day1 mon_aug5 + day2 tues_aug6

The tasks that I initially planned for this week:
1. Complete the planning exercises
The planning exercise is taking longer than expected. This is because I had a hard time figuring out the question and then understanding the solution. I was also not sure how helpful this exercise would be in our research as a whole and did not know how much priority I should give to completing it.

2. Read the two research papers listed in the Summer Plan document.
I have started reading the first research paper on Hierarchical Skeleton-Guided motion planning algorithms. The difficulty I have with this paper is trying to visualize the whole process and trying to understand how our research relates to it. It is not that the paper is not great, it is. But, I think it is more of my experience and understanding of the topic of Robotics Motion Planning. I want to go through these two papers slowly to understand and find out how to bridge the gap between the progress of our research and our tasks for the fall.

3. Complete the Duckietown Developer Manual.
I had trouble completing this because I was becoming unsure of how to use this guide for our purposes. There are a lot of software packages that have been pre-installed on docker images and we will be using them so that we can create our own code and push it onto the duckiebots and the duckietown simulator. However, I am having some difficulty envisioning what is already pre-installed and what we can add to their templates. They mention ROS and how we can use it, but I do not have much knowledge about ROS to understand the purpose of their ROS related instructions in the manual.

All in all, I think there are a list of things that it would be helpful to learn so that we can progress well.

1. ROS
2. What is the difference between the Parasol Library and ROS? Are they used together or does the Parasol Library not need ROS? (I talked to Professor Diane, and we thought about setting up a meeting with the people from Duckietown).
3. Check Assumption: we will be using the Parasol Library.
4. How are skeletons and planners encoded? What does the final result of a Path Planning Algorithm look like that the Motion Planning Algorithm will have to build on it?

day3 wed_aug7

A step by step plan:

1. Push through with the planning exercise
2. After completing the planning exercise, I will have a code that I hope I can use to use it on the Duckietown simulator and the Duckiebot.
3. Complete the developer manual.
4. Figure out how to push my planner onto the duckie-bot (Prof. Diane mentioned that we will not have to code our own collision checker, is there a library that I can try to implement?
5. If the planner does not work, I want to try implementing the parasol library in duckie-town. (Need to talk to Prof. Diane about it).

Alternatively:

1. Finish reading the papers.

todo today

1. FInish the planning exercises.

for tomorrow

1. finish following tutorials in the developer manual.

day4 thurs_aug8

Progress + Ideas for Next Steps

1. Completed Planning 1 Exercise. Will not be moving on with Planning 2 Exercise
2. Working on Completing the Developer Manual
   – (this time I am taking notes to see how we can change the development to implement our own libraries like the PPL?)
   – Goal: Find out if there is a way to implement libraries and algorithms from PPL
   – Today I learned that we need(?) ROS in our DTProjects so that we can push to code to the duckie-bot hardware (camera and wheels)
   
   Need More Knowledge:
   – Docker: to see how much independent code I can implement to duckie-town.
   – Research other duckie-town GitHub repositories and templates, and take notes about what they can do.
   – Simulator: 1. We can only run the simulator in an environment with specific libraries. 2. We are using keys to move around in the simulator. Is there a way to see the simulator from above and also automatically move the duckie-bot around?
   
   (It would be great if we could set up a meeting with people from Duckietown to see how much can be accomplished and to learn about their code base)
   
3. Learn more about implementing the PPL (download it on the Ubuntu Dell)
   
   Ideas to get ahead:
   1. Is it possible to duplicate the PPL repository (a github repo) so that I can implement DTProjects in it?
   – But is there a point?
   – PPL already has its own simulator.
   – What is the problem with the PPL that we are choosing Duckietown over PPL?
   – What is the purpose of using Duckietown if we will be implementing code from PPL anyways? To have a physical example of our algorithms?
   
4. Found a nice video explaining the challenges and opportunities of Robot Motion Planning
   https://www.youtube.com/watch?v=N7M4-XP7igk

Immediate Next Steps

Will be taking a break this Friday due to other commitments. Will be back on Saturday to make up for it and work on the Developer Manual.

Goals:

1. Finish the ROS template
2. Take notes about other templates and repositories
3. Reach out to Duckietown people for next steps.
4. Get a reply from Prof. Uwacu.