This month, DataScience.com received support from Amazon Web Services (AWS) through the company’s Public Datasets program that, for two years, will cover the cost of hosting datasets for the Transportation Data Challenge, an initiative led by the National Science Foundation (NSF)’s Regional Big Data Innovation Hubs in collaboration with DataScience.com and partners.

The challenge is a collection of “data science for good”-themed hackathons, training sessions, and community problem-solving events held around the country to spur data science projects that advance transportation safety. With traffic deaths across the United States increasing 14 percent over the last two years, cities are introducing data-driven measures that help save lives such as road diets, which are designed to reduce pedestrian deaths by limiting the number of lanes in high-traffic areas and repurposing that space for features like sidewalks, bike lanes, or center turn lanes.

Through the AWS Public Datasets program, DataScience.com will make the following datasets, and several others, available to challenge participants:

  • Traffic fatality data from the National Highway Traffic Safety Administration (NHTSA)
  • Highway Performance Monitoring System (HPMS) data from the Federal Highway Administration (FHWA)
  • Data from 4,000 traffic monitoring stations across the U.S. managed by the FHWA
  • Roadway weather data from the Department of Transportation

“The AWS Public Datasets program exists to support initiatives like the Transportation Data Challenge that use data to improve communities,” said Jed Sundwall, Open Data Global Lead at Amazon Web Services. “We look forward to working with DataScience.com to host these datasets. By making these datasets available through the AWS Public Datasets program, we can help lower the cost of research, accelerate innovation, and democratize access to important data and tools.”

As part of the collaboration, AWS also awarded the NSF’s West Big Data Innovation Hub credits through its Cloud Credits for Research program to provide computing power for the DataScience.com Platform, which challenge participants are using — in conjunction with other technology donated by challenge partners — to analyze large transportation datasets and identify life-saving trends. Previous beneficiaries of the Cloud Credits for Research program include Stanford University’s Archaeology Center and The International Centre for Radio Astronomy Research, among others.

amazon-transportation-map-meetup.jpg

Both the compute resources and hosted datasets were used during a meetup at DataScience.com headquarters on October 12, “AI for Good: Big Data Challenges for Disaster Response and Recovery.” The event, a recording of which can be viewed here, is aimed at providing relief for communities affected by recent storms in Puerto Rico, Texas, and Florida. Participants contributed to the Humanitarian OpenStreetMap Team and the OpenStreetMaps Foundation, projects that make geospatial data available to first responders to help them reach those in need during catastrophic events.

“Here at DataScience.com, we want to not only enable enterprise data science work, but also to use our expertise to give back to our community and improve collaboration between government agencies and the public,” said Dave Goodsmith, managing data scientist at DataScience.com, who cohosted the meetup with West Big Data Innovation Hub Founding Executive Director Meredith M. Lee. “We’re very excited to that this project has been on the receiving end of generous support from Amazon Web Services and that we’ve been able to work closely with their open data team. These resources will go a long way in helping us inspire more people to use big data and machine learning for good and create data models that will have lasting positive effects on transportation safety.”


To learn more about the Transportation Data Challenge, please visit the National Science Foundation’s Big Data Regional Innovation Hubs’ website.


 

Brittany-Marie Swanson
Brittany-Marie Swanson

Web marketing manager at DataScience.com.