Optical Skin Project, Sarika Bajaj

The Optical Skin Project revolves around the concept of providing robots, specifically humanoid robots, a sensor equivalent to the human sense of touch. While other solutions to this problem have incorporated a variety of sensors from force responsive sensors to capacitive touch sensors, the Optical Skin Project relies on fish eye cameras tracking the deformation of specific markers present right outside the lens of the camera, which results in a system of force tracking that has much less noise and gives much more precision then the other commonly used systems of force tracking.

My specific effort on the project will involve the following steps: 1) recreating Prof. Yamaguchi's set-up of optical skin for Baxter's finger (to experimentally learn what difficulities such a process entails) 2) determining and executing a system which would allow for a full body optical skin to be made, which is relatively universal for a wide variety of body shapes (at least for every segment of Baxter's body) 3) using the new methodology and process developed from the full body optical skin, trying to determine if any changes can be implemented on the original finger scale version of this project.

The main skills needed for this project will involve material selection and experimentation (especially in plastic molding), a basic implementation of computer vision to determine whether each version of the optical skin is successful or not, and CAD ability to create the molds and casting spaces necessary for the complex shapes that have to be explored for this project.

Currently, there are no materials to be ordered for the project (and will need to be ordered later on in the semester after some preliminary experiments have been executed).