Sisir Karumanchi
Highlights:
May 21: Our journal papers on the llama and roman series of robots which were developed as part of the seminal Army Research Lab's RCTA program were accepted for publication in Field Robotics [Llama System, Complex Terrain Walking, Roman Human Scale Manipulation, Roman Large Object Grasping]
Oct 20: Our IROS 2020 paper on Llama was a finalist for the Best Application Paper Award [pdf]
Jun 20: Started as a Senior Applied Scientist with Amazon Robotics AI.
Oct 19: Built and tested our second-gen dynamic robot (Llama 1.1) with across the board improvements from the first-gen robot. [videos: youtube playlist].
Oct 18: Built and tested our first-gen dynamic robot (Llama 1.0) with custom actuators within eight months. Currently building the second-generation prototype . [videos: youtube playlist]
Jun 18: A journal paper on our work in autonomous truss assembly with RoboSimian was published in the Journal of Field Robotics (JFR), https://doi.org/10.1002/rob.21792 . Extensive experiments were conducted in early 2017 [videos of autonomous truss assembly][online version][preprint].
May 18: Accompanying videos to our ICRA paper (Proprioceptive Inference for Dual-Arm Grasping of Bulky Objects Using RoboSimian) are here [spotlight video][extended video].
Jun 15: Team JPL finished fifth in DARPA Robotics Challenge Trials [video] [DRC finals official webpage][DARPA Press Release][JPL Press Release]
Apr 14: Started as a Robotics Technologist with the robotics section at Jet Propulsion Lab, NASA.
Dec 13: Team MIT finished fourth in DARPA Robotics Challenge Trials
Jun 13: Team MIT finishes third in Virtual Robotics Challenge. [System Summary]
Sisir is currently a Senior Applied Scientist at Amazon Robotics AI. He spent 6 years as a Robotics Technologist at NASA’s Jet Propulsion Lab, Caltech. He was the software lead for the JPL team at the DARPA Robotics Challenge (DRC) finals. Before joining JPL, Sisir was the manipulation lead in the MIT Team for the early phases of the DRC program.
Sisir got his Bachelor’s degree in Mechatronic Engineering from the University of Sydney in 2005. In 2010, he got his Ph.D. with the Australian Centre for Field Robotics at the University of Sydney. His thesis was on non-parametric learning techniques to aid off-road navigation. His doctoral advisors were Dr. Steve Scheding and Dr. Tim Bailey. During 2011-2014, he was a postdoc at Massachusetts Institute of Technology. During this time he worked on semi-autonomy. First on off-road driving with Dr. Karl Iagnemma in the Robotic Mobility Group. And later on mobile manipulation with Prof. Seth Teller and Prof. Russ Tedrake.
In broad strokes, Sisir's research focus is on closing the loop from perception to action. His past experience ranges from wheeled robots, limbed robots to limb wheel hybrids. His specific research interests include:
Mobile Manipulation
Whole Body Coordination
Motion Planning
Guaranteed Motion Safety
Soft hazards (beyond obstacles, e.g. deformable soil, slopes etc.)
Bayesian non-parametric inference
Vehicle-terrain interaction
Optimization
