Friday, March 17, 2017

Academic Alliance workshop focuses on autonomy R&D

ISEE honors retired Sandia explosives engineer Dale Preece

Sandia researcher Bill Hart, right, moderates a panel discussion during the inaugural Academic Alliance workshop on Autonomy and Complex Systems. Alliance university participants, from left, included researchers from the University of Illinois at Urbana-Champaign, University of Texas, Austin, University of New Mexico, Purdue, and Georgia Tech.

Representatives from Sandia and the Labs’ Academic Alliance university partners convened March 3 for the inaugural Academic Alliance workshop on Autonomy and Complex Systems. The workshop focused on how academic partnerships can leverage the intersection of complexity and autonomy R&D for success.

The workshop was designed to align with Sandia’s Resiliency in Complex Systems Research Challenge, one of the Labs’ 11 active research challenges, which are intended to produce breakthroughs that impact the mission and contribute in their own right to advancing the frontiers of science and engineering.

Participating Alliance schools included Georgia Institute of Technology, Purdue University, University of Illinois at Urbana-Champaign, University of New Mexico, and the University of Texas at Austin.

Recent advances in artificial intelligence, spurred on by the availability of new hardware that is enabling deeper and cheaper machine learning, has raised interest in autonomous systems among academia, government, and industry.

Workshop organizers say these new technologies present both opportunities and challenges for national security. Solving the toughest of these national science and technology problems, they add, will require partnerships between national laboratories and universities.

The Academic Alliance initiative links faculty, students, and researchers at key universities with Sandia scientists and engineers to develop collaborative solutions to mission-critical challenges.

Workshop participants explored the use of complex systems approaches across autonomy-related topics such as human-machine teaming, artificial intelligence and machine learning, and distributed autonomy and cooperative systems. Participants discussed how the application of complex systems science can help researchers better understand the opportunities and risks in these research areas, and identified technical gaps that complex systems research can address.

In a wrap-up at the end of the workshop, participants found areas of common interests, but with universities and Sandia researchers focusing attention on different aspects of autonomy research. Specifically, university researchers were drawn to push the technical limits of the possible and “cool” for autonomous systems, while the Sandia researchers, with an eye on mission requirements, expressed concern about issues of resilience, reliability, transparency, security, and trust in the adoption of these systems for national security applications. 

A long term, sustained initiative in autonomy for national security, participants agreed, will require a new paradigm and capabilities for testing and evaluation during design that can provide assurance of autonomous systems under myriad conditions.