Soft Robotic Cellbot for Extra-terrestrial Locomotion

点击翻译按钮获取中文摘要

This project will address two research challenges: 1. Actuation, shape change and coalescing/decomposition: The soft robot must be able to change its shape in order to move efficiently over different terrains on remote planets. Multiple soft robotic 'cells' (small soft robotic units) must be able to come together and fuse, or be able to decompose into smaller independent robots. They will exploit these two capabilities to enhance their locomotion capabilities. 2. Intelligence and control: Cellbots must sense their environments and intelligently decide the best morphology to adopt according to the terrain they are on, or how they should coalesce in order to form new 3D shapes, to obtain most morphological advantage in manoeuvring within the terrain. Project description: In hazardous and unapproachable environments, such as outer space, there is a need for robots to autonomously and efficiently manoeuvre over a variety of uneven topographies. Autonomy is required because there is commonly a delay in communication between robot and base station. A soft robotic swarm with the ability to morphologically adapt based on the surroundings would provide an effective way to tackle this problem. The inspiration is taken from nature based on the movement of animals, to move on unpredictable geomorphologies. The aim is to create a modular soft robot swarm that is composed of simple spherical elements that can coalesce and coordinate their volumetric actuation, thereby generating locomotion suitable to the encountered environment. Having a large number of small soft robotic units ensures a robust, cost-effective, and high fault tolerant solution. Softness in the robot allows for easy change of shape of the robotic units and enhances its compliance by allowing it to deform according to roughness of the terrain. In the session, the robot can easily be deflated and densely packed to reduce transportation costs. So far, the project has developed a computational model of a cellbot in Simscape where the locomotion cycle and fundamental model are shown in figure 1. We have focused on optimizing robot model parameters and understanding the effect of friction coefficients on locomotion, to tune them to achieve an optimal locomotion strategy. Foot displacement values were different for different feet during starting few cycles when static friction was in transition range. Figure 2 shows the relative phase plot for two feet. The model was adjusted to calculate average foot displacement over all actuation cycles for different frictional values after subtracting this starting time. Results from this will be later applied to a more complex 3D robot model. Tests on the model created will include dropping it from height to test for crash landing on Mar's surface. This will help in impact testing and understand robot locomotion after deployment.