Heterogeneous Collective (Swarm and Colony Robot) Testbed Infrastructure
Description: The HMTLab is building a swarm infrastructure testbed that allows for developing and evaluating capabilities for heterogeneous biologically inspired swarms of 50 – 150 vehicles. The infrastructure incorporates simulation-based capabilities, as well as two new, indoor hardware-based capabilities. The mobile swarm platform can accommodate up to two ground vehicles and four quadrotors. The full 12 ft x 14 ft swarm platform will support spatial swarm and colony-based collectives currently composed of 100 ground vehicles and 50 quadrotors.
People: Jennifer Leaf, Grace Diehl, Victor Agostinelli, Thomas Snyder, Tyler Prall, Matthew Lohr, Chloe Fleming, Ovunc Tuzel, Jason Cody
Description: This effort addresses the need to define and measure the resilience of Naval systems in coping with noisy and incomplete data sources, integrating information from multiple sensing modalities, and coherently propagating and outputting measures of uncertainty, by adapting resilience definitions and approaches from the ecological domain. This project is developing and formalizing an underlying theory of resilience, defining and validating associated resilience metrics, and deriving resilient system design principles that will enable comparison of the performance of teams in various configurations and under different classes of perturbations from the environment or the data sources. A formalized definition of resilience will account for augmenting agents’ computational potential, physical embodiment, information networks and the system’s capacity to provide services that will accomplish the overall mission, even when simple or complex disturbances are encountered. Metrics will be developed to assess resilience to perturbations (a) in the environment, (b) in robot embodiments, and (c) in communication and information networks. These metrics will be validated through a series of experiments using progressively more complex team configurations and classes of perturbations.
Sponsor: ONR Advancing Artificial Intelligence for the Naval Domain
Collaborators: Michael Goodrich, BYU (Co-PI) and Matthias Scheutz, Tufts (Co-PI)
Personnel: Jennifer Leaf
Description: This collaborative effort is developing core capabilities for the OFFensive Swarm Enabled Tactics (OFFSET) program that will support swarm autonomy and human-swarm interaction. Specifically, this effort will deploy up to 250 heterogeneous ground and aerial unmanned vehicles in a ten-block urban environment. This effort is intended to revolutionize the science, devices and systems to support future swarms. Our efforts focus on developing a swarm interaction grammar, conducting evaluations of the developed human-swarm interaction and supporting field exercises.
Sponsor: DARPA OFFSET
Personnel: Grace Diehl, Garrett Fleetwood, Larissa Letaw
Resilient Emergent Properties for Autonomous Agent InteRactions (REPAIR)
Description: REPAIR seeks to provide biologically inspired metaheuristic control algorithms for autonomous swarm agents, along with a framework for ongoing development and evaluation. Metaheuristics are a class of algorithms that guide the selection or operation of lower-level policies. The primary objective is to capture, in theoretical and algorithmic form, the unique protective and defensive capabilities exhibited in nature, and to demonstrate those capabilities in simulated scenarios of distributed swarms. This effort studies honeybee behaviors in order to inform the metaheuristic design.
Sponsor: DARPA SBIR
Personnel: Sumedh Mannar, Insuk (Patrick) Joh, Liqiang He, Ovunc Tuzel, Chloe Fleming
Managed Bio-Inspired Collectives
Description: This project is developing the underlying theory and identifying effective systems of humans and artificial agents when these systems include both hierarchical organizations and distributed collectives. The term agents in this context can include unmanned aerial, ground, surface, and underwater vehicles, and other autonomous systems including intelligent user interfaces and decision support systems. Since the described system depends on the autonomy the distributed entities possess, the research approach includes introducing and operationalizing a set of novel definition of autonomy and enabling human interaction with hub-based colonies.
Sponsor: ONR Science of Autonomy
Collaborators: Michael Goodrich, BYU (PI)
Personnel: Karina Roundtree, Jennifer Leaf, Teresita Guzman Nader
Biologically Inspired Algorithms for Heterogeneous Collectives
Description: Biological spatial swarms (i.e., schools of fish, flocks of birds) and colonies (i.e, honeybees or ants) provide resilient representative systems. Artificial collectives are hypothesized to be as adaptive and resilient as biological systems, but to do so, they need to incorporate algorithms that support such biological systems. The HMTLAb has developed and validated algorithms representative of spatial swarm communication, namely the visual (perceptual), metric and topological models; sequential consensus decision making; swarm leadership, and distributed multiple target search.
Sponsors: ONR Science of Autonomy, The United States Army
Personnel: Jennifer Leaf, Chloe Fleming, Jason Cody, Ovunc Tuzel, Aeijan Bajracharya, Ross Hoffman, Thomas Snyder, Musad Haque, Abigail Rafter, Douglas Kirkpatrick, Yifan Guo, Christopher Ren, Daniel McClanahan
Hybrid Mission Planning with Coalition Formation
Description: This project builds on the HMTLab’s prior work that has developed i-CiFAR, a framework of existing and developed coalition formation algorithms to support allocating teams of humans and unmanned systems. The current efforts combine this coalition formation capability with mission planning to support teams deployed in dynamic, uncertain, high stress environments (i.e., first response to natural or manmade disasters). Coalition formation and mission planning are both difficult to solve in real-time, as is required in such domains. Most existing systems cannot account for all the encountered factors; thus, the goal is to combine coalition formation and mission planning to a) make the problem more tractable and b) provide solutions that are more relevant and valuable.
Support: ONR Science of Autonomy
Personnel: Gilberto Marcon dos Santos, Miguel Ruiz, Zayed Shureih, Anton Dukeman, Sayan Sen, Travis Service, and Brian Okorn
Adaptive Human-Machine Teaming
Description: Human teaming with machines requires systems that can understand and adapt to their human teammates. This project incorporates objective metrics to determine the performance patterns of the human teammates (i.e., workload) in real-time, predict near term human performance and use this information to adapt the machines’ interactions with the human teammates or (re-)allocate tasks among the team members in order to improve the individual teammate’s or overall team’s performance when completing tasks. The work relies on using objective metrics (i.e., heart rate variability, speech rate, noise level) to assess performance. The current state of the art algorithm requires known tasks, but future efforts will focus on incorporating real-time task identification.
Sponsors: DARPA/NASA, AFOSR
Personnel: Jamison Heard, Julian Fortune, Rachel Heald, Caroline Harriott, Glenna Buford, Leah Guest, Corey Wollaeger, Matthew Manning
Real-Time Human Task Recognition – Hands Free Documentation
Description: This project is developing infrastructure that uses wearable devices (i.e., Myo) and environmental sensors (i.e., cameras) to automatically detect tasks performed by a human in real time. The current application is the ability to track and create a medical record of clinical procedures given to a patient while in transport to the hospital.
Sponsors: Department of Defense
Collaborators: Vanderbilt University (PI)
Personnel: Jamison Heard, David Greiner, Anjali Vasisht
Taskable and Adaptable Autonomy for Heterogeneous Marine Vehicles
Description: The project is developing autonomy capabilities that facilitate on-vehicle intelligence, leading to optimized neglect tolerance and longer duration deployments of unmanned underwater and surface vehicles. Human operators will provide high-level specifications for information gathering missions, such as searching for an area of interest and tracking targets of interest (e.g., biological hot spots). The vehicles use online representations, learned online, to expand the overall search area and coordinate activities based on automatic cuing, such as redirecting vehicles to potential target areas. Due to current on-board processing and between-vehicle communication limitations, data processing, learning, and planning are being handled on a manned vessel initially. The ultimate objective is to develop intelligent autonomy capabilities that extend the functionality of unmanned marine systems and reduce the need for human intervention and control.
Sponsor: NSF Smart and Autonomous Systems
Collaborators: Geoffrey Hollinger, (PI), Seth McCammon, and Matthew Frantz, Robotics, and John Barth, Kipp Shearman, and Jonathan Nash, Oceanography, Oregon State University
Personnel: Gilberto Marcon dos Santos, Gretchen Rice
Unmanned Aerial Systems
Description: The HMTLab has a number of efforts related to unmanned aerial systems technology. Prior efforts have involved some of the first work in using unmanned aerial systems to develop three dimensional maps of environments, and designing a new platform for fighting forest fires. Future research efforts are focused on developing autonomy as well as multiple vehicle and swarms of unmanned aerial systems. Dr. Adams is the Oregon State University primary investigator to the FAA ASSURE Center of Excellence. This center focuses on fundamental research for the integration of unmanned aerial systems into the national airspace.
Personnel: Karen Harper, Ben Lester, Justin Wayne, Eli Hooten, Blake Wulfe, Matthew Black, Andrew Hoofnagle, Ammar Abdelwahed, and Mark Bailey