One of the ways humans learn about their world is through touch. They can determine an object's temperature, shape, roughness, etc. Nick Dadds focused his Honors project on trying to give robots the same capabilities.
The three selected object properties to be determined are material type, surface roughness and surface temperature. The data is collected using a sensor suite that is positioned using a robotic manipulator. This project spanned over two semesters in which the first semester was used to research state of the art sensor technology, specify desired object properties, acquire necessary parts and begin development of the sensors. The second semester was used to continue development of the sensors while also designing a sensor suite.
The three selected object properties to be determined are material type, surface roughness and surface temperature. The data is collected using a sensor suite that is positioned using a robotic manipulator. This project spanned over two semesters in which the first semester was used to research state of the art sensor technology, specify desired object properties, acquire necessary parts and begin development of the sensors. The second semester was used to continue development of the sensors while also designing a sensor suite.
On great thing about the project was that it pulled together a lot of the technologies covered in his courses. Material type is based on the dielectric constant of the object and is determined through the use of a capacitance sensor (sounds like EE301). Surface roughness is determined through signal analysis obtained from a phonograph cartridge and stylus. Surface temperature is based on the heat energy emitted from the object which is determined through the use of an infrared temperature sensor (1/C do that in ES405). The target objects are located in the workspace using a Fixed Camera Tracking system developed in Advanced Robotics (ES452)... who says you never use all the things you learn in your courses?
Posts
