Our goal is to develop and analyze algorithms for robust multi-robot coordination where communication is used as a key enabler to achieving this goal. The key to effective coordination is information exchange. Our research looks at ways that use wireless signals to improve communication in multi-robot teams, enhance situational awareness, and secure core multi-agent algorithms. We are interested in applications of our research to positioning systems indoors, human-robot collaboration, and making robot-robot teams robust and secure against cybersecurity threats. Most generally, our work centers around trust and coordination in multi-robot systems. Prof. Gil has been granted an NSF CAREER award (2019) and has been selected as a 2020 Sloan Research Fellow. She also held a Visiting Assistant Professorship at Stanford University (2019) where she was working with the Stanford Robotics Lab. Please see links to our work that has been reviewed in MIT News, as well as several other news outlets including Wired and the Forbes!
Our REACT Lab
Robotics, Embedded Autonomy
& Communication Theory Lab
Our REACT Lab Heterogeneous Air/Ground
Testbed
...where we do fast algorithm prototyping and testing for multi-robot decision making and control
Our Students
Our students learn to control robots, track them using our new motion capture lab, and program them to work collaboratively using communication as a sensor for intelligent decision-making.
OUR RESEARCH
Communication As A Sensor
Multi-robot Sequential Decision Making
Resilience in Multi-robot Coordination
JOIN US!
We are always looking for talented and motivated PhD, M.S., and B.S. students to join our team. Email sgil@seas.harvard.edu with your CV and list of relevant current and past projects!
Check Us Out on YouTube!
Recent Publications
- Approximate Multiagent Reinforcement Learning for On-Demand Urban Mobility Problem on a Large Map (extended version)
- Learning Trust Over Directed Graphs in Multiagent Systems
- Learning Trust Over Directed Graphs in Multiagent Systems (extended version)
- Multi-Robot Adversarial Resilience using Control Barrier Functions
- Dynamic Crowd Vetting: Collaborative Detection of Malicious Robots in Dynamic Communication Networks
- Multiagent Reinforcement Learning for Autonomous Routing and Pickup Problem with Adaptation to Variable Demand