual solutions that collectively may not address the grand challenge. In other words, the whole may be less than the sum of its parts. The challenge of this year’s Call for Code global challenge is how we can build a portfolio of smart technology, internet of things (IoT), and AI to build the capacity and to unify disparate information sources, all while protecting sensitive data and keeping information secure.
Using smart technology to leverage big data
In the context of disaster recovery, smart technologies enable the leveraging of big data, which often involves sensitive data. All smart technologies rely on integrating newly generated and mapped data with existing data. For instance, after a wildfire is under control, the need emerges for rapid mapping of affected areas and the extent of damage, often at individual household levels. AI and machine learning (ML) can bridge this knowledge gap by coalescing data from autonomous mapping drones with textual, visual, and geoinformation gathered from active or passive crowd-sourcing. The challenge is to develop technologies for rapid mapping while keeping sensitive data protected in the cloud. Different data gathering techniques yield heterogenous data, but smart technologies use AI and ML to assess reliability and quantify uncertainties, without putting sensitive data at risk.
Applying smart technology to identify wildfire triggers
Before the disastrous Camp Fire, Noah Diffenbaugh, professor of earth science at Stanford University, predicted an increase in wildfire risks due to trends from higher temperatures and a drier climate (The Independent, United Kingdom, July 31, 2018). In the case of wildfires, the question is whether we can proactively detect a nascent wildfire trigger before it is out of control. Smart technologies, drones, machine vision, AI, and ML integrate and frame data to proactively identify hazards. It is also possible to better integrate the data obtained from these technologies from human activity data, such as social media information, aerial drone detection, and video integrations. With development, AI, ML, and data analytics can map potential hazards to enable proactive mitigation, all while keeping personally identifiable data protected.
See what one of the runner-up teams did with wildfires and machine learning from last year’s Call for Code challenge.
Smart tech to coordinate efforts of government, NGO, and volunteer responders
Responders may be professionals and volunteers; employees of governmental agencies or humanitarian organizations; civilians or members of the military. A disaster triggers aid from multiple organizations, but often with suboptimal coordination among efforts and stakeholder groups. In this complex layering of stakeholder groups, personal and smart technologies facilitate coordination of logistics by sorting out “who does what.” Smart platforms must also be protected during such turmoil. Smart technologies can serve as active hubs during the emergency and in its aftermath, allow first responders to log in to learn of the most critical needs to provide evacuation assistance, household needs, water, or road clearing. Automatic net notifications have significantly improved coordination of international humanitarian crises – all powered by cloud technologies, such as those offered via IBM. Development of AI and ML-based systems could aid in identifying responders and resources, mapping them to actual needs on the ground, all while supplying technical solutions for coordinating this information in real time. Imagine the power of such technological advances and how coordination could be significantly improved.
Reduce inequalities in aid to vulnerable populations
In addition to the complexities outlined above, there is another factor to consider. Vulnerable populations are often overlooked by high-end solutions and natural disasters often leave vulnerable populations even more exposed. The elderly, disabled, or those with very low incomes are less likely to have easy access to technologies, such as mobile phones or connection to the Internet, which can hinder how smart technologies can help them during a chaotic time. According to a recent UN report and this journal publication from Christina Muñoz and Eric Tate, those who were in one of those vulnerable categories were also not provided an equal distribution of aid, whether it be financial or otherwise. For such populations, technology has the potential to serve as an empowering force to reduce social inequality, all while still securing sensitive data. For instance, by monitoring and analyzing social media, responders can detect signs of distress among the elderly, low-income households, and disabled persons. Or, if access to technology is not so readily available, technologies like Project Owl’s “ducks” allow anyone to log into the OWL emergency network and connect to provide feedback on their status.
Technologies for disaster relief should be human-centric, not necessarily technology-centric, with solutions focused on identifying and prioritizing efforts to lessen hardships on distressed populations, while not exploiting the need to protect personally identifiable information. Technology has the power to lessen the gap between satisfied and unmet needs of these population segments.
As disaster responses increasingly rely on emerging technologies, it is essential to remember that the core of the response is data, and often involves the most sensitive data available at individual levels. Foremost in the advancement of disaster relief technology must be the protection of data. Losing critical data during events would exacerbate vulnerabilities and deteriorate disaster response. With proper data security and encryption, smart technologies enable a more intelligent disaster response to respond quicker to disasters and rapidly map disaster scenarios.
IBM Cloud Hyper Protect Services
IBM Cloud Hyper Protect Services can help you create and implement highly secure AI, ML, and data analytics solutions that empower developers and data scientists like you to build cloud applications using a portfolio of cloud services powered by IBM LinuxONE.
Enter your submission
Are you interested in addressing natural disasters through technology? IBM is a founding partner in Call for Code, a global coding competition that asks software developers, data scientists, and technologists to build sustainable solutions that address natural disaster preparedness, response, and recovery.
With submissions now open, Call for Code 2019 is asking you to accept the challenge to innovate like Ali Mostafavidarani, Ph.D.
Ali Mostafavi is an Assistant Professor in the Zachry Department of Civil Engineering at Texas A&M University. Ali Mostafavi supervises the Urban Resilience, Networks, and Informatics Lab. His research focuses on analyzing, modeling, and improving network dynamics in the nexus of humans, disasters, and the built environment to foster convergence knowledge of resilient communities. His research also focuses on integrating human and machine intelligence for smart disaster response through artificial intelligence. His review on social media providing critical information during a disaster touches on how to effectively acquire disaster situational awareness information, support self-organized peer-to-peer help activities, and enable organizations to hear from the public. He has received various awards and honors such as the NSF CAREER Award and Early-Career Research Fellowship from the National Academies Gulf Research Program.
This year’s challenge is specifically focused on healthcare, access to medical health records, the vulnerable, and more. Read the CTO’s letter to developers to understand this year’s focus.
- Extreme weather is here to stay: How developers can help
- Call for Code runner-up: Project Lali predicts and detects wildfires with sensor networks
- Code pattern: Predict wildfire intensity
- Survey wildfire-damaged neighborhoods to identify burned homes and intact homes
- Predict wildfire intensity using NASA satellite data and machine learning
- Ali Mostafavi’s study on “Social media for intelligent public information and warning in disasters: An interdisciplinary review”
For more information, visit developer.ibm.com/callforcode.