The aim of this project is to develop a leg exoskeleton robot.  The exoskeleton can be worn on the legs and would provide power and load sharing capacity during walk/exercise. First, the human leg motion during walk will be studied using a motion capture system, based on which the optimal leg exoskeleton will be designed. The leg joints will be designed using 4-bar mechanisms, motors and sensors that will provide support for load carrying and also emulate human motion. An experimental study will also be carried out to understand the relation between the surface EMG signals from the leg muscles and the leg force control in humans. Based on this study a learning based controller will be designed to control the exoskeleton using the wearers EMG signals. The developed exoskeleton will have applications in active prosthetics, support for the elderly and rehabilitation of persons with neuro-muscular disorder of lower limbs etc.