Proposal 5
2) no
3)6
4)7

Your proposal needs some work. Overall, I thought your idea was a great continuation of the

previous work done with buttons in the Developmental Robotics lab, but you’ll need to articulate what

you’re proposing in a more understandable manner. Below, I’ll detail what I liked and what I think you

can improve on, as well as concerns that I have with this proposal.

First of all, I liked the overall idea of your project. It’s a clear continuation of Vlad’s button

pushing algorithm he’s presented in class and I think this is a great adaptation of that previous work. I

like that you considered using the exiting set-up, I think it will save you a lot of time and shows that

you’ve taken time to consider the experimental set-up. Also, I think your proposed approach should work

for this application, though I was a little worried at first as to whether the robot can detect the small tactile

change of pushing a button. Though your description of having the robot hit the wall and then push,

you’ve relieved my worries and I think this project can work. Furthermore, you seem to have a good hold

on what machine learning algorithms you are going to use and I really liked that you could mention the

ones you intended to use by name. Finally, I think the general idea is very doable within the remaining

time of the course.

There are, however, many things that I think you need to change, and some things you could

change to make this proposal better. First of all, get to the benefits of your research faster. You seem to

explore the works of predecessors in the field, but you don’t look at why this research is important. In the

introduction, you should engage the reader and peak his interest. Also, you have several occasions where

you detail related work outside the related work section. I think it’s fine if you reference these sources

outside the section, but keep it short and concise.

Thirdly, and I cannot emphasize this enough, do not mention Wikipedia in the paper. Wikipedia

is a great place to get started in your research, but it is not a respected source, since none of the

information is verifiable and anyone can change the information there. If you want to mention something

you read on Wikipedia, find the source the Wiki used. Next is your formatting, though only in one

section. In section 4, Research Contents, you make a complete break from your previous format.

Consider changing this, as it makes no sense to do this in the middle of your paper.

Two final changes that I insist that you make are: add more pictures and have this proofread by a

native English speaker. Pictures are a great way to illustrate your proposed algorithm, and a wonderful

way to break up the monotony of plain text. I think your descriptions are a little confusing and that

pictures could go a long way in alleviating this confusion. The second recommendation is just a matter of

making the paper a bit more readable. It is clear to me that English is not you native language, and I think

your proposal suffers because of this. Just have an American friend look over and make slight corrections

to your grammar and it will go a long way to making your paper more readable.

Overall, I think that your proposal is so-so and if I had to give it a grade it would be around a C

plus. I also believe that with a little work, this could be a great document and a great project can come

out of it.