From inducing cybersickness to diminishing trust in equipment, cognitive attacks exploit the connection between users and mixed reality (MR) systems. In critical operations, protection against adversarial attacks is vital but not offered through current MR systems on the market. These problems are underexplored due to limited models available for reasoning about cognitive attacks and their potential impact on human behavior, engagement, and trust in MR systems. These factors prompted the Defense Advanced Research Projects Agency’s (DARPA) Intrinsic Cognitive Security program to issue a call for innovative projects to address this gap, which researchers like Professor of Electrical and Computer Engineering Tian Lan answered.
Over the next three years, Lan will support a collaborative effort between academia and industry, backed by $9M in DARPA funding, to build tactical MR systems to safeguard human users from cognitive attacks and offer verifiable cognitive guarantees. The study, “VeriPro: Verified Probabilistic Cognitive Reasoning for Tactical Mixed-Reality Systems,” is co-led by Pennsylvania State University (PSU) and GW Engineering. Additional collaborators include Northeastern University, the University of Southern California, Kennesaw State University, and industry partner Design Interactive.
“This exciting collaboration will enable us to investigate fundamental research problems in the cognitive security of MR and translate the research into significant practical impact. As co-lead of the project, GW will collaborate with PSU and other partners in building novel cognitive models and leveraging them for rigorous reasoning,” said Lan.
The team’s ultimate goal is to leverage computational science and machine learning (ML) to address vulnerabilities and improve the security of MR systems. By addressing DARPA’s dual objectives of creating cognitive guarantees and developing cognitive models for MR systems, this project addresses a critical need for advanced protective measures in an era where MR equipment’s role in critical operations is quickly expanding. Lan’s expertise in developing novel ML algorithms is a driving force in this project, with multiple ongoing projects to safeguard emerging technologies like MR underscoring his invaluable contributions.