Military leaders must make timely, high-quality decisions on a regular basis. That process will become increasingly challenging in the fast-paced and dynamic environment of cyberspace and the broader information domain.
To help future officers to prepare for these challenges, Lt. Col. Nathaniel D. Bastian, academy professor, division chief, Data & Decision Sciences (D2S), senior research scientist and chief data scientist at the Army Cyber Institute (ACI), and assistant professor in the Departments of Systems Engineering and Mathematical Sciences; and Dr. Robert Thomson, cognitive scientist and senior research scientist at ACI and associate professor in the Department of Electrical Engineering and Computer Science (EECS), who teaches and conducts research in the Department of Behavioral Sciences and Leadership (BS&L), have received $1 million of a $6.4 million four-year award from the Defense Advanced Research Projects Agency (DARPA) Information Innovation Office (I2O) Assured Neuro Symbolic Learning and Reasoning (ANSR) program for their proposal titled “TrinityNS: Trustworthy, Robust and Interpretable Neuro-Symbolic World Modeling and Reasoning” to work with collaborators from SRI International, Carnegie Mellon University (CMU) and University of California Los Angeles (UCLA) to research, design and develop a Common Operating Picture (COP) to support risk-aware decision-making and enhance warfighter situational understanding through deep and iterative integration of symbolic reasoning and deep learning artificial intelligence (AI) methodologies.
Bastian is primary investigator (PI) and Thomson is co-PI on the project, joined by ACI D2S Division members and researchers within the ACI Intelligent Cyber-Systems and Analytics Research Lab (ICSARL), including Maj. John Pavlik, ACI software systems engineer and senior research scientist and assistant professor in EECS; Maj. Iain Cruickshank, ACI operations research and AI scientist and senior research scientist; Capt. David Bierbrauer, ACI data scientist, data engineer and research scientist and instructor in the Department of Mathematical Sciences; Dr. Alexander Berenbeim, ACI AI/ML researcher; Dr. Pooja Rani, ACI postdoctoral research fellow; Ms. Emily Nack, ACI Information Technology specialist; and Mr. Brandt Davis, ACI software developer.
SRI research key personnel are Dr. Susmit Jha, technical director of Neuro-Symbolic Computing and Intelligence Research and primary investigator, and Dr. John Rushby, program director of Methods and Dependable Systems, Computer Science Laboratory.
The UCLA research key personnel are Dr. Paulo Tabuada and Dr. Mani Srivastava, professors of Electrical and Computer Engineering at the Samueli School of Engineering.
The CMU research key person is Dr. Corina Pasareanu, principal scientist at CyLab and a courtesy appointee in Electrical and Computer Engineering.
Lt. Col. Bastian elaborates on how the DARPA project supports Secretary of the Army Christine Wormuth’s priority to use data to leverage emerging technologies that will prepare the force to defend against cyber threats to weapons, networks and data.
“This research effort enables U.S. Military Academy faculty and cadets to help achieve this Army priority through the development of an assurance-centric learning and reasoning framework that can learn efficiently with better generalization from less data,” Bastian said.
Thomson explains how the research will expose cadets to emerging technology while preparing them for the cyber-based challenges they will confront in the future.
“This project places USMA at the forefront of AI research with a focus on practical applications at the intersection of cyber and military operations,” Thomson said. “This work will directly support cadet understanding of the multi-domain battlefield of the near future.”
West Point is the ideal location for the project, with its unique combination of military operational/domain experiences for mission critical applications and subject-matter technical expertise in support of Army and Department of Defense (DOD) relevant projects and its position to help facilitate knowledge and technology transition opportunities to the Army, other military services, DOD and Intelligence Community.
An outcome of this research will provide Soldiers with a framework for mission-critical applications requiring trustworthy and deliberative AI-powered autonomous systems that will produce a rationale for decisions and recommendations with a COP that will both allow commanders to be more decisive, effective and agile in multi-domain operations.
The cadets’ understanding of how AI systems will integrate into future battlefield decision-making will be enhanced through capstone, independent study and integrated curriculum opportunities in an interdisciplinary team effort including the Departments of Systems Engineering, Mathematical Sciences, EECS and BS&L.
Furthermore, the research will allow the ACI ICSARL researchers and collaborators, as well as faculty from across the USMA departments, to augment their knowledge in such areas as mathematical optimization, machine/deep learning, uncertainty quantification, human computer interaction, computational cognitive modeling, and intelligent systems.
To learn more about the Army Cyber Institute, visit https://cyber.army.mil.
Social Sharing