Howdy! My name is Cameron Smith and I am a 2012 CSCE-REU/USRG Student. This summer I worked in the Parasol Lab in the Department of Computer Science and Engineering, here at Texas A&M. My research project was multi-robot caravanning, a proposed framework for distributed group coordination of movement amongst heterogeneous agents in an environment. Caravanning is useful because it scales well in large environments as well as those involving a significant number of agents. To learn more about multi-robot caravanning an its applications, see the deliverables, below!
Colleges/Universities: Tarrant County College (A.A. Degree, Spring 2012); Dallas Baptist University
Major: Computer Science
Expected Date of Graduation: Spring 2015
Faculty Mentor: Dr. Nancy Amato, Co-Director of the Parasol Labs in the Department of Computer Science and Engineering at Texas A&M University. Her research interests in include: Computational Biology, Motion Planning, Animation and Robotics, Parallel and Distributed Computing.
Direct Supervisors: Andrew Giese (MS Student) and Jory Denny (PhD Student)
Teammate: Rachel Glockenmeier (Undergraduate)
Deliverables: Research Plan, Progress Report, Final Research Paper, Poster
I spent most of this week getting acclimated with my research project for the summer -- multi-robot caravanning. I also completed the first in a series of 13 literature review assignments, reading a paper on motion planning. After reading each paper, we have a group review session where we share any thoughts or critiques we had about the text. Additionally, since I didn't have much experience in programming, Jory and Andy supplied me with a number of helpful resources to get me familiar with the language of C++.
I completed the first in a series of four "crash course" assignments, designed to help CSCE-REU students, become familiar with the Parasol Motion Planning Library (PMPL) code we would be working with for the duration of the summer. I also developed my research plan and completed two reading assignments on Probabilistic Roadmaps (PRMs) as well as simulated flocking methods in robotics.
I completed the next two additional crash course assignments (2 and 3). Assignment 3 dealt with checking and updating the standards and procedures of PMPL code. Since I was still getting used to using C++, this assignment helped me better understand how to use functions. In addition, I started on my initial Webpage, as well as worked with Rachael, my teammate, to develope a map of the environment we would be using for our project -- which was the 4th floor halways of the H. R. Bright Building on campus. Furthermore, I completed both literature assignments for the week, reading articles on Configuration Space as well as an extension on flocking methods.
Because there were some things I didn't fully understand, I was advised to revisit assignment 3 of the crash course. During this time, Jory was able to sit down and go over the assignment with me, to make sure I understood everything correctly. I also started on the first programming assignment for my actual research project. Also, as usual, I completed both literature readings for the week, reading, this week, more about roadmaps as well as simulating group behaviors in pursuit/evasion scenarios. Finally, Rachel and I placed over 60 markers around the hallways of fourth floor. These markers are what the robot uses to tell exactly where it is in the environment (formally called localization), each having a unique ID and specific location in the environment, based on a coordinate system we defined.
I completed two of three C++ programming assignments given to me by Jory, given to help my understanding of vectors, iterators and maps. I also completed both reading assignments (6 and 7) reading about roadmap sampling methods as well as behavior-based evacuation planning in robotics. Additionally, both Rachel and I continued 'decorating' the 4th floor hallways with markers for robot localization. In addition, we also added new lines of code to our project, inputting information about each individual marker.
I continued working with Rachel to input marker information into our code. In addition, this week marked our first testing of our code on one of the actual robots. While we were still in the process of debuging our code, I added samplers from the PMPL code for our robot to utilize. Also, I completed both literature review assignments for that week, having read in-depth articles about, both, obstacle-based and medial axis PRM samplers.
This week concluded our literature review assignments, us each having read on topics we chose (I read two papers on group formations in robotics). Also, I wrote and completed my progress report, with the help of Dr. Amato. Although it took me much longer than I had anticipated, this report helped build my confidence for writing the final research report. I also attended an enlightening Brown Bag session regarding how to prepare posters and presentations for the sessions at the end of the program. Rachel and I continued testing our robot code, testing the effects of different samplers. Additionally, I began the process of creating series of function in our code to utilizes our robots' (iRobot Creates) bump sensor.
I spent time continuing the process of developing a series a functions that would allow our the robot to recover from collision. Although, ideally, the robot is not supposed to run into obstacles, it has a tendency to do so anyway, due to the fact that it has poor odometry. Therefore, the bump function should work such that the robot is able to recover from contact with an obstacle, if its bumper is pressed. In addition to doing this, Rachel and I also created an outline for our final research paper, following the format outlined by the International Conference on Robotics and Automation (ICRA).
This week, Rachel and I worked on our poster as well as practiced for our presentation. Also, we made changes slight changes to our report abstract and continued to write our final report. We also began the process of incorporating a 'follower' behavior for a second robot. Although it will not be finished in time for our poster presentation, we will definitely include it in our future work, in order to demonstrate the full potential of our multi-robot caravanning. Additionally, we updated our map of the 4th floor.
During our final week, the CSCE-REU students presented at two poster session. The first was more of an exhibition, in that we explained our poster to those who were interested. However, for the USRG poster session, there will be actual judging involved. We accomplished during this week, including finishing our final research report. As the program comes to a close, I feel like I accomplished a lot this summer, as well as have a better understanding of what to expect from graduate school.
Parasol Home | Research | People | General info | Seminars | Resources
Parasol Lab, 301 Harvey R. Bright Bldg, 3112 TAMU, College Station, TX 77843-3112
email@example.com Phone 979.458.0722 Fax 979.458.0718
Department of Computer Science and Engineering | Dwight Look College of Engineering | Texas A&M University
Privacy statement: Computer Science and Engineering Engineering TAMU
Web Accessibility Policy and Law - Web Accessibility and Usability Standards - Contact Webmaster