Engineers for Exploration

Project Details

Project Lead
Sameer Tilak 
Project Manager
Sameer Tilak 
Project Members
Antonella Wilby, Eric Lo  
Institution
UCSD, Calit2, UCSD  
Discipline
Ecology (608) 

Abstract

Engineers for Exploration (http://e4e.ucsd.edu) is a one of kind program that matches engineering students with scientific collaborators from domains including archaeology, biology, ecology, and oceanography. The projects aim to develop new technological solutions for the problems faced by the scientists. These include create high resolution topographic maps of important ecological areas, developing 3D models of ancient artifacts, finding and tracking endangered animals, and understanding the behaviors of captive animals in zoos and wildlife.

These projects require a substantial amount of computation. For example, we heavily utilize a technique called "structure from motion" that stitches hundreds of images taken from aerial robotic platforms into a topological model. Or we combine billions of points from close to one hundred LIDAR scans of tunnels and tombs underneath Maya temples into a coherent and accurate 3D model. Finally, we use advanced digital signal processing to post process videos captured at zoos from around the country in order to develop machine learning techniques that can subsequently be utilized to classify the animals behaviors.

Access to the FutureGrid infrastructure will allow us to run significantly more experiments. And those experiments can individually handle more data. The end results are more accurate models and techniques which ultimately enable our scientific collaborators to better understand our environment and our cultural heritage.

Intellectual Merit

These will develop technology to create high resolution topographic maps of important ecological areas, developing 3D models of ancient artifacts, finding and tracking endangered animals, and understanding the behaviors of captive animals in zoos and wildlife.

Broader Impacts

Access to the FutureGrid infrastructure will allow us to run significantly more experiments. And those experiments can individually handle more data. The end results are more accurate models and techniques which ultimately enable our scientific collaborators to better understand our environment and our cultural heritage.

Scale of Use

We will need access to multiple nodes (approximately 16-32) of a cluster.