Active Learning of Probabilistic Movement Primitives
New method of efficiently acquiring human demonstrations for robot manipulation tasks.
Adam Conkey
Roboticist - Python Developer - Rust Enthusiast
I am a Senior Software Engineer at Berkshire Grey. I’m passionate about robotics software development, particularly open source software for ROS and Linux. I’m excited about Rust - in my spare time, I am tinkering with implementing computational geometry algorithms from scratch in Rust.
I received my Ph.D. in Computing and Robotics from the University of Utah under the supervision of Dr. Tucker Hermans as a member of the Utah Learning Lab for Manipulation Autonomy. My dissertation work investigated skill planning under state and goal uncertainty for robot manipulations tasks.
If you’d like to chat about any of the projects I’ve worked on, or just want to connect, please find me on LinkedIn!
Projects
Learning Task Constraints from Demonstration for Hybrid Force/Position Control
Approach to learning force-constrained robot manipulation tasks from human demonstrations.
Teleoperation in Virtual Environments with Haptic Feedback
Rendering forces in simulation environments for robot manipulation tasks using haptic feedback.
Simulated Contact Sensors for ReFlex Robot Hand
Augmented robot model for the Gazebo simulator enabling simulation of distributed contact sensors.