Stroke patients lose their neural ability to activate leg muscles. As a consequence, they unable to perform daily life activities as a healthy person does. Powered assistive devices can provide external force to the affected leg in order to achieve better balance during walking compared to passives ones. Nowadays, various autonomous control strategies are explored to control powered assistive devices and exoskeletons. For example, gait event detection, locomotion mode recognition, prediction of walking speed, and step length prediction have become popular. These strategies utilize various wearable sensors to recognize either the intention of users or environmental conditions. EMG, goniometer, inertial sensors, force myography have been utilized for user’s intention recognition. Also, pressure sensors have been used to record the interaction between the user and the environment. For recognition of environmental conditions, depth cameras, RGB cameras, and LiDAR sensors are used. The efficiency of various signal processing, feature extraction, and machine learning approaches is also explored for this purpose. The methodology and outcomes of a few such state-of-the-art techniques will be discussed in this seminar.