AI-EDGE: pursuing the brink of the Internet and beyond

Posted: August 25, 2021
highway

When it comes to the future of artificial intelligence, the applications are only limited to the imagination. One day a virtual personal assistant could pick your kids up from school in your self-driving car. Fully automated factories may deliver your personally designed furniture straight to your door without human intervention.

This is the vision of the new $20M National Science Foundation’s Artificial Intelligence Institute for Future Edge Networks and Distributed Intelligence (AI-EDGE) led by Ohio State and partially funded by the U.S. Department of Homeland Security.

As director of the institute, Ohio State Professor Ness Shroff said the goal is to boost technological and cloud computing network advancements decades into the future.

Ness Schroff
Ness Schroff

An Ohio State Eminent Scholar, Center for Automotive Research Collaborating Faculty and Chaired Professor of both Electrical and Computer Engineering and Computer Science Engineering, Shroff said the collective team will develop a future intelligent communication edge network by interconnecting billions of end devices providing services and solutions for the benefit of all.

“AI technologies will bring revolutions to distributed edge networks,” Shroff said, “The next generation edge networks will have distributed and intelligent agents to enable intelligent transportation, remote healthcare, distributed robotics, smart aerospace, and more.”

Most of the industry growth, he said, is expected to come with wireless devices, services, and applications at the edge rather than the traditional network core. What this means is moving computation away from traditional data centers toward the edge of the network – computers, smartphones; wi-fi, and wired access; as well as computing and data servers.

The challenge, he said, is making AI more efficient, adaptive, interactive, privacy-aware, and robust to failure and attacks.

“The Internet is the most significant example of a distributed system. And now what we want is a distributed intelligent system. If we can enable that then you can use the power of collaboration across wide geographic regions, in order to solve very large and very important problems,” Shroff said. “If you think of the next generation of the Internet, it will need to be very adaptive to your needs and to your requirements of the application.”

For example, he said, if each application requires different data rates, some applications may require instant data, whereas others may only need data at certain times.

“These networks should have the ability to satisfy all these various new requirements, as well as be self-adaptive in case part of the network breaks down, they should automatically be able to reconfigure and still give you functionality. It should also be autonomous, which basically means it does not need a whole lot of human intervention,” Shroff said.

AI-EDGE includes 30 prominent researchers, scientists, and engineers from 11 universities, four leading companies in networking and Artificial Intelligence, and three premier Department of Defense Research Labs. Collectively, the team will create new AI tools and techniques to ensure communications networks are self-healing and self-optimized. The Institute will span over two broad synergistic themes: AI for Networks and Networks for AI. One theme is to apply AI to design the next-generation edge networks that are highly efficient, reliable, robust, and secure.

Shroff said developing AI and machine learning (ML) algorithms toward networking, with minimal or no human oversight, poses many challenging research questions.

He said the team will explore four research directions: re-engineering the physical fabric for 6G and beyond wireless networks through AI, designing AI-enabled network resource allocation, developing multi-agent AI-based distributed network control, and devising new AI-powered network security techniques. 

Another end result of the institute is researching three wireless edge use cases in-depth, including real-time sensing and networking, human-machine mobility, as well as programmable and virtualized 6G-and-beyond networks. AI-EDGE then works with its industry and DoD partners to facilitate translation and adoption.

This leads to the next AI-EDGE goal – increasing the number of young engineers who know how to make this happen. The institute will develop educational content to launch programs in AI and networks spanning K-12, to university students and faculty. The impact could bolster decades of future engineers and advancements.

“It is a national priority to educate students, professionals, and practitioners in AI and networks, and substantially grow and diversify the workforce,” Shroff said.

Ohio State President and ECE Professor Kristina M. Johnson said AI advancements created by AI-EDGE engineers will improve personalized health care, augment cybersecurity, boost smart mobility, enhance food security, and so much more.

"Not only will they continue the ongoing and incredibly vital work to fully realize the potential of AI,” she said, “but they will also offer the opportunity to increase the number and diversity of individuals with the skills necessary to meet the long-term workforce needs of this important field."

AI-EDGE is one of 11 new U.S. National Science Foundation National Artificial Intelligence Research Institutes, building on the first round of seven institutes funded in 2020. The combined investment of $220 million expands the reach of these institutes to include a total of 37 states. Ohio State is among the few universities picked to launch two of those 11 institutes. The second being the AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE).

By: Ryan Horns ECE Communications Specialist | Horns.1@osu.edu | @OhioStateECE