2014 Poster Sessions : Demo: Simulation of Dynamics, Control, and Haptic User Interface for a Humanoid Underwater Robot

Student Name : Gerald Brantner
Advisor : Oussama Khatib
Research Areas: Artificial Intelligence
Exploring and monitoring these oceanic resources is expensive and challenging because it requires human divers, who can only explore these environments during short periods of time and within limited depths. While underwater vehicles have proven to be useful for safely exploring oceans at greater depths, they lack human dexterity, which is necessary for performing fine manipulation tasks like collecting reef samples. To help scientists safely explore currently inaccessible underwater environments, we are designing a semi-autonomous underwater robotic explorer. The robot combines two force-controlled lightweight arms that are remotely operated using an intuitive two-handed haptic interface providing force-feedback guidance to the operator. We simulate the dynamic characteristics, haptic interaction, and control architectures that can decouple fine manipulation tasks from global disturbances due to unpredictable currents acting on the robot explorer.

Gerald Brantner is a PhD candidate in the Stanford AI-Lab researching humanoid robotic systems. His research interests are control, interface design, biomechanics and haptic interaction. Current and previous projects include designing control architectures and interfaces for a humanoid underwater robot, decoding human motion primitives based on functional neuroimaging, co-developing an fMRI-safe haptic device, and the implementation of AI algorithms for tactile and visual perception. Gerald's collection of academic degrees include MS'12 in Management Science and Engineering from Stanford and BS/MS'10 in Industrial-Mechanical Engineering from the Vienna University of Technology and he was a one-year exchange student at Columbia University.