video: In the past 10 years, computing capabilities increased dramatically, which is particularly important for platforms where there is limited space and power, for example. Plus, new machine learning techniques that can take advantage of advances the ability collate and annotate data, and be able to learn advanced models have significantly changed the field, and dramatically impacted the field of research in autonomous robots. Now, researchers can really think of having a complete understanding of environments in which robots operate and labeling the objects in the environments, labeling regions and not only that, also understanding people their activities, which is much further than what could be done a few years ago. Here how research collaborator Dr. Martial Hebert, Dean of Computer Science at Carnegie Mellon University explains it. view more
Credit: T'Jae Ellis and Jim Nelson for CCDC ARL
ABERDEEN PROVING GROUND, Md. (Sept. 10, 2019) -- Army officials will demonstrate intelligence behaviors that are foundational to the autonomy for the next-generation combat vehicles at an Oct. 17 integration showcase at Carnegie Mellon University's National Robotics Engineering Center in Pittsburgh.
The work stems from research within the Army's Robotics Collaborative Technology Alliance, a 10-year program for state-of-the-art basic and applied research related to ground robotics technologies, with an overarching goal of developing autonomy in support of manned-unmanned teaming and serves as the foundational research in support of future combat ground vehicles.
Led by the U.S. Army Combat Capabilities Development Command's Army Research Laboratory, the partnership that includes university and industry researchers pushed the frontiers of scientific study in four critical areas of ground combat robotics that affect the way U.S. warfighters see, think, move and perform as a team.
Dr. Stuart Young, the laboratory's RCTA collaborative alliance manager, said the work fundamentally disrupted thinking in government and industry around autonomy and showed the Army how to think of the realm of the possible.
"We created new knowledge and understanding in ground robotics autonomy was created that benefits government, academia, and industry, and we forged the ARL Autonomy Enterprise to operationalize artificial intelligence for maneuver," Young said.
The program is not without its challenges, he said.
"Robots are slow so we have to address that they have to go faster," he said. "Soldiers are not going to accept robots that break easily so we have to create more resilience not just physically but also in software and in intelligence."
Young also serves as program manager for the lab's Artificial Intelligence for Maneuver and Mobility Essential Research Program.
The RCTA was awarded in 2010 and is expected to conclude by early fiscal 2020. In the final years of the program, the program has brought together the research focusing on three capability thrusts: Operational Tempo Maneuvers in Unstructured Environments; Human-Robot Execution of Complex Missions; and Mobile Manipulation. Demonstrations of technology that will form the foundation of autonomy for the next-generation combat vehicles include high-speed perception and mobility in rough terrain, situational awareness in unstructured environments, collaborative human-robot mission planning and execution, multi-modal human-robot dialogue and dexterous manipulation in cluttered environment; and finally, how this research will transition to future vehicles.
Early on, the group identified the need for deep understanding of the environment, said Professor Martial Hebert, director, National Robotics Engineering Center at Carnegie Mellon University, who's been part of the alliance since the beginning.
"That is the critical need to be able to make intelligence decision for an autonomous system," Hebert said. "The problem at the time is that this understanding was still very primitive. We could recognize some objects, we could do some level of understanding of the environment but at the cost of large amounts of computing which was not possibly for this kind of application, and in a way that was fairly specific for different types of understanding so we were still very, very far from the level of understanding that is needed to be able to make the decisions that we need."
The improvement in sensor technology over the past 10 years, he said, has made a considerable difference, resulting in "much better cameras, much better range sensors. Computing capabilities increased dramatically, which is particularly important for this kind of platform where we are very limited in terms of space, power, etc."
Availability of data -- very large amounts of data -- is critical, he said.
"Not only the availability of data but the availability of technology to collate, annotate that data," he said. "Finally all of that comes together with new machine learning techniques that can take advantage of all of those advances to be able now to learn advanced models."
RCTA partners addressed research and development required to enable the deployment of future military unmanned ground vehicle systems ranging in size from man-portables to ground combat vehicles. The group identified four key technology areas expected to be critical to the development of future autonomous systems, namely: Perception, Intelligence, Human-Robot Interaction and Dexterous Manipulation and Unique Mobility. It further emphasized the overlap and interplay among these technologies and the need to better understand their interactions through relevant integration and assessment activities. Some work in the effort will migrate to other Army research programs focused on future combat solutions.
###
The CCDC Army Research Laboratory (ARL) is an element of the U.S. Army Combat Capabilities Development Command. As the Army's corporate research laboratory, ARL discovers, innovates and transitions science and technology to ensure dominant strategic land power. Through collaboration across the command's core technical competencies, CCDC leads in the discovery, development and delivery of the technology-based capabilities required to make Soldiers more lethal to win our Nation's wars and come home safely. CCDC is a major subordinate command of the U.S. Army Futures Command.