The purpose of the paper is to describe a wall climbing Hexapod having Shape Memory Alloy actuated suction grippers. Recent researches have shown that SMA provides an alternative for traditional actuators like motors, pneumatics or hydraulics. The property of shape memory alloy to forcefully regain their shape with temperature change is used to develop controllable suction grippers. The principle of joule heating is utilized for actuation of Nitinol springs. These controllable suction grippers can be integrated with standard hexapod gait for the wall climbing application. High energy usage and one-time evacuation of air are the major disadvantages. However, reduction in weight, bulkiness and noise of actuation are significant advantages. Usage of inherent material properties for actuation of suction grippers is the novelty introduced. This feature can also be applied in various legged and non-legged robotic applications.

The underwater robotic vehicle presented here is inspired by an Ostraciiform form of swimming with three oscillatory fins to propel itself and control its orientation. A mathematical model is made to simulate the motion of the vehicle based on the fin oscillations to aid in the real-time vehicle control. The model predicts the forces produced by the oscillating rigid fins in the water. A new motion control approach for an underwater robot is proposed and investigated. In this paper, the oscillating fin arrangement is under actuated and carries out four degrees of freedom (DOF) motion with three fins. The threes fins include: once caudal fin and two pectoral fins. The dynamic model developed for the robot has a highly nonlinear thrust vector map because of the thrust generation from these oscillation angles. Nonlinear control methods such as Backstepping control is applied to this robot model due to the continuous oscillatory motion of fins for achieving different DOF. Contrary to the feedback linearization technique which cancels potentially useful nonlinearities, this control scheme avoids the cancellation; resulting in a less complicated controller. Lyapunov stability theory was used to prove the system stability. RMS values of the error were used for tuning the controller gains constants.

This is a unique 11 DOF Quadruped robot developed at IIT Indore for my Bachelor Thesis work. It has 4 2 DOF legs and 3 DOF active spine uniquely configured to enable it to climb any tree from ground up. It will be able to climb and re-position itself on the trunk, transition from branch to trunk and vice-versa. The active 3 DOF spine enables the robot to maintain its stability by shifting its Centre of Gravity inside the support polygon while in motion. In this bachelor thesis, I detail my work on the development of the same.