My research focuses on motion planning for human-robot collaboration in manipulation tasks.  When human's perform manipulation tasks in close proximity to other humans, reaching motions are performed quickly, rarely colliding with one another with little to no verbal communication.  However, when a human is replaced by a robot collaborator in this domain, reaching motions often lose this "human-like" fluidity; a human may have to slow down sacrificing efficiency in order to avoid collisions.  I am thusly interested in identifying what latent, necessarily human features are currently missing from robot manipulation strategies and applying such missing features in the context of a robot's motion planner in order to generate safer, more efficient human-robot collaboration.