Difference between revisions of "Talk:Peek-a-Thief"

From ESE205 Wiki
Jump to navigation Jump to search
(Created page with "-Positives- *List of Challenges is very well thought out given the current project objective. *Overview is humorous and out there but intriguing enough to encourage further di...")
 
m (Wluer moved page Talk:Peek-a-Squirrel to Talk:Peek-a-Thief)
 
(2 intermediate revisions by 2 users not shown)
Line 1: Line 1:
-Positives-
+
==Proposal Review==
*List of Challenges is very well thought out given the current project objective.
 
*Overview is humorous and out there but intriguing enough to encourage further discovery about the project.
 
  
-Negatives-
+
===Strengths===
*Missing the Gantt chart, there will be multiple other times when images will need to be uploaded so it is an important process to figure out.
+
* List of Challenges is very well thought out given the current project objective.
*Include lab provided materials like the Raspberry Pi in the budget.
+
* Overview provides a clear intuition of the desired goals, and also implies an interest for discovery and experimentation.
*Difficult to imagine a safe and interactive demo were the device is used for its intended purpose. No way can you bring a squirrel into Lapota and put him in the box to take his picture.
 
  
-General Thoughts-
+
===Weaknesses===
*I know that it is the one of the motivating objects behind this project, but real, live squirrels cannot be used and trapped using Wash U money and resources. Furthermore, trapping of squirrels cannot be endorsed by the university as an activity for one of its classes. This is a cool general project idea but it must be re-envisioned without a live animal as the focal point for the pictures.
+
I am afraid that this project, in its current form, cannot be approved.
 +
WashU has strong regulations when live animals are involved in research and educational experiences [https://research.wustl.edu/Offices_Committees/DCM/Pages/default.aspx].
 +
Since dealing with those regulations is unrealistic for a lower-level undergraduate course like ours, you will have to reframe your project in a way in which no animals are involved.
 +
For example, you can design a security camera that works under particular considerations (say, to secure the interior of a drawer).
 +
 
 +
I will continue reviewing the rest of the proposal so you can understand the main weaknesses, yet these comments will likely become invalid once you modify the goal of your project.
 +
 
 +
There is a disconnection between the sensors you want to install, and the goals you want to achieve.
 +
For example, you want to take ''"high definition"'' photos at 0.5 meters (or less, I assume).
 +
That is a challenging task even for advanced expensive cameras, unless a special macro lens is used.
 +
In particular, none of the Pi cameras are capable of performing in these conditions.
 +
Similarly, you want to use ultrasonic sensors to detect when an animal has entered the box, but those sensors can be triggered by many things that are not necessarily squirrels, such as leaves or insects.
 +
Also, no weather sensor is listed in the budget, yet one of the objectives is to stop the device from operating under bad weather (I'm not sure how turning off a camera will avoid a short circuits due to rain though).
 +
Your choice of sensors should match the needs of your project, and if no sensor can provide the performance you are looking for then you should probably modify the desired goals.
 +
 
 +
Some of your challenges are particular hard to solve, or even do not have a clear solution.
 +
Just as an example, there is no way you can design an efficient way of luring animals into an enclosure unless you understand how the animals behave.
 +
Since you want to focus on squirrels, then you will have to spend a significant amount of time reading and observing squirrels, something that is hardly a good use of your time in this course.
 +
 
 +
Other comments:
 +
* Gantt chart is missing. By now you should have learned how to properly upload and link an image to your wiki. Take a look at other groups' project source code to figure out how they solved this problem.
 +
* Include lab provided materials, like the Raspberry Pi, in the budget.
 +
* This project is almost impossible to demo. Are you going to bring squirrels to the Lopata gallery for the demo session?
 +
 
 +
 
 +
[[User:Hgonzale|Humberto]] 00:20, 8 February 2017 (CST)

Latest revision as of 21:57, 11 February 2017

Proposal Review

Strengths

  • List of Challenges is very well thought out given the current project objective.
  • Overview provides a clear intuition of the desired goals, and also implies an interest for discovery and experimentation.

Weaknesses

I am afraid that this project, in its current form, cannot be approved. WashU has strong regulations when live animals are involved in research and educational experiences [1]. Since dealing with those regulations is unrealistic for a lower-level undergraduate course like ours, you will have to reframe your project in a way in which no animals are involved. For example, you can design a security camera that works under particular considerations (say, to secure the interior of a drawer).

I will continue reviewing the rest of the proposal so you can understand the main weaknesses, yet these comments will likely become invalid once you modify the goal of your project.

There is a disconnection between the sensors you want to install, and the goals you want to achieve. For example, you want to take "high definition" photos at 0.5 meters (or less, I assume). That is a challenging task even for advanced expensive cameras, unless a special macro lens is used. In particular, none of the Pi cameras are capable of performing in these conditions. Similarly, you want to use ultrasonic sensors to detect when an animal has entered the box, but those sensors can be triggered by many things that are not necessarily squirrels, such as leaves or insects. Also, no weather sensor is listed in the budget, yet one of the objectives is to stop the device from operating under bad weather (I'm not sure how turning off a camera will avoid a short circuits due to rain though). Your choice of sensors should match the needs of your project, and if no sensor can provide the performance you are looking for then you should probably modify the desired goals.

Some of your challenges are particular hard to solve, or even do not have a clear solution. Just as an example, there is no way you can design an efficient way of luring animals into an enclosure unless you understand how the animals behave. Since you want to focus on squirrels, then you will have to spend a significant amount of time reading and observing squirrels, something that is hardly a good use of your time in this course.

Other comments:

  • Gantt chart is missing. By now you should have learned how to properly upload and link an image to your wiki. Take a look at other groups' project source code to figure out how they solved this problem.
  • Include lab provided materials, like the Raspberry Pi, in the budget.
  • This project is almost impossible to demo. Are you going to bring squirrels to the Lopata gallery for the demo session?


Humberto 00:20, 8 February 2017 (CST)