U.S. National Science Foundation AI Institute for Edge Computing Leveraging Next Generation Networks is committed to educating and developing our workforce, cultivating the next generation of technical leaders who will ensure the ethical and fair use of AI. We work with schools, districts, and other nonprofit partners to bring evidence-based activities to students all over the east coast. Together with its industrial and government partners, Athena will spearhead collaboration and knowledge transfer to maximize the translation of AI-empowered edge computing into new business opportunities.
- K-12 Activities
Athena partners with non-profit organizations to conduct activities for aspiring STEM students. Our year-round programs include Community Code with Knox Street Studios, Summer in STEM with Inspiring Minds, Academic Year Research with North Carolina A&T Early STEM College, and TechArts and Crafts with the Whitaker Project. We've reached over 900 students with more than 250 contact hours. Contact dukeliftlab@duke.edu if you're interested in partnering with us. Upcoming events are announced on X and Instagram.
- Summer Research Opportunities
Summer 2025 Project Descriptions
Daniel Limbrick, North Carolina A&T
Student researchers will investigate the impact of physical faults (e.g., soft errors, laser fault injection, voltage glitching) on microcontrollers running machine learning tasks, such as object recognition. Students will create a microcontroller-based ground vehicles equipped with cameras to emulate faults by randomly flipping values in internal memory at specific times. The goal is to understand when and how likely a single bit flip will be critical.
(First years - Seniors; Dates: 6/2/2025 - 7/25/2025)
Olga Russakovsky, Princeton University
We are looking for students interested in building computer vision systems at the edge. They will have a chance to collaborate with PhD students exploring topics such as data-efficient AI, AI systems interacting with human users, or understanding social biases in edge AI systems. They will be part of a virtual community of REU students at Athena and an in-person community of REU students at Princeton. (https://www.cs.princeton.edu/diversity-and-outreach/researchexperience) (http://visualai.princeton.edu)
(First years - Seniors; Dates: 6/1/2025 - 8/3/2025)
Miroslav Pajic, Duke University
Deep brain stimulation (DBS) is the principal treatment for patients with Parkinson’s disease (PD), whose motor fluctuations and dyskinesias cannot be sufficiently managed with medication. However, conventional DBS methods are limited by providing only a continuous, predetermined stimulation which can paradoxically lead to worsening motor functioning in patients over time. Our research objective in the Cyber-Physical Systems Lab is to develop a closed-loop, adaptive DBS system which intelligently controls stimulation parameters as a real-time response to neurological and physiological signals so that patient treatment and quality of life is significantly improved. Our collaboration with Duke University Hospital and the Biomedical Engineering department is a unique opportunity for students to build skills in applied medical research.
We are looking for curious and motivated students who are interested in one or more of the following areas: control theory, reinforcement learning / machine learning, edge computing and distributed systems. As a member of our team, you will be involved with the research and development of novel controller design, neural systems testing (in-clinic, test-bench), and edge-computing for medical devices. The REU will provide you with the opportunity to make an independent contribution to the closed-loop DBS project, and time permitting will be used for our ongoing clinical study. Additionally, we will provide you with mentorship and guidance to compile your research into an academic presentation. (https://cpsl.pratt.duke.edu/ )
(First years - Seniors; Dates: 5/2025 - 7/29/2025)
Abhishek Bhattacharjee , Yale University
We will build hardware for implantable brain computer interfaces that can help treat neurological disorders and enable next-generation neuroscience research.
(Juniors/Seniors; Dates: 7/1/2025 - 8/31/2026)
Suman Banerjee, University of Wisconsin-Madison
The students plan to undertake the following activities, subject to change:
- Assist in designing integrating LLM capabilities into tiny/mobile devices that enable efficient operations.
- Work on implementing autonomous systems, e.g., drones and robots, in navigating indoor spaces.
- Understand how computation can be efficiently split between edge systems and end systems.
- Work towards an integrated demonstration with these capabilities.(https://wings.cs.wisc.edu/easyvizar/ )
(First years - Seniors; Dates: 6/1/2025 - 8/31/2025)
Bhuvana Krishnaswamy, University of Wisconsin-Madison
As part of Athena, we are working on collecting and processing heterogeneous data from multiple sensors - a camera, some environment sensors, that are attached to drones. We aim to attach many sensors to a drone, fly to a desired location, collect data and repeat. This data is partly processed at the edge and partly at the cloud. However, different sensors have different distance resolutions and cannot capture all of this from the same height as that of the drone. In this work, students will design a mechanism to autonomously drop sensors to different heights from the drone and pull it back up once the measurement is complete. (https://uwconnect.ece.wisc.edu/)
(Juniors/Seniors; Dates: 5/2025 - 8/2025)
- Undergraduate, Masters, and PhD Research
Athena Students: Research in AI and Edge Computing
Undergraduate, Masters, and PhD students at the Athena AI Institute institutions are conducting research across several key areas of artificial intelligence and edge computing. Their work aims to address fundamental challenges in these fields and develop practical applications. Here's an overview of their research areas:
- Developing techniques to compress large AI models for deployment on resource-constrained devices
Investigating methods for dataset distillation to enable more efficient training of AI models
Image
- Exploring ways to improve the robustness and generalization of AI systems across different environments
- Creating algorithms for real-time object detection and scene understanding in augmented reality applications
- Investigating techniques for collaborative augmented reality experiences with multiple users
- Designing algorithms for situational awareness in autonomous vehicles and drones
- Developing robust decision-making systems for autonomous agents in complex environments
- Investigating methods for secure and efficient communication between multiple autonomous systems
- Developing techniques for low-latency, privacy-preserving distributed computing
- Investigating novel architectures for disaggregated memory and storage in edge data centers
- Investigating the robustness of AI systems to data poisoning and other security threats
- Developing AI-assisted systems for medical imaging and surgical guidance
- Creating edge computing solutions for real-time health monitoring and diagnostics
These research projects involve a combination of theoretical work, algorithm development, and practical implementation. Students are working with cutting-edge technologies such as deep learning frameworks, augmented reality devices, autonomous vehicle platforms, and edge computing hardware to advance the state of the art in AI and edge computing.
- Postdoctoral Research Positions
We are currently not accepting applications for 2025. Please check back for next year.
The NSF-funded National AI Institute for Edge Computing Leveraging Next Generation Networks invites applications for Postdoctoral Fellowships from early-career engineers, natural and social scientists, mathematicians, statisticians, and educators. The Athena postdoctoral fellowships are ideally suited to scholars seeking interdisciplinary and multidisciplinary opportunities to develop and apply AI to edge computing devices and systems. Our Institute is organized in four thrust areas: 1) AI-powered networking at the edge, 2) AI-powered computer systems at the edge, 3) Advancing AI for mobile networks, and 4) AI-enabled services and applications for next-generation mobile networks. Preference will be given to candidates who can facilitate cross-thrust collaborations among research groups at Athena-affiliated institutions: Duke, MIT, NC A&T, Princeton, University of Michigan, University of Wisconsin, and/or Yale. Selection will be made based on the candidate’s academic and/or industrial experience, expected impact of the proposed research, and potential to broaden participation in the Institute by women, underrepresented minorities, and persons with disabilities.
To be eligible for an Athena Postdoctoral Fellowship, individuals must: 1) have received a doctoral degree (Ph.D. or equivalent), and 2) be able to start the fellowship in residence with an Athena-affiliated research group. Initial fellowship awards will be made for one-year and may be extended. Interested applicants are strongly encouraged to contact one or more Athena investigators (listed at https://athena.duke.edu/who-we-are) to discuss their research interests prior to making an application.
To apply, please provide a current curriculum vitae, research statement to include a proposed research plan (maximum 2 pages), name(s) of potential host Athena investigator(s), and the names and contact information of three references. These materials should be e-mailed to info-athena@duke.edu.
- Knowledge Transfer
The Athena AI Institute actively engages in knowledge transfer activities with industry partners, fostering a dynamic exchange between academia and the corporate world. These collaborations take several forms, creating multiple pathways for sharing expertise and innovations. Students from the institute have opportunities to participate in internships with corporate partners, gaining real-world experience and bringing fresh perspectives to industry challenges. Faculty members are encouraged to spend summer months or sabbaticals at corporate sites, facilitating deep knowledge exchange and cross-pollination of ideas. Additionally, Athena members collaborate with corporate partners on solving problems in the public domain, ensuring that the institute's research has practical applications and societal impact. This multi-faceted approach to industry collaboration not only enhances the institute's research but also helps bridge the gap between academic advancements and real-world implementation of AI and edge computing technologies.
Image
- Athena Introduction for the Non-expert
The Athena AI Institute: Bringing Big AI to Small Devices
The Athena AI Institute is an exciting initiative that's pushing the boundaries of artificial intelligence (AI) and edge computing. But what does that mean for you? Let's break it down:
What is Athena?
Athena is part of a national effort to advance AI research and development. It brings together experts from 9 universities and 8 affiliate organizations to work on making AI more accessible and useful in our daily lives.
The Big Idea: AI Everywhere
Imagine having the power of advanced AI not just in big data centers, but in the devices you use every day. That's Athena's vision: "Big AI, Small Devices." We're working to bring complex AI capabilities to smaller devices like your smartphone or smart home gadgets.
Three Main Areas of Focus
- Edge Computing Systems: Making AI work efficiently on devices close to you, rather than in far-away data centers.
- Collaborative XR: Enhancing virtual and augmented reality experiences that multiple people can share.
- Situational-Awareness & Autonomy: Improving how AI understands and reacts to the world around it, like in self-driving cars or drone systems.
Impact on Education and Society
Athena isn't just about technology. We're also focused on:
- Educating the next generation of AI experts
- Ensuring AI benefits everyone from the research
- Partnering with industry to bring innovations to marketThe Big Picture
By bringing together experts from different fields and focusing on practical applications, Athena aims to make AI more powerful, accessible, and beneficial to society. Their work could lead to smarter devices, more immersive digital experiences, and AI systems that better understand and interact with the world around us.
As Athena continues its five-year journey, we can look forward to exciting developments that could change how we interact with technology in our daily lives.
