#latepost
Sharing our #RSS@RoboticsSciSys Conference #RSS2025 paper "Resolving Conflicting Constraints in Multi-Agent Reinforcement Learning with Layered Safety" joint work with @UCBerkeley & @MITAeroAstro
Project page๐ https://t.co/ktrOIVa1Ey
Paper ๐ฐ https://t.co/dWV4B3WYft
Our results, demonstrated through both simulations and physical drone experiments, show a substantial improvement in safety with minimal impact on operational efficiency.
This work demonstrates how the applied learning methods provide an interpretable framework for characterizing satellite behavior and decision-making, enabling greater transparency and understanding in autonomous operations. #Aiaa#MIT#satellite#gnn#graphs
๐ Finally, the study evaluates how information sharing affects collision rates and path planning. The results are compelling: sharing maneuver data significantly reduces collisions and enhances planning efficiency.
Hamsa Balakrishnan, the William E. Leonhard (1940) Professor in @MITAeroAstro, has been appointed associate dean of the MIT School of Engineering effective Aug. 1.
https://t.co/5sTTrZyaHC
Had such a fun time at the @cocomarl2024 workshop at RLC on Friday! Thank you to all that attended, from our incredible speakers to our poster presenters ๐
Prof. Hamsa Balakrishnan has been appointed associate dean of the MIT School of Engineering. Her new leadership role will focus on efforts to attract, retain, and support top talent across all academic levels. @DINaMo_MIT
https://t.co/aOUunH838T
Work done by @sidnayak98, Adelmo Morrison Orozco, Marina Ten Have, Vittal Thirumalai, Jackson Zhang, Darren Chen, Aditya Kapoor, Eric Robinson, Karthik Gopalakrishnan, @1_Anuj_M_1, @brian_ichter, @jmes_harrison, Hamsa Balakrishnan, @DINaMo_MIT@MITAeroAstro
We believe LLaMAR represents a step towards systematically analyzing the use of LLMs in multi-agent robotics, making autonomous systems more adaptable and capable in complex environments.
Check out the full paper for more details: https://t.co/hYGAIuq4G2 10/10