Author: Fausto Vega, Email: vegaf1@unlv.nevada.edu
Date Last Modified: 06/30/2019

These are the main robot that they have in the lab, and they are called the PR2 and Baxter robot. These robots are currently being used for picking and placing tasks as well as collaborative human robot interaction. This is a computer science laboratory, so they primarily focus on computer vision and algorithms for sequential tasks. For example, teaching the robot how to make tea or a sandwich with the toys placed on the table. It is also something new seeing psychology have a role in robotics. This lab also conducts human robot interaction studies in which they create an experiment and then call volunteers to experience it and fill out a survey on it. This was the project they wanted me to do in the beginning, yet I'm not interested in psychology so I switched over. The PR2 robot is an impressive piece of hardware as it can do various tasks from path planning to even plugging in its own extension cord for charging. They are using the ROS package Move It for all of the manipulator controls.

In terms of lab culture, the lab is not that organized (Dongbin would go crazy haha). I spent time looking for an extension cord because everything is just randomly placed in certain areas.

The third week I continued to work on what I proposed last week. I successfully made a subscriber that subscribed to the messages being sent by the laser package, and it states the number of people that are in the frame. I am currently working on making a service which will print the locations of the people whenever the service is called. Next week I will try to get the Pioneer started up and learn how to operate it. I will also install the navigation and localization packages it requires and study their purpose. This weeks workshops consisted of peer review and editing which provide useful insight when editing papers. I also finished the related work section of the paper deliverable this week.

This week I completed the service that will display the position of the person whenever it is called using rosservice call. I also learned the basics of the Pioneer robot such as how to ssh into it as well as the general packages it uses for control. It currently does not have a navigation stack applied to it, therefore I will need to search for existing packages and modify them to work with the topics that the pioneer robot is already publishing. One idea is to send command velocity messages over to an existing topic that will move the wheels to the goal. I also have looked into machine learning to learn the basics of exactly what it is. Since this lab is computer science related, all the projects incorporate some sort of machine learning. I at least want to understand what is going on when these algorithms are used therefore I decided to do some reading. I then configured a laptop with nvidia drivers and cuda in order to use an existing package called Darknet YOLO. This package is for object detection and contains existing weights for certain objects. I was surprised with the open source resources that are available. I also was reading up on Python basics and realized that it is similar to matlab syntax. This week we did a peer review on the background section of the paper deliverable as well as a workshop on Next week, the goal is to get the Pioneer running with some localization and navigation packages.

I learned that I get stressed out with computer science in general, but that I should just be patient as the end results will come. With so many errors in the code, I spend most of the time debugging the code, yet by doing this I learn the proper structure and will avoid making the same mistake in the future. Overall it is making me a stronger engineer as most ME's do not go through a programming class.

I have also learned that I try to figure out the problem myself and stay stuck for a couple of days instead of asking for help. Yet, it is good to ask for as there is a set timeline for the project, and the more progress, the better. It is not like I am asking for them to do the project for me, yet just guidance in the right direction. That is something I need to do whenever I spend too long for a problem.

I met Dr. Louis this week and he gave me an introduction on his work. It involves Unity with Dr. La's bridge inspection robot. His grad student was out of town, so I will get to see a demo this upcoming Friday and will include it in the next blog.