In an increasingly connected world where edge devices play a crucial role in our daily lives, ensuring both their intelligence and security is paramount. Hokeun Kim, an assistant professor in the School of Computing and Augmented Intelligence at Arizona State University, has been awarded a grant from ATTO Research to tackle this challenge head-on.
Kim’s research focuses on developing a federated learning toolset tailored for edge devices, such as smartphones, doorbell cameras, and fitness trackers. These devices, often overlooked in terms of security, are now storing significant amounts of sensitive data locally. Kim’s goal is to leverage artificial intelligence to enable these devices to learn from user data while safeguarding privacy.
“Edge devices were historically fairly secure, as they primarily transmitted data to centralized data centers for processing,” explains Kim. “But with advancements in technology, these devices are now more powerful and storing sensitive data locally, making them vulnerable to security breaches.”
The proliferation of AI-capable edge devices further complicates the security landscape. Even small single-board computers like the Raspberry Pi can run sophisticated AI models directly on the device. Kim’s grant aims to address these challenges by creating a middleware solution that empowers edge developers to build smart and secure devices.
Central to Kim’s approach is secure federated learning, a form of AI where edge devices learn autonomously and share insights with other devices without exposing raw data. By designing a platform for secure federated learning, Kim aims to equip developers with the tools needed to harness AI’s power while preserving data privacy.
“We’re not reinventing the wheel; instead, we’re integrating existing security standards and cryptography protocols to enable efficient federated learning on edge devices without compromising security,” says Kim.
The potential applications of secure federated learning are vast. For instance, in healthcare, edge devices could analyze patient data locally and share insights with medical professionals without compromising patient privacy. Kim’s work also extends to smart city infrastructure and law enforcement, where edge devices play a crucial role.
Ross Maciejewski, director of the School of Computing and Augmented Intelligence at ASU, emphasizes the importance of Kim’s research in enabling edge devices to leverage AI while safeguarding sensitive data.
