Week 2

As I said in my previous blog post, I transitioned into week 2 with Labor Day holiday on the Monday and I moved on Tuesday. I went straight into the module 2 of the pmpl crash course which was a lot more hands-on than modules 0 and 1. Although my plan to complete the module did happen, I was still able to get through half of it and also had the article discussion which covered another paper (one of the three I had to read for module 1). The first section of module 2 required that I worked on a motion planning problem of my choice; since my project is about proteins which undergo folding, I decided to choose a box folding problem. Unfortunately, I couldn’t generate a roadmap for the problem, not even after trying all the samplers. The problem is that all the samplers are probabilistic roadmap strategies and so it was impossible for me to know whether the program would generate the roadmap and path if I let it run for a long time; I lost patience after 5 hours.

Figure 1 showing the nodes generated for the box folding problem

Since my own project wasn’t much helpful, I switched to using a 3D environment example that was already coded when I was working on the next section. Module 2.2 was about changing the samplers for the code so I could compare their efficiency, an exercise that helped me to understand the code and just the topic of motion planning better. I had 5 samplers to compare: Uniform, Gaussian, obstacle-based, bridge and medial-axis samplers.

image Figure 2 shows the nodes added by the Bridge sampler.

image Figure 3 shows the nodes added by Gaussian sampler

image Figure 4 showing the nodes added by the Medial-axis sampler

image Figure 5 showing the nodes added by the obstacle-based sampler

image Figure 6 showing the nodes added by the uniform sample

Further analysis of statistics generated for each sampler helped to compare the samplers. For next week, I plan to get back on track and finish the second module and start module 3. I also hope to gain a better understanding of the probabilistic roadmap article by reading it again and attending a presentation for it.

Written on May 31, 2021