SCS Professor Earns Air Force Award to Enable Machine Learning in Drones

Contact

Devin M. Young

Communications Assistant

Sidebar Content
No sidebar content submitted.
Summaries

Summary Sentence:

SCS Assistant Professor Hadi Esmaeilzadeh receives a highly respected Young Investigator Research Program grant to conduct research on drone reliability and autonomy via machine learning and probability algorithms.

Full Summary:

No summary paragraph submitted.

Media
  • Hadi Esmaeilzadeh_Stairs Hadi Esmaeilzadeh_Stairs
    (image/jpeg)

Georgia Institute of Technology School of Computer Science (SCS) Assistant Professor Hadi Esmaeilzadeh is receiving one of 58 grants totaling nearly $20.8 million from the Air Force Office of Scientific Research’s (AFOSR) Young Investigator Research Program (YIP).

Esmaeilzadeh’s YIP award, approximately $360,000 over three years, will support his research to develop a novel many-core architecture for machine learning and control algorithms in drones and autonomous vehicles.

“This is a significant accomplishment for Hadi,” said SCS Chair Lance Fortnow. “It will allow him to move forward with this important work and it furthers Georgia Tech’s reputation for groundbreaking research.”

Esmaeilzadeh’s research proposal, “Accelerated System Design for Perception and Control in Energy-Constrained UAVs,” was selected from among hundreds of proposals submitted to make the final list of funded projects.

Esmaeilzadeh’s primary research areas include computer architecture, programming languages, and machine learning. These research areas are key to pushing past current limitations of the existing generation of drones and unmanned vehicles.

“The use of drones for all types of missions has increased dramatically during the past decade,” said Esmaeilzadeh. “However, a lack of true autonomy, short battery life, and the need for constant communication make them vulnerable.”

To counter these shortcomings, Esmaeilzadeh will conduct basic research to develop a unified, cross-stack solution that will help make drones and similar autonomous vehicles smarter, more energy efficient, and more self-sufficient. This research plans to merge software and hardware to accelerate machine learning and stochastic control via a novel many-core accelerator, allowing unmanned vehicles to better perceive their surroundings and make more informed decisions. Other improvements would include extended flight time and a reduced need for human guidance.

The grant will also support Esmaeilzadeh’s research in field programmable gate arrays (FPGA), which have shown promise in improving the efficiency of machine learning. Esmaeilzadeh received recognition for his FPGA research earlier this year during the 22nd International Symposium on High-Performance Computer Architecture, where he and his fellow researchers earned the “Distinguished Paper Award.”

The prestigious YIP grant is given to scientists and engineers at research institutions who have received their Ph.D. or equivalent degrees in the past five years and that demonstrate “exceptional ability and promise for conducting basic research,” according to the award’s webpage.

Additional Information

Groups

College of Computing, School of Computer Science

Categories
No categories were selected.
Related Core Research Areas
National Security
Newsroom Topics
No newsroom topics were selected.
Keywords
No keywords were submitted.
Status
  • Created By: Devin Young
  • Workflow Status: Published
  • Created On: Nov 7, 2016 - 1:50pm
  • Last Updated: Nov 10, 2016 - 1:00pm