Caelan Garrett

PhD Student
Learning and Intelligent Systems Group
MIT CSAIL
[first-name]@csail.mit.edu


Research

I am a PhD student in the Learning and Intelligent Systems Group in MIT CSAIL with Professors Tomás Lozano-Pérez and Leslie Pack Kaelbling. My research involves integrating Neoclassical Planning from Artificial Intelligence with Motion Planning from Robotics. I'm interested in understanding both theoretical properties these algorithms and problems as well as experimental results on problem instances typically encountered in applications. My goal is to develop fast algorithms for planning in high dimensional, continuous spaces that can be applied to autonomous systems to perform human-like tasks such as cleaning a room or cooking a meal.

Research Pages

My research interests are planning (discrete, motion, probabilistic, ...), reinforcement learning, and hyper-parametric function approximation (deep learning). My general interests are artificial intelligence, robotics, machine learning, complexity theory, and decision and game theory.

Education

  • Doctor of Philosophy in EECS - MIT (2015 - present)
    Thesis: TBD
  • Master of Engineering in EECS - MIT (2014 - 2015)
    Thesis: Heuristic Search for Manipulation Planning. [paper]
  • Bachelor of EECS and Bachelor of Mathematics - MIT (2011 - 2015)
    Thesis: FFRob: An efficient heuristic for task and motion planning. [paper]
  • Advanced High School Diploma - Thomas Jefferson High School for Science and Technology (2007 - 2011) Thesis: An Image Processing Algorithm for Enhancing Perceptual Visibility of Imagery

Preprints

  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. Sampling-Based Methods for Factored Task and Motion Planning, 2018.

Journal Publications

  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. FFRob: Leveraging Symbolic Planning for Efficient Task and Motion Planning, The International Journal of Robotics Research (IJRR), 2017. [arXiv] [doi]

Conference Publications

  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. Sample-Based Methods for Factored Task and Motion Planning, Robotics: Science and Systems (RSS), 2017. [paper]
  • Michael X. Grey, Caelan R. Garrett, C. Karen Liu, Aaron D. Ames, and Andrea L. Thomaz. Humanoid Manipulation Planning using Backward-Forward Search, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2016. [paper]
  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. Learning to Rank for Synthesizing Planning Heuristics (Ordinal Regression for Learning Planning Heuristics), International Joint Conference on Artificial Intelligence (IJCAI), 2016. [arXiv]
  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. Backward-Forward Search for Manipulation Planning, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2015. [paper] [arXiv]
  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. FFRob: An Efficient Heuristic for Task and Motion Planning, International Workshop on the Algorithmic Foundations of Robotics (WAFR), 2014. [paper]
  • Mark A. Livingston, Caelan R. Garrett, Zhuming Ai. Image Processing for Human Understanding in Low-visibility, ASNE Human Systems Integration Symposium, 2011. [paper]

Workshop Publications

  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. STRIPS Planning in Infinite Domains, ICAPS Workshop on Planning and Robotics (PlanRob), 2017. [arXiv]
  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. STRIPS Planning in Infinite Domains, RSS Workshop on Task and Motion Planning, 2017. [paper]
  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. STRIPS Planning in Infinite Domains, ICRA Workshop on AI Planning and Robotics, 2017. [paper]
  • Caelan R. Garrett, Michael X. Grey, C. Karen Liu, Aaron D. Ames, Andrea L. Thomaz. Humanoid Task and Motion Planning using Backward-Forward Search, RSS Workshop on Task and Motion Planning, 2016. [paper] [video]
  • Caelan R. Garrett, Tomás Lozano-Pérez, Leslie P. Kaelbling. Heuristic Search for Task and Motion Planning, ICAPS Workshop on Planning and Robotics (PlanRob), 2014. [paper]

Other "Publications"

  • Caelan R. Garrett, Clement Gehring, Gustavo Goretkin, Zelda Mariet, Zi Wang*. Motion Planning using Naturally Annoying Grammars (NAGs), SIGTBD, 2016. [paper]

Patents

  • Livingston, Mark A., and Caelan R. Garrett. System and Method of Image Processing Perceptual Visibility of Imagery. The United States of America, as Represented by the Secretary of the Navy (Washington, DC), assignee. Patent 8,705,857. 04 Apr. 2014. [link]

Presentations

  • STRIPS Planning in Infinite Domains, ICAPS Workshop on Planning and Robotics (PlanRob), Singapore, May 2017. [slides]
  • Learning to Rank for Synthesizing Planning Heuristics, International Joint Conference on Artificial Intelligence, New York, NY, July 2016. [slides] [poster]
  • Humanoid Task and Motion Planning Using Backward Forwards Search, RSS Workshop on Task and Motion Planning, Ann Arbor, Michigan, June 2016. [poster]
  • Backward-Forward Search for Manipulation Planning, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Hamburg, Germany, October 2015. [slides]
  • Heuristic Search for Fast Robotic Manipulation Planning, MIT Corporation Partners Robotics@MIT Event, Boston, MA, Dec. 2014. [slides]
  • Heuristic Search for Fast Robotic Manipulation Planning, MIT EECS Department Presentation, Boston, MA, Oct. 2014. [slides]
  • FFRob: An efficient heuristic for task and motion planning. International Workshop on the Algorithmic Foundations of Robotics (WAFR), Istanbul, Turkey, Aug. 2014. [slides] [poster]
  • Heuristic Search for Task and Motion Planning. ICAPS Workshop on Planning and Robotics, Portsmouth, NH, Jun. 2014. [slides] [poster]
  • Heuristic Search for Task and Motion Planning. MIT EECScon, Boston, MA, Apr. 2014. [poster]
  • Constraint Based Methods for Integrated Task and Motion Planning. MIT SuperUROP Conference. Boston, MA, Dec. 2014. [poster]
  • A Data Structure for Rapidly Testing Reachability for Robotic Motion Planning. MIT EECScon, Boston, MA, Apr. 2013. [poster]
  • An Image Processing System for Enhancing Perceptual Visibility of Imagery. Siemens Competition National Finals, Washington DC, Dec. 2010.

Awards


Teaching

  • MIT Introduction to EECS (6.01) - Teaching Assistant (Spring 2015)
  • MIT Introduction to Algorithms (6.006) - Tutor (Fall 2013 - Fall 2014)
  • TJHSST Introduction to Computer Science - Teaching Assistant (Summer 2009)

Theses Supervised

  • Yonadav Shavit, Learning Search Heuristics for POMDPs and Other Planning Problems, Bachelor of CS, 2016.

Internships

  • Optimus Ride (2017) - Computer Vision and Perception
  • Google (2014) - AdQuality
  • Twitter (2013) - Revenue Analytics
  • Lockheed Martin (2012) - Simulation, Training, and Support Internal Research Lab
  • Naval Research Laboratory (2010) - Information Technology Division

Projects

  • Kernel Regression for Planning Heuristics. 6.867 (Machine Learning) Final Project. Dec. 2014. [paper]
  • Local Cheeger Inequalities and Sparse Cuts. 18.404 (Seminar on Spectral Graph Theory) Final Project. May 2014. [paper] [slides]
  • Survey on Online Bipartite Matching and Its Variants. 6.854 (Advanced Algorithms) Final Project. Dec. 2013. [paper]

Referred Conference Reviewer

  • IEEE International Conference on Robotics and Automation (ICRA) - 2015, 2017
  • IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) - 2015, 2017
  • Association for the Advancement of Artificial Intelligence (AAAI) - 2016, 2017
  • International Joint Conference on Artificial Intelligence (IJCAI) - 2016, 2017
  • International Conference on Automated Planning and Scheduling (ICAPS) - 2016, 2017
  • Robotics Science and Systems (RSS) - 2017
  • Neural Information Processing Systems (NIPS) - 2017

Websites


Other

  • I sang in and was president of the MIT Chorallaries, MIT's first coed a cappella group.