Home research People General Info Seminars Resources Intranet
| Algorithms & Applcations Group | Home | Research | Publications | People | Resources | News
Home Page for Daniel Andrew Latypov | Parasol Laboratory

Picture Daniel Andrew Latypov
Algorithms & Applications Group

Parasol Laboratory url: http://parasol.tamu.edu/~dlatipov/
Department of Computer Science and Engineering email:
Texas A&M University office: 407 HRBB
College Station, TX 77843-3112 tel:
USA fax: (979) 458-0718

I'm an incoming A&M freshman this fall. I'm working on the multi-agent systems library and on an extension to the RVO algorithm of accounting for rotations proposed by Andy Giese. RVO is meant for locally avoiding collisions between agents on their way to wherever they want to go, which is generally decided by a global motion planner like a PRM. My graduate mentor is Jory Denny and obviously my faculty advisor is Dr. Amato.

Weekly Journal

Week 1

This week started off with me graduating from high school on Monday, June 3. I arrived in College Station on Thursday and spent the first day trying to get my CS and my parasol accouts back up and running from last year. Afterwards, I checked out some of the changes that had been made over the past year in the code and the computers. My somewhat trusty (due to buggy behavior) computer "Shanghai" was no longer in the lab along with Nanjing. Instead there was a "joefriday." On Friday, I took care of the REU and employment paperwork in Teague. Then I drove back home to bring the rest of my stuff to the Tradition the next week.

Week 2

This week started off on a bad note as I was unable to access the CSE vpn through Linux and I had to switch to Windows and configure it for SSTP while not being able to access the CSE Wiki due to a weird password issue. While waiting for that I read some papers about the method I was going to be working with, Reciprocal Velocity Obstacles (RVO), which solves for velocities robots can assume to avoid collisions and minimize the deviation from their course, or if it can't, at least tries to minimize how hard they'll hit. I read about a few of the variants, but none of them aimed to enhance the capabilities of the robot like Andy is doing by allowing them to rotate (which wouldn't normally matter as most of the work is done assuming circular robots). This will allow therm to hopefully squeeze through gaps in dense crowds which RVO performs poorly in. I also downloaded the RVO library created by the same group who published the paper on RVOs and Andy's graphical front-end to it. I slightly modified it to use function pointers for the update and preferred velocity planning routines to make it more general. I was also told that I would be helping rewrite the Group Behaviors library (the software for motion planning for multiple robots or agents) along with Karen, Andres, and Colton.

Week 3

This week I got the new and old versions of the Group Behaviors code, got both of them to compile and added in timing into the old one as a benchmark to test the efficiency of the newer version later on. I messed around with the new version a bit, implementing a Behavior that makes robots go in a circle and a Lorenz attractor behavior which I had to visualize with gnuplot due to the lack of any graphics in the new code. I also started working on implementing parts of the code I'd need such as elliptical robots that can rotate and some of the basic infrastructure. Such as handling the flow of data to InformationSources from Sensors and other places for access by anyone who needs it. I finally got the Cisco VPN to work from Linux and celebrated Shuvra's birthday by eating pizza and singing Happy Birthday.