Biomechanical model for adaptive gait planning in lower limb exoskeletons

In the last decade, powered lower limb exoskeletons (LLEs) have advanced as upright bipedal walking therapy methods and free personal mobility devices for spinal cord injured, paraplegic or lower limbs disability patients. Even though they have affirmed as rehabilitation devices in clinical settings, LLEs haven't been adopted on a vast scale as a mobility system in real life environments. This discrepancy can be traced back to the fact that the majority of LLEs available on the market are able to carry out the walking task as a repetition of identical steps, without any awareness of the environment they are moving through. For this reason, they are able to guarantee the user's safety only in controlled environments. In order to overcome this limitation, a new research field which aims to develop environment-adaptive gait planning methods for LLEs, has been established. In this thesis, I propose two different gait trajectory biomechanical models, which can be customized by environment-dependent parameters: the "Single step biomechanical model" and the "Continuous walking biomechanical model". The single step biomechanical model drives the LLE for a "step-by-step triggered" walking task. This model only describes the LLE's kinematics in single support, i.e. while one of the two legs is off the ground, swinging. The continuous walking biomechanical model drives the LLE in order to carry out the walking task autonomously, without waiting for the user's command in between each step. This model describes the LLE's kinematics during the whole gait cycle. Unlike the single step biomechanical model, it also drives the LLE in double support, i.e. while both legs are touching the ground. The proposed models could be useful especially if the parameters by which the trajectories can be customized, were calculated by Computer Vision techniques coupled with Artificial Intelligence algorithms, able to detect the environment, and compute the best parameters the trajectories should fit in, order to avoid collision with obstacles.