*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Atlanta, GA | Posted: June 13, 2017
Summary Sentence:
A team of researchers from the Schools of Aerospace Engineering and Interactive Computing has received a $100K grant to further its work on autonomous driving
Full Summary:
No summary paragraph submitted.
Prof. Evangelos Theodorou
James Rehg compressedA team of Georgia Tech researchers headed up by School of Aerospace Engineering professor Evangelos Theodorou and School of Interactive Computing professor James Rehg has been awarded a $100,000 fellowship by Qualcomm for its proposal, “Autonomous Racing Using Deep Learning and Game Theoretic Optimization.”
The GT proposal is one of eight nationwide that were chosen for the 2017 fellowship, which also includes a one-year mentorship by Qualcomm engineers.
Theodorou says the innovation fellowship will help him, Rehg, and graduate students Grady Williams (College of Computing) and Paul Drews (School of Electrical and Computer Engineering) to bring their research to place where it will have a transformative impact in the transportation industry.
“Autonomous driving is one of the most important sub-fields in robotics,” said Theodorou. “However, autonomous vehicles driving hundreds of millions of miles are likely to get into situations where it is necessary for them to perform aggressive maneuvers to avoid collision. Our work can have an impact on that.”
The team’s work focuses on the problems faced by two or more autonomous racing vehicles in an environment that has not been previously mapped out. Potholes, bumps, and other irregularities are expected, but cannot be precisely predicted at the onset. Any system seeking to travel over such terrain must be able navigate new decisions on the fly. Each racing vehicle is necessarily pushed to its handling/acceleration limits, a condition that requires even more simultaneous sensing of the environment and other intelligent agents.
“There is only a small margin of error on both the control and perception side when racing against a capable adversary,” said Theodorou. “This research will address fundamental questions in autonomy by bringing together concepts on stochastic optimal control, game theory and deep learning. "
