BU-led Research Team Wins Competitive $7.5 million MURI Grant to Create Neuro-Autonomous Robots

in Research
August 22nd, 2019

By Maureen Stanton, CISE

Dream Team of Engineers, Computer Scientists, and Neuroscientists from BU, MIT, and Australia to develop neuro-inspired capabilities for Land, Sea, and Air-based Autonomous Robots

A Boston University-led research team was selected to receive a $7.5 million Multidisciplinary University Research Initiative (MURI) grant from the U.S. Department of Defense (DoD).  With this prestigious grant, the researchers will develop a novel category of neuro-inspired autonomous robots for land, sea, and air that the investigators have termed “neuro-autonomous.”

The initiative will tackle the challenge of how to make robots truly autonomous as well as provide important insights into the process of learning and memory formation. The project will be led by Yannis Paschalidis, director of the Center for Information and Systems Engineering, and professor in the College of Engineering at Boston University.

The winning team of researchers is a Dream Team of experts from Boston University and the Massachusetts Institute of Technology. It includes experts in neuroscience, robotics, computer science, computer vision, artificial intelligence, mathematical systems theory, and a host of other related advanced technology domains. The project will also benefit from collaboration with renowned researchers from the University of Melbourne, Macquarie University, Queensland University of Technology, and the University of New South Wales.  (See researcher biographies.) 

Yannis Paschalidis, Director, BU Center for Information & Systems Engineering

Clearly, humans and animals are more effective at navigating than robots are,” said Professor Paschalidis.  “While there have been tremendous advances in intelligent systems, state-of-the-art autonomous systems are still limited in that they are programmed for specific objectives in well-structured environments.  Our research focus is to create a new class of neuro-autonomous robots, inspired by the fusion of multiple sensor modalities, spatial awareness, and spatial memory inherent in biological organisms.  These systems will have unprecedented capabilities for self-learning and on-the-fly adaptation to environmental novelty, enabling their ability to pursue complex goals in highly dynamic environments.”

Advancing a Science of Autonomy

Under this MURI grant entitled “Neuro-Autonomy: Neuroscience-Inspired Perception, Navigation, and Spatial Awareness for Autonomous Robots,” the research team will conduct experiments to gain insights into how humans and animals process visual and other stimuli to engage in goal-directed navigation.  With this greater understanding, the researchers will develop new algorithmic methods that can do the same for robots, enabling them to navigate autonomously. The implementation of the core algorithms will be agnostic to the specific robot used, enabling validation studies with ground, aerial, and underwater autonomous robots using experimental testbeds at BU, MIT, and Australian (AUS) institutions.

“This is fascinating science,” adds Professor Paschalidis. “We will develop the ability to make behavioral observations of animals and humans, correlate behavior with activity in the brain, and use the data to design control policies that will guide autonomous systems. This truly is the next frontier in advancing the field of robotics and autonomous vehicles.”

Cross-disciplinary Global Investigator Team

This MURI grant makes possible the multidisciplinary collaboration of an extraordinary team of researchers who bridge the neuroscience, engineering, and computer science worlds.  The team spans two continents and brings together some of the pre-eminent experts in neuroscience, with emphasis on localization, mapping, and navigation functions, with experts in robotics, computer vision, control systems, and algorithms.

The BU research team includes BU College of Engineering Professors John Baillieul, Yannis Paschalidis, and Roberto Tron, computer science Professor Margrit Betke, as well as neuroscience Professors Michael Hasselmo and Chantal Stern of the College of Arts and Sciences. The Massachusetts Institute of Technology is a subaward recipient on this grant. MIT Computer Science & Artificial Intelligence Laboratory (CSAIL) faculty John Leonard, Samuel C. Collins Professor of Mechanical and Ocean Engineering, and Nicholas Roy, Bisplinghoff Professor of Aeronautics and Astronautics, are the co-principal investigators.

Remarks from Principal Investigators

Michael Hasselmo, Director, BU Center for Systems Neuroscience

“While there are some very exciting computational theories, computational neuroscience has not yet fully accounted for the mechanisms and the function of all these interesting experimental phenomena,” says Professor Hasselmo, director of the Center for Systems Neuroscience at Boston University. “This project offers an exciting opportunity to collaborate with engineering in order to provide the mathematical and theoretical framework necessary. Further, this project offers the potential for some major theoretical breakthroughs for understanding cognition. While we will be focusing on navigation, elements of the algorithms we will develop could apply to a broad range of different types of intelligent behavior.”

John Leonard, MIT Computer Science and Artificial Intelligence Laboratory (CSAIL)

“As a roboticist, I am excited about combining insights into how scientists think the brain might work to make better robots and to exploit some of the recent advances in things like deep learning and object detection,” says Professor Leonard, MIT CSAIL. “Taking the best from biologically-inspired models and combining those with real robot experiments is very exciting. The potential impact of the research is awe-inspiring.  We understand now that memory formation is coupled to how an animal or human knows their position; there is a coupling there that perhaps could one day lead to better insights that ultimately might lead to better therapies for memories.”

Chantal Stern, Director, BU Cognitive Neuroimaging Center

“The MURI project is a fantastic opportunity for us to combine knowledge and expertise across disciplines,” says Professor Stern, Director, BU Cognitive Neuroimaging Center.  “As a neuroscientist, I find it fascinating to think about the speed and flexibility of human cognition, and I’ve learned a tremendous amount about what state-of-the-art robotics can and cannot do from my interactions with roboticists on the MURI team.  Spatial navigation is an ideal test bed for thinking about spatial awareness and problem solving across animals and humans, and I’m looking forward to working with the engineering and neuroscience faculty here at BU as well as with our robotics colleagues, John Leonard and Nick Roy, at MIT.”

John Baillieul, Distinguished Professor of Engineering, BU

Professor Baillieul, a Distinguished Professor of Engineering at Boston University states, “My career has been devoted to systems and control engineering.  The research that we’ll be doing under this MURI is focused on the most interesting control system out there–the brain and its coordination of the neurosensory and neuromuscular systems in the body.”

 

 

Margrit Betke, Co-director, BU Artificial Intelligence Research Initiative

“I am excited to lead the MURI research efforts around the video recordings of animals and their navigation behavior, as we monitor their brain activities,” said Professor Betke, co-director of the BU Artificial Intelligence Research Initiative. 

Professor Girish Nair at the University of Melbourne will lead the Australian-based researchers. “The Australian team is thrilled to participate in this project,” says Professor Nair.  “We look forward to a fruitful and exciting collaboration with our partners at Boston University and MIT.”

The BU-led MURI is one of 24 MURI awards announced by the DoD on April 3, 2019. The highly competitive MURI program supports basic research in science and engineering at U.S. institutions of higher education and is focused on multidisciplinary research efforts where more than one traditional discipline interacts to provide rapid advances in scientific areas of interest to the DoD.  

For details about this ONR MURI grant, see “Project Abstract approved for Public Release.”