Wearable robotics have been proposed as one possible solution for increasing mo- bility and enhancing the stability and strength of users requiring prosthesis or musculoskeletal augmentation. However, most robotics literature neglects the in- ternal flexibility of any anatomic joints. It views the human limbs (and therefore their replacements) as rigid elements connected by revolute joints with a confined range of motion. Tensegrity-inspired designs replicate the musculoskeletal connections using compressive and tensile components within a self-stabilizing structure, preserving the hybrid (rigid/flexible).
Traditionally, monitoring biomechanics parameters requires a significant amount of sensors to track exercises such as gait. Both research and clinical studies relied on intricate motion capture studios to yield precise measurements of movement. The advantage of the IMU-based (inertial measurement units) system used is that it does not require a fixed array of infrared (IR) cameras; therefore, it is not only significantly cheaper, but it is also less restricted by its environment. Biomechanic simulation environments, such as OpenSim, are commonly used for their complex multi-body dynamics solvers [Delp et al., 2007], and this work extends the frame- work to simulate the behavior of flexible-rigid robotic systems within the same environment.
Specifically, in this thesis I discuss a method that captures motion indepen- dently of optical hardware with the specific goal of tracking joint-angle measure- ments with IMU sensors. I extended this approach to the control strategy and the sensing network for flexible-rigid robotics and human movements like walking for gait phase classification. Additionally, I interpreted the biomechanics solvers from OpenSim to generalize a set of required maximum joint-moments to transition between stages of gait based on physical characteristics (e.g., height, weight, etc.). To my knowledge, integrating flexible-rigid structures and human subjects using IMU sensors with complex biomechanics solvers has never been proposed before. This work could help applications beyond monitoring the phases of gait with direct applications to medical and assistive technology fields.