By Army Research Laboratory Public AffairsApril 18, 2019
ADELPHI, Md. -- Army researchers will lead a discussion and tutorial on innovative underground robotics research at the 2019 Institute of Electrical and Electronics Engineers International Conference on Robotics and Automation in May.
The tutorial, "Subterranean Autonomy: Open Challenges, Resilient Design, and Virtual Benchmarking," introduces the unique subterranean robotics problem domain and a concrete simulation-based framework for initiating and carrying out research into complex systems-oriented solutions, and will be supported by collaborators from the Open Source Robotics Foundation, the University of Southern California and the Defense Advanced Research Projects Agency.
Drs. Ethan Stump, Jon Fink and John Rogers from the U.S. Army Combat Capabilities Development Command's Army Research Laboratory, the Army's corporate research laboratory also known as ARL, serve as scientific advisers for the DARPA Subterranean Challenge and will be leading the tutorial and discussion.
There are many obstacles that both warfighters and emergency responders face while operating in underground environments, the researchers said.
Due to visual, technical, or environmental challenges, subterranean settings pose a great risk to personnel who are trying to carry out a mission (think time-sensitive scenarios such active combat operations or disaster response).
Stump believes that robotics offers a compelling answer for the challenges inherent to subterranean operations, but success will require systems-level thinking and research and development that can integrate across a broad range of technologies, including autonomy, perception, networking and mobility.
Success in using autonomous systems in subterranean environments will be important for future research.
"Subterranean environments are a fantastic domain for trying to understand the extreme limits of what can be done with autonomy," Stump said. "While it is possible to communicate information into and out of a deep subterranean environment, it is expensive and limited, and so it is important to understand exactly what parts of a complex mission you need human input for and what parts can be safely and reliably delegated to autonomy. This understanding has implications for many military and civilian applications, such as interplanetary science and underwater exploration."
During their discussion and tutorial, the researchers hope to provide further insight into the challenges that exist for subterranean robotics, while providing a space for learning and collaboration that can lead to even more innovation in this field.
"We will present the topic of subterranean autonomy as an invitation to the robotics community, focusing on key open challenges with an eye towards possible contributions and on a concrete starting point for building and demonstrating systems that can begin to address this complex problem space," Stump said.
The discussion of open challenges will be centered on the difficulties of perception and wireless communications in subterranean environments, and the team will present a holistic open-source simulation framework that can be used as a starting point for demonstrating the effectiveness of proposed approaches in complex environments.
Perhaps one of the most important aspects of the discussion and tutorial, the researchers will discuss what it means to run a virtual experiment and will introduce a virtual benchmark and simple leaderboard for comparing solutions.
Army researchers have worked on the technologies underlying subterranean autonomy for more than a decade, to include mapping GPS-denied environments as part of the Micro-Autonomous Systems Technologies, or MAST, Collaborative Technology Alliance, and, more recently, work on multi-robot mapping and decision-making as part of the Distributed Collaborative Intelligent Systems Technologies, or DCIST, Collaborative Research Alliance.
In the DCIST alliance, the emphasis of the research is on building maps that encode knowledge about the meaning of the environment (i.e., what are the different types of objects or rooms) and being able to learn from experience and human guidance.
According to the researchers, the planned session may appeal to graduate students in regards to thesis topics, practitioners interested in techniques for scalable simulation of complex environments, roboticists looking for insight into systems thinking for multi-agent applications and research groups looking for good entry points into a challenging research domain with an emerging community.
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.