Talk:WALL-E

Proposal Evaluation

Objective is not associated to an application

The proposal states that the vehicle will be autonomous, will avoid chairs and tables, and will run around a room. Although that is not a bad set of objectives, it is unclear why do we want a vehicle with these features. In order words, your objective lacks a clear motivation. For example, do you want the vehicle to find something in particular? to deliver an object to a point? to create a map of the room?

Avoid duplicating efforts with past projects

Your project uses similar tools, and aims to similar goals, than the Leap Controlled RC Car project last semester. I assume that you discussed this with Andrew in your meeting. If that is the case, then you should cite the previous project's work, explain what you will and won't use from that project, and explain what you will improve.

It is perfectly ok to work on top of another project (either from this course or from somewhere else, such as a webpage), as long as you provide clear references and you explain the improvements you will implement.

Not enough sensors to detect obstacles

Currently your budget considers one ultrasonic distance sensor. With that sensor you can find walls or big obstacles, but you will not be able to find chair or table legs, as explained in your objectives, since thin objects do not reflect well sound waves. Moreover, a single sensor will not allow you to make a control decision to continue moving around the room. Indeed, once you find something in front of you, what is the control action? You won't have information if moving left is better than moving right, nor how many angles.

You can fix this problem by either adding more sensors (e.g., an array of 5 distance sensors pointing at different angles, but they need to be synchronized to avoid cross-talk between the speakers/microphones), installing your sensors over a servo motor (so it can be rotated), or design your own sensor (clearly the hardest option, but still doable within one semester).

Gantt chart is unsatisfactory

The weeks should have actual dates, not just numbers. There is no clear division of the tasks among the members in the chart. Also, some tasks (such as "Senser control") are not obviously related to your objectives. How do you plan to control the sensor?

Overall this is a good proposal, but needs a bit of improvement. Your modified proposal should be ready on Friday September 23 at 5pm.

-- Humberto 17:02, 19 September 2016 (CDT)