2012 Poster Sessions : Robotics-Based Reconstruction and Synthesis of Human Motion

Student Name : Emel Demircan
Advisor : Oussama Khatib
Research Areas: Artificial Intelligence
Abstract:
Understanding human motion requires accurate modeling of the kinematics, dynamics, and control of human musculoskeletal system to provide the bases for the analysis, characterization, and reconstruction of their movements. These issues have much in common with the problems found in the studies of articulated body systems in robotics research. Here, we present the robotics-based algorithms and methodologies leveraged to provide novel musculoskeletal modeling, analysis and control methods.

In motion analysis, we present methodologies to characterize human postural behaviors and dynamic skills in a uni ed framework including the task, posture and additional constraints such as contact with the environment and physiological capacity. We develop human performance metrics and exploit the information given by the musculoskeletal models mapped into the motion of the human in a task oriented simulation framework.

In motion control and synthesis, we present algorithms for redundancy resolution and real-time control of human musculoskeletal system. We establish a marker space control structure for the reconstruction of human motion by direct tracking of marker trajectories. Our human motion reconstruction methodology provides the full motion dynamics in real-time and is applied to analyze several human dynamic skills. Finally, we present an approach for muscle redundancy resolution based on a hybrid electromyography and conventional computed muscle control method and apply it for dynamic simulations of human movement.

Robotics-based reconstruction and synthesis approaches presented in this work provide a basis for understanding natural human motion and the tools applicable for efficient robot control, human performance prediction or synthesis of novel motion patterns in the areas of robotics research, athletics, rehabilitation, physical therapy and computer animation.

Bio:
Emel Demircan is a Mechanical Engineering PhD student in Professor Oussama Khatib's robotics group within the Arti cal Intelligence Laboratory. Her research focuses on the application of dynamics and control theory for the simulation and analysis of biomechanical and robotic systems. She is broadly interested in human motion synthesis to investigate high performance skills for athletic training and performance improvement and rehabilitation of patients with impaired sensory-motor function. Her research interests include human motion dynamics, control and simulation; sports biomechanics; robotics for rehabilitation; and motion analysis for physiotherapy exercises. Emel received her MS in Mechanical Engineering from Stanford University and BS with Honors in Mechanical Engineering and in Industrial Engineering (double-major) from Robert College.