Ezra Tal

Ezra Tal

Research Scientist at MIT, working on planning and control for fast and highly-maneuverable robots.

Ezra Tal: Research Scientist, LIDS, MIT

I am a Research Scientist at the Laboratory for Information and Decision Systems (LIDS), where I work with Prof. Sertac Karaman in the Autonomy and Embedded Robotics Accelerated (AERA) group. Previously, I was a Postdoctoral Associate at LIDS. I obtained my PhD degree from the Department of Aeronautics and Astronautics at MIT, and BSc and MSc degrees from Delft University of Technology.

I am an applied researcher, with a love for fundamental science. My research on innovative mobile autonomy is at the intersection of aerospace/mechanical engineering and robotics, and combines aspects of control theory and motion planning with machine learning, optimization, applied aerodynamics, and vehicle and electronics design. I am particularly interested in leveraging methods from robust control theory and differential games to safely bring fast and highly-maneuverable robots out of research labs and into the real world.

News

  • December 2022 New paper on Age-of-Information-based Wireless Networking for UAVs accepted to IEEE INFOCOM 2023: Video, PDF.

  • October 2022 We’ll present our recent work on Trajectory Generation and Tracking for an Agile Fixed-Wing VTOL Aircraft at IROS in the Workshop on Agile Robotics: Perception, Learning, Planning, and Control.

  • September 2022 Our paper on a novel learning-based trajectory generation algorithm was accepted for presentation at CoRL 2022: OpenReview.

  • August 2022 Our paper on INDI for agile maneuvering of a tailsitter aircraft is accepted for publication in AIAA Journal of Guidance, Control, and Dynamics and available Open Access.

  • July 2022 Uploaded two exciting new videos with aerobatic tailsitter maneuvers (including an airshow with three aircraft!) to YouTube and the project webpage.

  • July 2022 Two new pre-prints on tracking and generation of aerobatic fixed-wing trajectories are online: arXiv:2207.13218 and arXiv:2207.03524.

  • April 2022 Our paper on Cooperative Multi-Agent Trajectory Generation with Modular Bayesian Optimization is accepted at RSS 2022.

  • November 2021 I’ll continue working at MIT as a Postdoctoral Associate.

  • October 2021 I successfully defended and submitted my PhD thesis on Algorithms for Generation and Tracking of Fast and Agile Flight Trajectories!

  • August 2021 MIT News published an article about our research on multi-fidelity trajectory optimization: MIT News.

  • August 2021 Our AIAA Aviation 2021 paper on Global Trajectory-tracking Control for a Tailsitter Flying Wing in Agile Uncoordinated Flight has been awarded Best Student Paper in V/STOL Aircraft Systems!
    image

  • August 2021 Our new paper on trajectory-tracking flight control for a tailsitter flying wing will be presented at AIAA Aviation 2021 Video, PDF.

  • August 2021 We’ll present our paper on 3D-printed jet vanes for thrust vectoring rockets at AIAA Propulsion and Energy Forum 2021 PDF.

  • July 2021 Our work on Multi-Fidelity Bayesian Optimization of Quadrotor Maneuvers has been accepted for publication in The International Journal of Robotics Research (IJRR). It is now available on the IJRR website.

  • June 2021 Our work on Multi-Modal Motion Planning Using Composite Pose Graph Optimization will be presented at 2021 IEEE International Conference on Robotics and Automation (ICRA) Video, PDF.

  • May 2021 We released a major update for FlightGoggles, including a photogrammetric model of the MIT Stata Center, new rendering settings, and a Python API. Check out our new website https://flightgoggles.mit.edu, and the updated pre-print on arXiv.

  • November 2020 I will give a talk on Sensor-based Control for Fast and Agile Aerial Robotics at MIT.

  • July 2020 Our RSS 2020 paper Multi-Fidelity Black-Box Optimization for Time-Optimal Quadrotor Maneuvers has been selected as a Finalist for Best Student Paper.
    image

  • June 2020 Our RSS 2020 paper on black-box optimization for time-optimal quadrotor maneuvers is now available on arXiv, and the corresponding video can be viewed on YouTube.

  • May 2020 Our paper on accurate tracking of aggressive quadrotor trajectories is accepted for publication in IEEE Transactions on Control Systems Technology. Video on YouTube and pre-print on arXiv.