Robot for Observational Gait Assessment and Rehabilitation

Open Access
Homich, Andrew Joseph
Graduate Program:
Mechanical Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
May 17, 2014
Committee Members:
  • Jason Zachary Moore, Thesis Advisor
  • Stephen Jacob Piazza, Thesis Advisor
  • Everett Carl Hills, Thesis Advisor
  • Robot
  • Robotics
  • Parallel Bars
  • Physical Therapy
  • Observational Gait Assessment
  • Rehabilitation
  • Complementary Filter
  • Walking
This thesis explores the design of a robotic device for observational gait assessment and rehabilitation, a method to estimate a patient's orientation within the rehabilitation device, as well as an optimal state space controller to actuate the rehabilitation device. Current rehabilitation methods require the patient to propel the assistive device or offer limited walking distance. Additionally, current devices do not measure the patient's reliance on the assistive device, possibly prolonging the rehabilitation period or even preventing satisfactory function to be regained. A novel Robot for Observational Gait Assessment and Rehabilitation (\deviceName) was designed to address the shortfalls of current assistive devices. A complementary filter was developed to estimate the patient's orientation within the device using a magnetometer and gyroscope. Experiments of the complementary filter on a test platform show that the filter provides estimates within 5 degrees of the true value over a range of angular velocities. An optimal state space controller was implemented using a linear quadratic regulator. The controller preforms well both in simulation and on the actual device. The simulations is able to predict the linear behavior with an average error of 2.7 cm and the angular behavior with an average error of 8 degrees.