VISION BASED FORCE SENSING FOR NANOROBOTIC MANIPULATION

摘要

Over the last decade, considerable interest has been generated in building and manipulating nanoscale structures. Applications of nanomanipulation include study of nanoparticles, molecules, DNA and viruses, and bottom-up nanoassembly. We propose a Nanomanipulation System using the Zyvex S100 nanomanipulator, which operates within a scanning electron microscope (SEM), as its primary component. The primary advantage of the S100 setup over standard scanning probe microscopy based nanomanipulators is the ability to see the object during manipulation. Relying on visual feedback alone to control the nanomanipulator is not preferable due to perceptual limitations of depth and contact within the SEM. To improve operator performance over visual feedback alone, an impedance-controlled bilateral teleoperation setup is envisioned. Lack of on-board force sensors on the S100 system is the primary hindrance in the realization of the proposed architecture. In this paper, we present a computer vision based force sensing scheme. The advantages of this sensing strategy include its low cost and lack of requirement of hardware modification(s). Force sensing is implemented using an atomic force microscopy (AFM) probe attached to the SI00 end-effector. Deformation of the cantilever probe is monitored using a Hough transform based algorithm. These deformations are mapped to corresponding end-effector forces following the Euler-Bernoulli beam mechanics model. The forces thus sensed can be used to provide force-feedback to the operator through a master manipulator.