Research
I recently graduated from MIT where I worked in the
Learning and Intelligent Systems Group within the
Computer Science and Artificial Intelligence Lab (CSAIL) with Professors
Tomás Lozano-Pérez
and Leslie Pack Kaelbling.
I work at the intersection of robotics and artificial intelligence, and am interested in
developing robots that have the ability to reason about low-level physics while trying to
accomplish long-horizon manipulation tasks. In my work I explore ways of leveraging active
learning to aid in efficient data collection for learning accurate action models.
I am also very passionate about issues of diversity, equity, and
inclusion in the spaces I occupy. You can see some of my recommended
resources here.
Education
Massachusetts Institute of Technology
PhD in Electrical Engineering and Computer Science, May 2022
Minor: African American Studies
Advisors: Leslie Pack Kaelbling, Tomás Lozano-Pérez
[thesis] Optimistic Active Learning of Task and Action Models for Robotic Manipulation
Northestern University
MS in Computer Science, December 2015
Advisors: Robert Platt and Rahul Chipalkatty (Draper)
[thesis] Multi-Agent UAV Planning Using Belief Space Hierarchical Planning in the Now
Cornell University
BS in Mechanical Engineering, September 2013
Talks
-
Wevolver Innovators Update: Teaching robots to get curious
[article and video]
-
MIT Technology Review's EmTech Digital Conference
Active Learning of Abstract Plan Feasibility [video]
Publications
-
Caris Moses, Leslie Pack Kaelbling, Tomás Lozano-Pérez. Learning to Plan with Optimistic Action Models. ICRA Worksop - Scaling Robot Learning, 2022. [paper]
-
Michael Noseworthy*, Caris Moses*, Isaiah Brand*, Sebastian Castro, Leslie Pack Kaelbling, Tomás Lozano-Pérez, Nicholas Roy. Active Learning of Abstract Plan Feasibility. Robotics: Science and Systems (RSS), 2021. [paper] [video]
-
Michael Noseworthy*, Caris Moses*, Isaiah Brand*, Sebastian Castro, Leslie Pack Kaelbling, Tomás Lozano-Pérez, Nicholas Roy. Curiosity-Driven Learning of Abstract Plan Feasibility. ICRA Worksop - Towards Curious Robots: Modern Approaches for Intrinsically-Motivated Intelligent Behavior, 2021. [paper]
-
Caris Moses, Jane Shi. Integrating State Estimation and Perception for Picking. IROS Worksop on Why Robots Fail to Grasp, 2020. [paper]
-
Caris Moses*, Michael Noseworthy*, Leslie Pack Kaelbling, Tomás Lozano-Pérez, Nicholas Roy. Visual Prediction of Priors for Articulated Object Interaction. IEEE International Conference on Robotics and Automation (ICRA), 2020. [website]
-
Caris Moses, Rahul Chipalkatty, Rob Platt. Belief Space Hierarchical Planning in the Now for Unmanned Aerial Vehicles. AIAA Infotech@Aerospace, 2016. [paper]
-
Rashi Tiwari*, Michael A Meller*, Karl B Wajcs, Caris Moses, Ismael Reveles, Ephrahim Garcia. Hydraulic artificial muscles. Journal of Intelligent Material Systems and Structures, 23(3), 301-312, 2012. [paper]
Internships
-
Amazon Robotics (2020)
Integrated state estimation with segmentation to predict errors in the
segmentation. See workshop paper for more details.
-
Mujin (2017)
Worked on verification methods for a bin-picking robotic system in order
to prevent failures before they were encountered on the deployed robots.
Selected Projects
-
Peg Insertion with Policy Search
Dynamic Programming and Stochastic Control, 6.231, Spring 2017, MIT
[report]
[slides]
-
Template Matching in Image Colorization
Advances in Computer Vision, 6.869, Fall 2017, MIT
[paper]
-
Localization and Reference Tracking in Mobile Robots
UCSD STARS, Summer 2012, UC San Diego
[real robot video]
[rviz view video]
[slides]
-
The 99% Robot - My very first robot!
Mechatronics, MAE 3780, Spring 2012, Cornell University
[video1]
[video2]
[report]
[slides]
Teaching
- Robotics: Science and Systems (6.141/16.405) - Teaching Assistant (Spring 2020)
Selected Awards
|
