Optical Tweezers, Path Planning, and the GPU

2009-current

Five Micron particles, courtesy WikipediaThis project is in its infancy. More will be added after we publish!

With Dr. Amitabh Varshney and Dr. Satyandra Gupta, I am working on a project involving using the GPU to simulate different interactions with nanoparticles, specifically using optical tweezers. In this project, we are primarily interested in three areas:

  • Development of a GPU-based simulation infrastructure for simulating the trapping of nanoparticles and other such interactions typically found in nanoassembly operations;
  • Implementation and optimization of path planning algorithms, for use when the optical tweezer(s) wants to move the trapped nanoparticle(s);
  • Development of algorithms to automatically construct multi-object nanoparticle models.

We are basing much of our work on the recent Ph.D. thesis of Dr. Gupta's student, Ashis Banerjee (now at MIT). In his thesis, Dr. Banerjee describes a basic framework for simulation of the trapping process as well as a set of algorithms for multi-particle path planning. This path planning is done with high-level human interaction and low-level automation. We are "GPU-izing" this, as it were, with an emphasis on decreasing computational time into the realtime realm.

Reference Thesis: Ashis Banerjee, Real-Time Path Planning for Automating Optical Tweezers based Particle Transport Operations, August 2009.

This project is in its infancy. More will be added after we publish!