Collective Perception and Decision Making in a Robot Swarm

Advisor: Prof. Radhika Nagpal, Harvard University

This research aims to improve the ability of a large group of robots to perceive and understand their environment by employing agents with different perceptual skills.

Some Results

Julia Ebert, Melvin Gauci and Radhika Nagpal. 2018. Multi-feature collective decision making robot swarms. In Proceedings of the 17th International Conference on Autonomous Agents and MultiAgent Systems, 1711–1719.

Collaborative Autonomy for Space Situational Awareness

Advisor: Dr. Michael Schneider, Lawrence Livermore National Laboratory

Tracking satellites is an important component of space situational awareness (SSA). However, current ground-based tracking approaches rely on centralized detection and require hours to accurately estimate an orbit. A constellation of low-cost, autonomous cube satellites could provide a fast and robustly decentralized architecture for SSA. We propose distributed particles filters as a method to iteratively refine orbit estimates with low communication bandwidth. We demonstrate the feasibility of this approach by implementing our algorithm in simulation. This simulator can also be used to evaluate the parameter space for future satellite constellation design, as well as test the system's robustness to failures.

Some Results

Julia Ebert, Joshua Meyers, William Dawson and Michael Schneider. 2018. Collaborative Autonomy for Space Situational Awareness. Poster at Lawrence Livermore National Laboratory Summer Student Poster Symposium (8 August 2018). Livermore, CA.

Cooperative Exoskeleton Control for Human Balance Recovery

Advisors: Prof. Etienne Burdet and Dr. Ildar Farkhatdinov, Imperial College London

Maintaining balance in the face of perpurbations is essential to walking and standing. For my masters thesis, I developed controls for LOPES (LOwer-extremity Powered ExoSkeleton, University of Twente) to assist humans with balance recovery after perturbations, using a combination of feed-forward and feedback control (such as hip torques and a PD controller). We found that even simple, single-joint torques are sufficient to reduce the time to a recovery movement and the energy used by subjects in recovery.

Some Results

Julia Ebert, Ildar Farkhatdinov, Geiss van Oort, Edwin van Asseldonk and Etienne Burdet. 2016. Preliminary Study on Assisting Balance Recovery with Lower Limb Exoskeleton. Poster at EuroHaptics 2016 (4–7 July 2016). London, UK.

Stability and Predictability in Human Control of Complex Objects

Advisor: Prof. Dagmar Sternad, Northeastern University

We examined human control of physical interaction with objects that exhibit complex dynamics, hypothesizing that humans exploited stability properties of the human-object interaction. Using a simplified 2D model for carrying a cup of coffee, we developed a virtual implementation to identify human control strategies. The specific task was to transport the cart-pendulum model of a cup of coffee to a target, as fast as possible, while accommodating assistive and resistive perturbations. To assess trajectory stability, we applied contraction analysis. We showed that when the perturbation was assistive, humans absorbed the perturbation by controlling cart trajectories into a contraction region prior to the perturbation. When the perturbation was resistive, subjects passed through a contraction region following the perturbation. Entering a contraction region stabilizes performance and makes the dynamics more predictable. This human control strategy could inspire more robust control strategies for physical interaction in robots.

Some Results

Salah Bazzi, Julia Ebert, Neville Hogan and Dagmar Sternad. 2018. Stability and Predictability in Dynamically Complex Physical Interactions. In 2018 IEEE International Conference on Robotics and Automation (ICRA), 5540–5545.

Salah Bazzi, Julia Ebert, Neville Hogan and Dagmar Sternad. 2018. Stability and predictability in human control of complex objects. Chaos, 28, 10. DOI: https://doi.org/10.1063/1.5042090

Bimanual Learning and Retention

Advisor: Prof. Dagmar Sternad, Northeastern University

Participants were asked to perform a task in which one arm is moved rhythmically while the other makes fast, discrete movements when cued. Over 20 sessions of practice, participants learned the task asymmetrically: while they learned to make much faster discrete movements, they failed to attenuate the perturbation these discrete movements caused in the rhythmic arm. After 6 months, subjects retained the skills they learned, including the asymmetry.

I proposed this research in my application the Barry Goldwater Scholarship and conducted the work for my undergraduate honors thesis. It was also funded by two Provost Undergraduate Research Awards.

Some Results

Julia Ebert, Se-Woong Park and Dagmar Sternad. 2015. Asymmetric Learning in an Asymmetric Bimanual Task. Poster at Society for the Neural Control of Movement 25th Annual Meeting (20–24 April 2015). Charleston, SC.