We understand that general purpose and generally intelligent robots are going to be built in our lifetimes. Not content to sit on the sidelines, we work efficiently towards this goal by combining human and artificial intelligence into something greater than the sum of its parts- a Tutor Intelligence.
We enable anyone to teach robots complicated vision and manipulation skills. Our robots observe, copy, adapt, and improve on what we show them. They are the students, we are the Tutors.
The robot problems we tackle range from the classical to the unsolved. We solve problems in planning, computer vision, SLAM, optimization, machine learning, and more.
If this sounds like you, read on! Across the board we write software to make our robots more capable, faster and safer. We work mostly in Python and Linux.
What follows is a list of things we find interesting. Depending on your passions, your role might fit one or all of the following.
\n Robot VisionImproving 2D and 3D sensing capabilitiesBuilding visual representations of the worldGeometric perceptionPerception of seen and unseen objectsSupervised learningSelf supervised learning Robot MotionControl of robot armsPlanning motion in the presence of obstaclesOptimizing grasps and other basic manipulation primitivesMotion planningTools such as Drake, OMPL, KLAMPT, etcMulti roboti planningReinforcement learning Robot SoftwareProficiency programming in a Python-Linux environmentComfort with programming linters (Flake8, Mypy)Software support of real-time systemsVisualization of robot dataAdvanced command of Python (dunder methods, lambdas, exception handling, decorators) Robot HardwareDesign and control of end effectorsExperience integrating and programming robot armsMechanical design of mounts and stands.
\n