ISR faculty leading, playing key roles in ARL cooperative agreement
ISR faculty and their students are playing key roles in the University of Maryland and University of Maryland Baltimore County’s (UMBC) new "ARTIAMAS" cooperative agreement with the U.S. Army Research Laboratory (ARL). The five-year agreement—worth up to $68 million—focuses on safe, effective, and resilient capabilities and technologies that work intelligently and cooperatively with each other and humans. The research will span engineering, robotics, computer science, operations research, modeling and simulation, and cybersecurity. Professor Derek Paley (AE/ISR), the director of the Maryland Robotics Center, is the lead researcher for the agreement.
The effort encompasses three research areas that will spur the development of technologies to reduce human workload and risk in complex environments. The agreement builds on a more than 25-year research partnership between UMD and ARL in AI, autonomy, and modeling and simulation.
Research Area 1: Collaborative Autonomy Research, Development, Test and Evaluation (RDT&E) Infrastructure
This research area addresses the lack of standard shared infrastructure, baseline scenarios, tools and common models for collaborative autonomous systems.
Professor Jeffrey Herrmann (ME/ISR) is the principal investigator for this research area. He is joined by Derek Paley; ISR alum, ARLIS Director for Systems Research and ISR Visiting Research Scientist Craig Lawrence; Executive Director of the Fraunhofer Center for Experimental Software Engineering and ISR-affiliated Professor Adam Porter (CS/UMIACS); ISR-affiliated Distinguished University Professor Dinesh Manocha (CS/UMIACS/ECE); Professor Ming Lin (CS/UMIACS); Assistant Professor Michael Otte (AE); and Matt Scassero, the director of the UMD Unmanned Aircraft System Test Site.
The researchers will focus on enabling collaborative RDT&E through shared infrastructure, digital twin development, robot navigation of complex terrain, simulation-based verification for autonomous systems, and long-distance collaborative autonomy.
“Currently, research teams have to build their own simulation testbeds to evaluate their algorithms and systems, which increases cost and delays the development of significant innovations,” says PI Jeffrey Herrmann. “We will collaborate with ARL researchers to plan and develop a modeling and simulation infrastructure that will include libraries of simulation models and tools for building and running them. This will reduce the time and effort required to evaluate how well new algorithms perform and test autonomous systems, which will reduce the resources needed for test & evaluation, validation & verification (TEVV), which is an important challenge in robotics and autonomy.”
Research Area 2: Harnessing the data revolution
This research area addresses networking and sensing, a battlefield Internet of Things (IoT) testbed, adaptive cybersecurity for battlefield IoT, OpenML cross-domain learning, integrated mission planning for urban environments, and predictive maintenance for Army assets.
Researchers will be concerned with enabling technologies such as Edge AI, intelligent surfaces, energy transfer and harvesting that are driving trends in reliable bandwidth, areal to volumetric spectral and energy efficiency, massively available small data and self-sustaining networks with applications to XR, robotics, and autonomous systems.
Research Area 3: Closer human-machine teaming
This research will explore how sophisticated artificial intelligence/machine learning sensor and computing systems can augment human performance for situational awareness, behavioral and physiological health assessment, battlefield uses, forensics, and metareasoning.
ISR-affiliated Distinguished University Professor Dinesh Manocha (CS/UMIACS/ECE) is the principal investigator for this area. He is joined by ISR Director and Professor Ankur Srivastava (ECE/ISR); Assistant Professor Huan Xu (AE/ISR); ISR-affiliated Professor Nikhil Chopra (ME); Professor Ramani Duraiswami (CS/UMIACS); Principal of Campbell Code Consulting and FPE Adjunct Lecturer Chris Campbell; Ming Lin; Craig Lawrence; Adam Porter; Jeffrey Herrmann; and faculty from UMBC.
Autonomous systems use metareasoning to control how they make decisions. This project will generate new knowledge about the performance of metareasoning approaches in human-autonomy teams that operate at different levels of human interaction, including supervising, advising, directing, and collaborating. Performance includes the quality of the solutions and plans that the reasoning algorithms generate, the time and computational resources that are required to reason, and the transparency of the approach. Ultimately, successful metareasoning approaches will improve the performance of human-autonomy teams while maintaining sufficient transparency needed for trustworthiness.
Research will include AI/ML on edge for situational awareness; individual and collective behavioral and physiological health assessment; explainable AI for the battlefield; perception-based interaction; forensics for human-machine teaming; metareasoning to improve team performance; and human-machine teaming and effective aggregation of information in complex systems.
Published May 26, 2021