Both the frequency and severity of natural disasters are increasing. This year alone, we have seen significant wildfires across the Western United States and in countries like Greece and Turkey; major floods across Europe; a 7.2 earthquake in Haiti, and Hurricane Ida’s impact to the coast of Louisiana. In response, governments, businesses, nonprofits, and international organizations are placing more emphasis on disaster preparedness and response than ever before. Many of these same entities are accelerating their efforts to make their data publicly available for others to use. Repositories such as the Registry of Open Data on AWS and Humanitarian Data Exchange contain treasure troves of data that is ripe for use by developers, data scientists, and machine learning practitioners.

At AWS, we believe that technology has the power to solve the world’s most pressing issues. As we build and improve new services such as Amazon SageMaker Studio Lab, we take pride in supporting the new and innovative ways our customers are using these technologies to deliver social impact. Through this hackathon, we hope to stimulate ways to apply machine learning to solve pressing challenges in natural disaster preparedness and response.

The scope of this hackathon is broad, but we have provided some parameters to guide participants’ work:

1. The solution must solve a challenge that is aligned with at least one of the phases of the disaster life cycle (mitigation, preparedness, response, recovery).

2. The solution must focus on a challenge that is present in one or more types of natural disaster (e.g. hurricanes/cyclones; wildfires; floods; earthquakes; drought; etc.).

3. The solution must use machine learning to solve the challenge. This includes the submission of a machine learning model that helps people predict, identify, categorize, prioritize etc. 

Example challenges include:

• How might we accurately and efficiently determine the extent of damage to individual homes in a given disaster-impacted area?

• How might we detect stress signals from radio communications?

• How might we predict where first-responder assets are most likely to be needed in a disaster (a) before it happens, or (b) immediately after it happens?

• How might we predict the spread of wildfires?

• Check out the inspiration page for more ideas.

The project should also focus on accessibility and an ease of use for disaster response organizations with limited technology and machine learning knowledge. 

$54,000 in prizes

• Potential Value
  Includes the extent to which the solution can be widely useful to disaster response organizations, easy to use, accessible, etc.
• Quality of the Idea
  Includes creativity, originality of the project, such as finding and using unique public datasets, or solving a challenge in a unique way.
• Implementation
  Includes how well the idea was prototyped by the developers using machine learning and Amazon SageMaker Studio Lab.