Computation to the rescue! Student teams use modeling and analysis to optimize disaster relief
IACS is pleased to announce the first IACS Computational Challenge.
As part of the second during Harvard's winter recess in January 2012, student teams will compete to solve a real-world humanitarian problem by applying modeling and other computational tools.
After an earthquake or other natural disaster, cities are often littered with debris that blocks main arteries and other roads, keeping rescue teams from reaching victims while restricting access to supply depots, hospitals and other vital infrastructure.
The goal of this exercise is to harness computational tools to optimize the actions of cleanup crews after such an event. Harvard graduate students are invited to join teams to tackle this challenge during January @ GSAS, Jan. 9-20, 2012. Cool technology prizes will be awarded for the winning solution.
Students will be organized into two or more teams depending on the number who register. Team membership will be adjusted to balance computational, modeling, project management and presentation skills, GIS (Geographic Information Systems) knowledge etc. Each team will be coached by a senior researcher and will have access to meeting space and computational resources provided by SEAS.
A scenario based on an actual catastrophe will be provided to the groups. Most of the data will be in GIS format. The information will include details about roads, intersections, major supply depots, hospitals etc. in a city. In this scenario, only partial knowledge about the distribution of the debris is available to the rescue teams or the orchestration teams; only as intersections are opened can the full extent of the debris be known.
As in a real disaster, teams will cope with limited availability of resources such as vehicles and workers to clean up roads.
The goal is to maximize the satisfaction factor, defined as the sum of all people with access to food, water, hospitals and at least one of the “exits” of the city.
The competition is being organized for IACS by Özlem Ergun, Visiting Associate Professor of Applied Mathematics, and Pavlos Protopapas, Lecturer in Computational Science and a Research Associate at the Harvard-Smithsonian Center for Astrophysics. Ergun is a Georgia Tech faculty member who specializes in applying algorithmic and analytical tools to real-world problems.
How to sign up
Registration is closed.
Who is eligible?
Any Harvard graduate student who can meet the requirements below. Undergraduates who can meet these requirements and will be in the area for other reasons (not requiring campus housing) may apply.
What is the required time commitment for the Computational Challenge?
Students must commit at least 36 hours of time during the competition, Jan. 10-20. They should have approval of their advisors in making this commitment.
How will the challenge be judged?
A diverse panel of judges, organized by IACS with the assistance of the Harvard Humanitarian Initiative, will listen to team presentations at the concluding event on Jan. 20 and will review submitted solutions. Each team will have three hours to run their final code to produce a solution for judging. The winning team will be announced the week after the competition ends (the first week of classes for spring term).
What software and hardware will the teams use?
Teams will choose their own tools. The compute resources available are described here: https://ac.seas.harvard.edu/display/USERDOCS/Overview+of+Compute+Resources
Can I enter with a friend?
Sure. Just note in your application that you would like to have your friend on your team. We reserve the right to adjust team membership to make sure all teams have a good mix of expertise.When will the teams receive the challenge?
Background information will be provided over the holiday break in order to help everyone prepare. The data and computational resources will become available on Jan. 10.
What can I win?
We’re working on the prizes. Whatever they are, they will be cool. And you will have the satisfaction of helping advance the use of computation and optimization for humanitarian relief.