The dismal state of disaster preparedness in the US and how data science can help

FEMA has released a January 2019 report entitled: "Building cultures of preparedness: report for the emergency management higher education community". The purpose of the report is to promote a new culture of preparedness approach to disaster preparedness, based on local trust relationships and institutional knowledge. But what really stands out is the dismal state of community and family/individual preparedness in the U.S. currently. The report starkly concludes that all prior efforts at encouraging preparedness such Ready.gov have failed, and Americans are no more prepared now than they were in 2007. Specifically:

14% saw preparedness as a "part of life" and were more likely to have started preparing or have been prepared for an extended period of time.

21% were "working on it" and have taken some preparedness actions but perceived low levels of risk.

18% said preparedness was "on their mind" and perceived themselves to be at high risk, but were unlikely to act on fear and take preparedness actions

46% said preparedness was "not on their radar".

The approach laudibly recommends engagement between higher education institutions, local cultural groups, emergency managers, and so on, catalyzed by cross-cutting "culture brokers" to build resilience and preparedness. It's unclear if any of this will succeed where past attempts have failed.

I think a broader question is exactly what kind of response from the public we should be trying to encourage right now, and how the priorities for preparedness are changing. Natural disasters are increasing in intensity and frequency, and FEMA has clearly indicated that it cannot respond proportionately. In addition to natural disasters, communities and individuals face increasing threats from cyberattacks and potential infrastructure failures, as well as growing domestic and international geopolitical challenges.

Here is an idea. Data science, AI and machine learning are having a transformative effect on all industries and walks of life (I know because I'm a Professor of Data Science, among other things). I am sure we can harness these capabilities to have approaches to preparedness we haven't even thought about yet.

For example emergency managers have long carried out periodic hazard analysis to prioritize the risks to their communities so they can mitigate and prepare for the most severe threats. Why can't we do this for everyone? We could build apps and tools to help people really understand the specific, customized-to-them risks they face, across health, financial, natural disasters and other events, and driven by big-data risk analysis and AI. Risk models can be complex, but modern data science is up to the task. Further, apps could then prioritize steps people can take to mitigate their specific risks. We could even gamify it, with rewards for taking preparedness actions.

Data Science for Preparedness is something higher education, startups and the emergency management community can come together for, but financial support will be needed to make it a success, either from government funding or far-visioned investors.

Also posted on AllHazards.Net and Linkedin.